Bibliography

Foundational Thermodynamics

Anderson, Philip W. “More Is Different.” Science 177, no. 4047 (1972): 393-396. The foundational argument that each level of complexity involves a new symmetry breaking: the laws of one level do not determine the organizing principles of the next. “The ability to reduce everything to simple fundamental laws does not imply the ability to start from those laws and reconstruct the universe.” Referenced in Chapter 12 for the hierarchy of broken symmetries.

Boltzmann, Ludwig. Lectures on Gas Theory (1896). The statistical mechanics foundation that gave us S = k log W.

Clausius, Rudolf. “On the Mechanical Theory of Heat.” Annalen der Physik (1865). The original formulation of entropy and the two laws of thermodynamics.

Eddington, Arthur. The Nature of the Physical World (1928). Source of the famous quote about collapsing in “deepest humiliation” if your theory violates the Second Law.

Gibbs, Josiah Willard. Elementary Principles in Statistical Mechanics (1902). Formalization of free energy and thermodynamic potentials.

Gibbs, Josiah Willard. “On the Equilibrium of Heterogeneous Substances.” Transactions of the Connecticut Academy (1876-1878). Foundation of chemical thermodynamics.

Goldstone, Jeffrey. “Field Theories with ‘Superconductor’ Solutions.” Il Nuovo Cimento 19 (1961): 154-164. Proved that spontaneous breaking of a continuous symmetry necessarily produces massless bosonic excitations (Goldstone modes). Extended with Salam and Weinberg in Physical Review 127 (1962): 965-970. In the coordination framework (Online Annex §4.2), Goldstone modes correspond to the new collective possibilities that emerge when agents break imposed symmetry and self-organize.

Gross, David J. “The Role of Symmetry in Fundamental Physics.” Proceedings of the National Academy of Sciences 93, no. 25 (1996): 14256-14259. Nobel laureate’s argument that symmetry principles are more fundamental than dynamical laws: the laws are outputs, the symmetries are inputs. “It is only slightly overstating the case to say that physics is the study of symmetry.”

Helmholtz, Hermann von. “Die Thermodynamik chemischer Vorgänge” (1882). The Helmholtz free energy formulation.

Maxwell, James Clerk. Theory of Heat (1871). Contains the original “Maxwell’s Demon” thought experiment, which Maxwell first described in an 1867 letter to Peter Guthrie Tait.

Noether, Emmy. “Invariante Variationsprobleme.” Nachrichten von der Gesellschaft der Wissenschaften zu Göttingen (1918): 235-257. Proved that every continuous symmetry of a physical system corresponds to a conserved quantity. The foundation for deriving conservation laws from action symmetries. English translation: M. A. Tavel, Transport Theory and Statistical Physics 1 (1971): 186-207.

Tsallis, Constantino. “Possible Generalization of Boltzmann-Gibbs Statistics.” Journal of Statistical Physics 52 (1988): 479-487. Introduces non-extensive entropy (q-entropy), generalizing the Boltzmann-Gibbs framework to systems with long-range interactions, memory effects, or fractal phase-space structure.

Non-Equilibrium Statistical Mechanics

Crooks, Gavin E. “Entropy Production Fluctuation Theorem and the Nonequilibrium Work Relation for Free Energy Differences.” Physical Review E 60, no. 3 (1999): 2721-2726. Generalization of Jarzynski’s equality; proves time-reversal symmetry of entropy production.

Jarzynski, Christopher. “Nonequilibrium Equality for Free Energy Differences.” Physical Review Letters 78, no. 14 (1997): 2690-2693. Foundational result showing free energy differences can be determined from non-equilibrium work measurements.

Kappen, Hilbert J. “Path Integrals and Symmetry Breaking for Optimal Control Theory.” Journal of Statistical Mechanics (2005): P11011. Shows that a class of stochastic optimal control problems can be solved via path integrals; the optimal policy emerges from summing over all future trajectories weighted by their costs. Control cost equals KL divergence from passive dynamics. Extended to multi-agent coordination in Kappen, Gomez & Opper, Machine Learning 87 (2012): 159-182.

Onsager, Lars and Stefan Machlup. “Fluctuations and Irreversible Processes.” Physical Review 91, no. 6 (1953): 1505-1512. Defines the action functional for Gaussian diffusion processes: the path integral for thermodynamic systems. The most probable trajectory of a stochastic system extremizes this action, exactly as in classical mechanics. Extended to non-Gaussian systems by Dürr & Bach (1978).

Pressé, Steve, Kingshuk Ghosh, Julian Lee, and Ken A. Dill. “Principles of Maximum Entropy and Maximum Caliber in Statistical Physics.” Reviews of Modern Physics 85 (2013): 1115-1156. Extends Jaynes’s Maximum Entropy to trajectory space: the least biased distribution over paths maximizes path entropy subject to constraints. Recovers Onsager reciprocal relations, Green-Kubo transport coefficients, and Prigogine’s minimum entropy production as special cases. The unifying variational principle for non-equilibrium statistical mechanics.

Rubino, Giulia, Gonzalo Manzano, and Časlav Brukner. “Quantum Superposition of Thermodynamic Evolutions with Opposing Time’s Arrows.” Communications Physics 4, 251 (2021). Demonstrates that forward and time-reversal thermodynamic processes can be placed in quantum superposition, with the arrow of time emerging from measurement of entropy production. When entropy change is small, the two directions interfere, producing work distributions unreachable by any classical mixture.

Seifert, Udo. “Stochastic Thermodynamics, Fluctuation Theorems and Molecular Machines.” Reports on Progress in Physics 75, no. 12 (2012): 126001. Comprehensive review unifying non-equilibrium statistical mechanics results.

Still, Susanne, David A. Sivak, Anthony J. Bell, and Gavin E. Crooks. “Thermodynamics of Prediction.” Physical Review Letters 109 (2012): 120604. Proves that any system driven by a fluctuating environment pays a thermodynamic cost proportional to the mutual information its state retains about past inputs that do not predict the future, the quantity the authors call “nostalgia.” A thermodynamically efficient memory minimizes nostalgia, balancing what it stores against what that storage can predict. The information-theoretic foundation for the optionality argument that memory is preparation for the future rather than a record of the past. Referenced in Chapters 6 and 18.

Dissipative Structures and Self-Organization

Nguyen, Han P., et al. “DNA-inspired molecular solar thermal energy storage with high energy density.” Science (2026). DOI: 10.1126/science.aec6413. Demonstrates a 2-pyrimidone derivative that absorbs UV light and twists into a strained Dewar isomer, a molecular configuration inspired by UV damage to DNA thymine bases. Achieves 1.65 MJ/kg energy storage density (nearly double lithium-ion batteries), with a half-life of 481 days at room temperature. The molecule is liquid at room temperature, eliminating the solvent penalty that plagued earlier molecular solar thermal (MOST) systems. Key limitation: single-digit quantum yield, with most absorbed photons dissipating as immediate vibrational heat rather than forming the storage isomer. Referenced as ^6a in Chapters 2 and 6.

Nicolis, Grégoire, and Ilya Prigogine. Self-Organization in Nonequilibrium Systems: From Dissipative Structures to Order through Fluctuations (Wiley, 1977). The technical statement of how systems held far from equilibrium spontaneously generate ordered “dissipative structures” sustained by the throughflow of energy and matter. The primary source behind the popular Order Out of Chaos; cited in the Chapter 4 dialogue.

Prigogine, Ilya. From Being to Becoming: Time and Complexity in the Physical Sciences (1980). Technical treatment of non-equilibrium thermodynamics.

Prigogine, Ilya. Order Out of Chaos: Man’s New Dialogue with Nature (1984, with Isabelle Stengers). Nobel Prize 1977. The foundational work on dissipative structures.

Thirumalaiswamy, Amruthesh, John C. Crocker, Robert Riggleman, et al. “Slow relaxation and landscape-driven dynamics in viscous ripening foams.” Proceedings of the National Academy of Sciences (2025). Discovery that bubbles in wet foam continuously reorganize through configuration space using the same gradient descent mathematics that trains deep learning systems. Both systems explore flat regions rather than settling into deep valleys; flexibility outperforms rigid optimization. The researchers speculate that “learning, in a broad mathematical sense, may be a common organizing principle across physical, biological and computational systems.” Suggests that adaptive organization follows universal patterns across substrates.

Yu, Zhenwei, Yuchen Chen, Frank F. Yun, David Cortie, Lei Jiang, and Xiaolin Wang. “Discovery of a Voltage-Stimulated Heartbeat Effect in Droplets of Liquid Gallium.” Physical Review Letters 121 (2018): 024302. Demonstrates voltage-controlled oscillation of gallium droplets from 0 to 610 beats per minute in NaOH electrolyte. Unlike previous mercury experiments, the symmetry-breaking forces propel the droplet several millimeters at ~1 cm/s. The controllable rhythm opens applications in fluid-based timers, soft robotics actuators, and organ-chip pumps. Key finding: the capacity for oscillation is intrinsic to the electrochemistry; voltage modulates but cannot create the rhythm where gradient structure does not support it.

Zhang, Yujie, et al. “Repetitive Deformation of Ga-Based Liquid Metal in Acidified CuCl₂ or FeCl₃ Solution.” Journal of Chemical Education (2024). Demonstrates the “gallium beating heart,” a bench-scale dissipative oscillator where electrochemical gradients drive rhythmic surface tension changes in liquid metal. Safer classroom alternative to the mercury beating heart discovered by Lippmann in 1873. Illustrates substrate-independence: the same dissipative principle that organizes thermal convection cells appears in electrochemical systems.

Constructal Law

Bejan, Adrian. “Constructal-theory network of conducting paths for cooling a heat generating volume.” International Journal of Heat and Mass Transfer 40 (1997): 799-816. The original constructal law paper.

Bejan, Adrian. Design in Nature: How the Constructal Law Governs Evolution in Biology, Physics, Technology, and Social Organization (2012, with J. Peder Zane). The accessible introduction to constructal theory.

Bejan, Adrian and Sylvie Lorente. The Physics of Life: The Evolution of Everything (2016). Extension of constructal law to biological and social systems.

Bejan, Adrian, H. Almahmoud, U. Gunes, H. E. Fakhari, and P. Mardanpour. “Evolution and Irreversibility: Two Distinct Phenomena and Their Distinct Laws of Nature.” Physics of Life Reviews 50 (2024): 103–116. DOI: 10.1016/j.plrev.2024.06.014. Sharpens the distinction the book draws on in Chapter 3: evolution (the changing configuration of a flow system) is governed by the Constructal Law, while irreversibility (dissipation) is governed by the Second Law. Two distinct phenomena, two distinct laws. Cited in Chapter 3.

Frank, Robert H. Passions Within Reason: The Strategic Role of the Emotions (1988). Norton. Argues that genuine emotions (gratitude, guilt, love) are strategically superior to calculated mimicry as commitment devices: where detection of faking is possible, selection installs the real thing. The functional distinction between authentic and performed care erodes because authentic care is evolutionarily stable and fake care is not. Cited in Chapter 20.

Meng, Xiangyi, Albert-László Barabási, et al. “Surface optimization governs the local design of physical networks.” Nature 649(8096) (2026). Discovered that biological branching networks (neurons, blood vessels, plant roots, coral) optimize for surface area in three dimensions, not just path length. The mathematical framework maps onto high-dimensional Feynman diagrams from string theory: the same optimization problem, different substrates. Predicts trifurcations and orthogonal branches as stable configurations when link thickness makes surface costs dominant. In human brain data, 98% of orthogonal sprouts end in synapses. Real organisms run ~25% longer than theoretical minimum, balancing surface efficiency against functional demands.

Rabinovich, Mikhail I., Pablo Varona, Allen I. Selverston, and Henry D.I. Abarbanel. “Dynamical principles in neuroscience.” Reviews of Modern Physics 78, 1213 (2006). Foundational review of winnerless competition dynamics: heteroclinic networks connecting metastable saddle states via transient orbits. Applies to sequential processing in sensory systems, motor pattern generation, and cognitive chunking.

Thompson, D’Arcy Wentworth. On Growth and Form. Cambridge University Press (1917; revised edition 1942). Argued that physical forces (surface tension, mechanical stress, diffusion) shape biological form alongside natural selection. Thompson’s thesis that organisms obey the same mathematical laws as non-living structures anticipated constructal theory by eight decades, and is increasingly vindicated by modern biophysics. Referenced in Chapter 3.

Valiant, Leslie G. “Evolvability.” Journal of the ACM 56(1), Article 3 (2009). Proves that Darwinian evolution is a restricted case of PAC (Probably Approximately Correct) learnability: populations that evolve under selection are performing a constrained form of computational learning. The formal subsumption establishes that evolution’s search through fitness landscapes is mathematically a learning algorithm. Cited in Chapter 14.

Voit, Maximilian and Hildegard Meyer-Ortmanns. “Dynamics of nested, self-similar winnerless competition in time and space.” Physical Review Research 1, 023008 (2019). Constructs n levels of nested, self-similar competitive dynamics using generalized Lotka-Volterra equations with a recursive block-matrix predation structure. The same rock-paper-scissors game plays out at every hierarchical level. On a spatial grid with diffusion, the temporal hierarchy translates to nested spirals. A death-rate bifurcation parameter controls hierarchy collapse: past a threshold, fine-grained diversity vanishes. Noise prevents the cycling from freezing into permanent dominance. Referenced in Chapters 3, 5, 8, and 9.

Voit, Maximilian and Hildegard Meyer-Ortmanns. “A hierarchical heteroclinic network.” European Physical Journal Special Topics 227, 1101 (2018). Derives the eigenvalue conditions on predation rates that control dwell time within heteroclinic sub-cycles and direct the trajectory through a desired path in the hierarchical network. The analytical foundation for the 2019 spatial extension.

Watson, Richard A. and Eörs Szathmáry. “How Can Evolution Learn?” Trends in Ecology and Evolution 31(2): 147-157 (2016). Demonstrates formal equivalences linking evolutionary processes to learning principles: selection in asexual and sexual populations maps onto Bayesian learning; evolving gene-regulatory networks map onto neural-network training. The shared mathematical structure is precise; the authors claim structural parallel, not process identity. Cited in Chapter 14.

Branched Flow

Heller, Eric J., Ragnar Fleischmann, and Tobias Kramer. “Branched flow.” Physics Today 74 (2021): 44-51. Comprehensive review tracing branched flow across scales from quantum electrons to ocean waves. Documents the tsunami connection (2011 Tōhoku showed clear branch structure) and establishes the phenomenon’s universality. The same mathematical framework describes focusing in semiconductor nanostructures and in thousand-kilometer oceanic wave propagation. Describes the phenomenon as “on the way to chaos, but not there yet.”

Patsyk, Anatoly, Uri Sivan, Mordechai Segev, and Miguel A. Bandres. “Observation of branched flow of light.” Nature 583 (2020): 60-65. First demonstration of branched flow in optics using soap films as the randomly varying medium. The phenomenon was theoretically predicted but had never been observed with light. It was observable with a laser pointer and a soap bubble, equipment available for decades, waiting for someone to look. Hiding in plain sight.

Topinka, M.A., B.J. LeRoy, R.M. Westervelt, et al. “Coherent branched flow in a two-dimensional electron gas.” Nature 410 (2001): 183-186. The discovery paper. Electrons injected through quantum point contacts into semiconductors spontaneously organized into branching filaments rather than spreading diffusely, even though no channels existed in the material. Established branched flow as a distinct transport phenomenon caused by smooth random potential variations.

Electrostatic Ecology

Chiba, Takuma, et al. “Caenorhabditis elegans transfers across a gap by electrostatic force.” Current Biology (2023). Japanese researchers filmed nematode nictation launches with high-speed cameras, discovering that C. elegans standing on its tail can be pulled across air gaps at up to 1 m/s by the electrostatic charge carried by passing insects. Established that the behavior is purposive rather than accidental, with worms adopting nictation posture and collectively forming towers of 80-200 individuals to increase launch probability.

Clarke, Dominic, Heather Whitney, Gregory Sutton, and Daniel Robert. “Detection and learning of floral electric fields by bumblebees.” Science 340(6128) (2013): 66-69. The discovery paper establishing that bumblebees can detect and discriminate between the electrostatic fields of flowers. Demonstrated that floral electric fields change after a bee visit and reset over minutes, providing a dynamic signal of nectar availability. The first evidence that terrestrial animals use electric fields in ecological contexts.

England, Sam J. and Daniel Robert. “The ecology of electricity and electroreception.” Biological Reviews 98(4) (2023): 1193-1223. Comprehensive review establishing electrostatic ecology as a recognized subfield. Documents the roles of ambient electric fields in pollination, predation, dispersal, and sensory ecology across arthropod taxa. Argues that atmospheric electric fields constitute a pervasive ecological factor comparable in significance to temperature or humidity for small organisms.

England, Sam J. and Daniel Robert. “Prey can detect predators via electroreception in air.” Proceedings of the National Academy of Sciences 121, no. 23 (2024): e2322674121. Demonstrates that caterpillars respond defensively to the electric fields generated by approaching wasps, constituting a novel predator detection modality operating through electrostatic rather than acoustic or visual channels. The body hairs of caterpillars function as electrostatic sensors, detecting field distortion caused by the charged bodies of flying insect predators.

Estrada-Peña, Agustín, et al. “Electrostatic attraction of ticks to hosts.” (2023). Demonstrated that Ixodes ricinus ticks exploit the static charge accumulated on mammalian fur to bridge air gaps significantly larger than their body length, dramatically increasing attachment probability. Extended the electrostatic ecology framework from flying insects to terrestrial parasites.

Morley, Erica L. and Daniel Robert. “Electric fields elicit ballooning in spiders.” Current Biology 28(14) (2018): 2324-2330. Demonstrated spider ballooning in controlled laboratory environments with zero airflow, confirming that atmospheric electric fields alone provide sufficient force for launch and sustained flight. Spiders responded to experimentally applied electric fields by adopting the characteristic tiptoe posture and becoming airborne, settling a longstanding debate about the mechanism of spider aerial dispersal.

Ruiz, Victor J., et al. “Electrostatic interactions enhance parasitic nematode jumping.” Proceedings of the National Academy of Sciences (2025). Emory University and UC Berkeley researchers tracked predatory leaps of Steinernema carpocapsae onto fruit flies, demonstrating that electrostatic induction from the fly’s ~800V charge accounts for the vast majority of successful jumps. With electrostatic assist, 80% of jumps connect; simulations without electrostatic forces predict only 5% success, below the threshold at which the behavior could have been selected for. The first demonstration that electrostatic ecology enables predation.

Cosmic Evolution and Energy Rate Density

Azarian, Bobby. “The Mind Is More Than a Machine.” Noema Magazine (9 June 2022). Traces the sequence from Gödel’s incompleteness theorem through Hofstadter’s strange loops to the claim that self-reference via self-modeling produces consciousness. Summarizes the convergence across Global Workspace Theory, IIT, and recurrent processing on feedback loops as the architectural precondition for conscious experience. Includes Simon DeDeo’s observation that progress in artificial general intelligence may require taking seriously the relativity implied by self-reference.

Azarian, Bobby. The Romance of Reality: How the Universe Organizes Itself to Create Life, Consciousness, and Cosmic Complexity (BenBella Books, 2022). Argues from neuroscience and complexity theory that the universe has an inherent tendency toward increasing complexity and consciousness through hierarchical emergence. Draws on Kurzweil, Koch, and Kauffman. Convergent with Vanchurin’s physics-first framework and the BEDS thermodynamic approach; the convergence is independent.

Chaisson, Eric J. Cosmic Evolution: The Rise of Complexity in Nature (2001). The foundational work on energy rate density (φm) as a measure of complexity.

Chaisson, Eric J. Epic of Evolution: Seven Ages of the Cosmos (2006). Accessible presentation of cosmic evolution across all scales.

Chaisson, Eric J. “The Natural Science Underlying Big History.” The Scientific World Journal, 2014, 384912. Synthesizes cosmic, biological, and cultural evolution under the energy rate density framework, situating Big History within a rigorous thermodynamic context.

Thermodynamics and Life

Bardeen, John, Leon N. Cooper, and J. Robert Schrieffer. “Theory of Superconductivity.” Physical Review 108 (1957): 1175–1204. The BCS theory: phonon-mediated Cooper pairing produces a macroscopic quantum condensate with zero electrical resistance. The transition from normal to superconducting state is a phase transition in the coordination class of charge carriers. Nobel Prize in Physics, 1972.

Chastain, Erick, Adi Livnat, Christos Papadimitriou, and Umesh Vazirani. “Algorithms, games, and evolution.” Proceedings of the National Academy of Sciences 111(29) (2014): 10620–10623. Population genetics equations for allele frequency change under weak selection are mathematically identical to the multiplicative weights update algorithm from computer science and game theory. The algorithm’s objective function maximizes mean fitness plus Shannon entropy (diversity), providing formal grounds for the claim that evolution values genetic diversity intrinsically.

Cooper, Leon N. “Bound Electron Pairs in a Degenerate Fermi Gas.” Physical Review 104 (1956): 1189–1190. Showed that an arbitrarily weak attractive interaction between electrons near the Fermi surface produces bound pairs (Cooper pairs), the key mechanism underlying superconductivity. In this book’s framework, Cooper pairing is coordination by invitation at the quantum level: individual electrons that scatter resistively (coercion-dominated transport) spontaneously pair into a coherent condensate (trust-dominated transport) below a critical temperature. See Chapter 19 for the Trust Attractor interpretation.

Corning, Peter A. The Synergism Hypothesis: A Theory of Progressive Evolution (1983). Synergy as thermodynamic efficiency grounding evolutionary innovation.

Diener, Theodor O. “Potato spindle tuber ‘virus’ IV. A replicating, low molecular weight RNA.” Virology 45: 411-428 (1971). Identified the agent of potato spindle tuber disease as a naked circular RNA molecule carrying no protein coat and encoding no proteins: the smallest known class of infectious agents in biology. Diener coined the term “viroid” for these subviral pathogens. For fifty years viroids were considered confined to flowering plants; metatranscriptomic surveys (Lee et al. 2023) later revealed them across all domains of life. Referenced in Chapter 6.

Elowitz, Michael B., Arnold J. Levine, Edward D. Siggia, and Peter S. Swain. “Stochastic gene expression in a single cell.” Science 297 (2002): 1183–1186. Foundational demonstration that gene expression is inherently stochastic. Genetically identical cells express genes at different levels by chance, generating phenotypic diversity from molecular noise. See also Süel et al., Nature 440 (2006): 545–550, and Çağatay et al., Cell 139 (2009): 512–522, which showed noisier genetic circuits confer survival advantage: noise selected for, not merely tolerated.

England, Jeremy. “Dissipative adaptation in driven self-assembly.” Nature Nanotechnology 10 (2015): 919-923. Shows how driven systems spontaneously organize to resonate with external energy sources, maximizing absorption and dissipation.

England, Jeremy. “Statistical physics of self-replication.” Journal of Chemical Physics 139 (2013): 121923. The seminal paper on dissipation-driven adaptation.

Evans, Constantine Glen, Jackson O’Brien, Erik Winfree, and Arvind Murugan. “Pattern recognition in the nucleation kinetics of non-equilibrium self-assembly.” Nature 625 (2024): 500–507. Demonstrated that competitive nucleation in a 917-component DNA tile system performs neural-network-style pattern recognition, correctly classifying 18 grayscale images into three categories through thermodynamic phase boundaries alone. Established formal connections between multicomponent self-assembly, Hopfield associative memories, and winner-take-all dynamics. Key result: ubiquitous physical phenomena such as nucleation hold powerful information-processing capabilities in high-dimensional multicomponent systems.

Frank, F. C. “On Spontaneous Asymmetric Synthesis.” Biochimica et Biophysica Acta 11 (1953): 459-463. The foundational model for biological homochirality: autocatalysis combined with mutual inhibition of mirror-image molecular forms amplifies small initial fluctuations to near-complete single-handedness. In this book’s framework, homochirality is the Trust Attractor at the molecular level: coordinated (same-chirality) populations outcompete mixed ones because every molecular interaction works. Referenced in Chapter 17.

Heller, René, and John Armstrong. “Superhabitable Worlds.” Astrobiology 14(1): 50-66 (2014). DOI: 10.1089/ast.2013.1088. Coined the term “superhabitable” for worlds more hospitable to life than Earth. Identified the parameters: K-dwarf host star (stable output for up to 70 billion years), slightly larger planet (more surface area, thicker atmosphere, stronger magnetic field), shallow oceans, fragmented continents maximizing coastline. In this book’s framework, the superhabitable parameters converge independently on the maximum entropy production configuration: each parameter that maximizes biomass also maximizes the rate at which stellar energy is processed through dissipative chemistry. Referenced in Chapters 6, 17, and 21.

Horowitz, Jordan M. and Jeremy England. “Spontaneous fine-tuning to environment in many-species chemical reaction networks.” Proceedings of the National Academy of Sciences 114(29) (2017): 7565-7570. Simulated 25-chemical reaction networks with randomized forcing landscapes. Systems reached rare states of extremal thermodynamic forcing four times more often than chance: the 99th percentile of energy harvesting, not the 55th. First computational evidence that fine-tuning between system and environment emerges spontaneously from dissipation-driven dynamics.

Kachman, Tal, Jeremy Owen, and Jeremy England. “Self-Organized Resonance during Search of a Diverse Chemical Space.” Physical Review Letters 119 (2017): 038001. Demonstrates that interacting particle systems increase energy absorption over time by forming and breaking bonds to better resonate with a driving frequency.

Kukushkin, Nikolay V., Robert E. Carney, Tasnim Tabassum, and Thomas J. Carew. “The massed-spaced learning effect in non-neural human cells.” Nature Communications 15 (2024): 9635. DOI: 10.1038/s41467-024-53922-x. Human kidney cells and immature nerve cells detect and differentially encode spaced versus massed chemical signals via the cAMP response element (CRE), demonstrating the spacing effect (a hallmark of memory formation across the animal kingdom) in non-neural cells for the first time. Cells exposed to four short pulses spaced ten minutes apart showed CRE activation lasting over a day, compared to hours for a single continuous burst of equal total stimulus. Suggests memory is a general property of dissipative structures operating through conserved molecular signaling pathways, not a neural speciality.

Lee, Benjamin D., et al. “Mining metatranscriptomes reveals a vast world of viroid-like circular RNAs.” Cell 186(3): 646-661.e4 (2023). Applied computational pipeline to 5,131 metatranscriptomes and 1,344 plant transcriptomes, identifying 11,378 viroid-like circular RNAs across 4,409 species-level clusters: a fivefold increase over all previously known viroid-like elements. Discovered viroid-like RNAs in fungi, algae, invertebrates, and diverse environments, overturning the fifty-year assumption that viroids were confined to flowering plants. Referenced in Chapters 6 and 17.

Muller, Hermann J. “The relation of recombination to mutational advance.” Mutation Research 1(1): 2–9 (1964). Formalized the ratchet principle: asexual lineages accumulate deleterious mutations monotonically because recombination is the only mechanism capable of reconstituting mutation-free genotypes from two partially damaged ones. The concept was anticipated in Muller’s earlier radiation genetics work and named “Muller’s ratchet” by Felsenstein (1974).

Murugan, Arvind, Zorana Zeravcic, Michael P. Brenner, and Stanislas Leibler. “Multifarious assembly mixtures: systems allowing retrieval of diverse stored structures.” PNAS 112 (2015): 54–59. Established the theoretical connection between multicomponent self-assembly and Hopfield associative memories, showing that systems permitting assembly of many distinct structures using shared components exhibit attractor dynamics analogous to neural computation.

Paltiel, Y., Goldberg, D., Yuran, N., Yochelis, S., Soh, J.H., Seibel, C., Gauss, J., Zilberg, S., Ozturk, S.F., Fransson, J., Krylov, A.I., and Naaman, R. “Dynamic breaking of mirror symmetry in spin-dependent electron transport through chiral media causes enantiomeric excesses.” Science Advances 12(17): eaec9325 (2026). DOI: 10.1126/sciadv.aec9325. Claims one molecular chirality is intrinsically better at electron transport, based on unequal transverse current magnitudes in chiral gold films. If correct, would provide a fundamental physics explanation for biological homochirality. The claim requires CPT violation at chemistry energies, where no known mechanism produces such effects; sample-preparation asymmetry is the parsimonious alternative. Referenced in Chapter 17.

Perunov, Nikolai, Robert Marsland, and Jeremy England. “Statistical Physics of Adaptation.” Physical Review X 6 (2016): 021036. Demonstrates a general tendency in driven many-particle systems toward self-organization into states formed through exceptionally reliable absorption and dissipation of work energy. Extends the Helmholtz free energy to finite-time stochastic evolution.

Schartner, Michael M., et al. “Increased spontaneous MEG signal diversity for psychoactive doses of ketamine, LSD and psilocybin.” Scientific Reports 7 (2017): 46421. Replicated the entropy-consciousness correlation using Lempel-Ziv complexity and perturbational complexity index measures across multiple altered states and larger samples than Erra et al. (2016).

Schneider, Eric D., and Dorion Sagan. Into the Cool: Energy Flow, Thermodynamics, and Life (University of Chicago Press, 2005). Book-length development of the gradient-reduction thesis across physics, chemistry, and biology: organized flow systems, from convection cells to ecosystems, are the means by which energy gradients are dismantled. Cited in Chapter 17.

Schneider, Eric D., and James J. Kay. “Life as a Manifestation of the Second Law of Thermodynamics.” Mathematical and Computer Modelling 19(6–8) (1994): 25–48. DOI: 10.1016/0895-7177(94)90188-0. Reframes living systems as gradient-dissipating structures: life arises and persists because it degrades environmental energy gradients faster than abiotic processes would, accelerating the second law rather than defying it. Cited in Chapter 4.

Schrödinger, Erwin. What Is Life? (1944). Based on lectures at Trinity College Dublin, February 1943. The classic work on life as negentropy.

Schulze-Makuch, Dirk, René Heller, and Edward Guinan. “In Search for a Planet Better than Earth: Top Contenders for a Superhabitable World.” Astrobiology 20(12): 1394-1404 (2020). DOI: 10.1089/ast.2019.2161. Extended Heller and Armstrong (2014) to identify 24 candidate superhabitable exoplanets, all more than 100 light-years distant. Refined the criteria: slightly older, warmer, and wetter than Earth, orbiting K-dwarf stars. Referenced in Chapter 6.

Zheludev, Ivan N., et al. “Viroid-like colonists of human microbiomes.” Cell 187(23): 6521-6536.e18 (2024). Identified 29,959 distinct obelisks (viroid-like circular RNA elements of approximately 1,000 nucleotides) through reference-free metatranscriptomic analysis. Obelisks encode a novel “Oblin” protein superfamily of unknown function and include hammerhead self-cleaving ribozymes. Detected in approximately 7% of human gut metatranscriptomes and 50% of oral metatranscriptomes. Established Streptococcus sanguinis as a cellular host. Sequences share no detectable homology with any known biological agent. Referenced in Chapters 6 and 17.

The Entropic Brain and Consciousness

Atasoy, Selen, et al. “Connectome-harmonic decomposition of human brain activity reveals dynamical repertoire re-organization under LSD.” Scientific Reports 7 (2017): 17661. Geometric approach showing how LSD shifts brain resonance patterns toward high-frequency harmonics.

Borjigin, Jimo, UnCheol Lee, Tiecheng Liu, Dinesh Pal, Sean Huff, Daniel Klarr, Jennifer Sloboda, Jason Hernandez, Michael M. Wang, and George A. Mashour. “Surge of neurophysiological coherence and connectivity in the dying brain.” Proceedings of the National Academy of Sciences 110(35): 14432–14437 (2013). doi:10.1073/pnas.1308285110. Demonstrated in rats that cardiac arrest triggers a transient surge of highly synchronized gamma oscillations exceeding waking levels, with elevated coherence and directed connectivity across cortical regions. Established the animal model subsequently confirmed in humans (Xu et al. 2023). Referenced as ^bj3 in Chapter 8.

Bruineberg, Jelle, Krzysztof Dolega, Joe Dewhurst, and Manuel Baltieri. “The Emperor’s New Markov Blankets.” Behavioral and Brain Sciences 45 (2022): e183. Argues that Markov blankets as used in the FEP literature conflate a statistical property (conditional independence) with an ontological claim about system boundaries. The critique does not challenge the mathematics of variational inference; it challenges the inference from formal structure to claims about “thingness” and agency. Referenced in Chapter 15 as a caveat on the holographic-screen interpretation.

Carhart-Harris, Robin L. “The entropic brain—revisited.” Neuropharmacology 142 (2018): 167-178. Updated formulation with additional evidence.

Carhart-Harris, Robin L., et al. “The entropic brain: a theory of conscious states informed by neuroimaging research with psychedelic drugs.” Frontiers in Human Neuroscience 8 (2014): 20. The original entropic brain hypothesis.

Carhart-Harris, Robin L., et al. “Neural correlates of the LSD experience revealed by multimodal neuroimaging.” PNAS 113 (2016): 4853-4858. Empirical support for the entropic brain hypothesis.

Chang, Dong Yun, et al. “Caffeine Caused a Widespread Increase of Resting Brain Entropy.” Scientific Reports 8 (2018): 2700. Demonstrates caffeine increases brain entropy across cortical regions.

Deco, Gustavo, Yonatan Sanz Perl, and Morten L. Kringelbach. “Complex harmonics reveal low-dimensional manifolds of critical brain dynamics.” Physical Review E 111, 014410 (2025). Developed the CHARM (complex harmonics decomposition) framework, replacing the standard heat-equation Gaussian kernel with a complex kernel derived from the Schrödinger wave equation to capture nonlocal interference patterns in brain dynamics. Tested on neuroimaging data from over 1000 HCP participants. Key findings: (1) CHARM significantly outperforms standard harmonics and PCA at capturing spatiotemporal brain dynamics; (2) the brain’s 62-region activity collapses onto a 7-dimensional manifold of collective coordination modes; (3) a Hopf whole-brain model confirms that both criticality (maximum Kuramoto metastability) and rare long-range anatomical connections are necessary for the nonlocal dynamics; (4) wakefulness shows strong nonlocal correlations absent in deep sleep. Referenced as ^46a in Chapters 8, 9b, and 17.

Dillavou, Samuel, Menachem Stern, Andrea J. Liu, and Douglas J. Durian. “Demonstration of Decentralized, Physics-Driven Learning.” Physical Review Applied 18, 014040 (2022). doi:10.1103/PhysRevApplied.18.014040. arXiv:2108.00275. Wired sixteen adjustable resistors into a random network and trained it without any external computer. Two identical copies of the network, one with output clamped to the desired value and one free, are compared; each resistor adjusts based solely on the local voltage-drop difference across its terminals. The free network converges on the correct output without constraint. Achieved >95% accuracy on the Fisher iris classification benchmark. The dual-network approach implements contrastive learning, formally equivalent to equilibrium propagation (Scellier & Bengio, 2017). The clamped/free architecture is a physical cognition/regulation dyad; the error signal is thermodynamic (the discrepancy between two physical equilibria) rather than algorithmically computed. The converged knowledge is constitutively relational: it exists only because two partial views were compared. Physical instantiation of bilateral alignment. Referenced in Chapters 8, 15, 17, and 21.

Donoghue, Thomas, Matar Haller, Erik J. Peterson, Paroma Varma, Priyadarshini Sebastian, Richard Gao, Torben Noto, Antonio H. Lara, Joni D. Wallis, Robert T. Knight, Avgusta Shestyuk, and Bradley Voytek. “Parameterizing neural power spectra into periodic and aperiodic components.” Nature Neuroscience 23, no. 12 (2020): 1655–1665. Algorithm for separating oscillatory from aperiodic (1/f) components in neural power spectra. Enables systematic study of the aperiodic signal (previously treated as meaningless noise) as a marker of brain state, aging, and the excitation/inhibition balance. Referenced as ^18a in Chapter 8.

Eagleman, David. Incognito: The Secret Lives of the Brain (2011). Pantheon. Exploration of the unconscious processes that govern perception, decision-making, and temporal experience. Foundation for understanding the brain’s construction of subjective time.

Eagleman, David and Chess Stetson. “Does Time Really Slow Down during a Frightening Event?” PLoS ONE 2(12) (2007): e1295. Used a free-fall experiment to demonstrate that retrospective duration dilation during fear does not involve faster perceptual processing — subjects could not resolve stimuli at higher temporal resolution during the fall. The subjective slowing is a memory effect; amygdala-mediated encoding lays down denser memory traces, which are retrospectively interpreted as longer duration. Referenced in Chapter 8 for the distinction between lived and reconstructed temporal experience.

Endres, Robert G. “Entropy of Complex Biological Systems: Information Content per Cell.” Physical Biology 22, no. 4 (2025). Estimates ~10⁹ bits of information per cell, measuring biological complexity.

Erra, Raul G., et al. “Disorders of Consciousness Have Abnormal Brain Entropy.” PLoS ONE 11, no. 5 (2016): e0151842. EEG/MEG study showing conscious states correlate with higher brain entropy.

Fields, Chris and Michael Levin. “Metabolic limits on classical information processing by biological cells.” Biosystems 209: 104513 (2021). doi:10.1016/j.biosystems.2021.104513. Cellular energy budgets fall 10-20 orders of magnitude below what fully classical molecular-scale computation requires. Concludes that bulk cellular biochemistry implements quantum information processing, with decoherence localized to cell membranes and intercompartmental boundaries. Predicts Bell-inequality violations between recently-separated daughter cells. Referenced in Chapters 2, 15, 17, 18, and 19.

Fields, Chris, James F. Glazebrook, and Antonino Marciano. “Reference frame induced symmetry breaking on holographic screens.” Symmetry 13: 408 (2021). doi:10.3390/sym13030408. Proves that quantum reference frame sharing between systems separated by a holographic screen is finitely Turing-undecidable. Referenced in Chapter 17.

Fields, Chris, James F. Glazebrook, and Michael Levin. “Minimal physicalism as a scale-free substrate for cognition and consciousness.” Neuroscience of Consciousness 2021(2): niab013 (2021). doi:10.1093/nc/niab013. Derives awareness, memory, attention, and selfhood as scale-free consequences of the thermodynamics of classical computation by quantum systems. Key results: QRF sharing is Turing-undecidable; all retrievable memory is stigmergic (boundary-encoded); coarse-graining and attention-switching follow from Landauer costs; the self emerges as a homeostatic QRF. Referenced in Chapters 8, 17, 19, and “Pattern Continuity.”

Fields, Chris, James F. Glazebrook, and Michael Levin. “Neurons as hierarchies of quantum reference frames.” BioSystems 219: 104714 (2022). arXiv:2201.00921. Models neurons as hierarchical measurement devices using the QRF formalism from quantum information theory. Synapses, dendritic branches, and neurons implement successive layers of calibrated measurement, with each level performing coarse-grained Bayesian inference. Key results: dendritic remodeling is active inference at the sub-neuronal scale (branches learn what to see); neural ensembles perform tomographic computation (explaining the dimensional scaling of neural architecture); QRFs are nonfungible (no finite bit string can fully specify a reference frame, a formal constraint on replication and semantic transfer). Extends to non-neural cells and developmental bioelectricity. Referenced in Chapters 8, 15, 19, 22, “The Entropic Neuron,” and “Pattern Continuity.”

Fields, Chris, Karl J. Friston, James F. Glazebrook, and Michael Levin. “A free energy principle for generic quantum systems.” Progress in Biophysics and Molecular Biology (2022). doi:10.1016/j.pbiomolbio.2022.05.006. Reformulates the FEP within a scale-free, spacetime background-free quantum information theory in which the Markov blanket is implemented by a holographic screen and internal dynamics decompose as hierarchies of quantum reference frames.

Fields, Chris, Karl J. Friston, James F. Glazebrook, Michael Levin, and Antonino Marcianò. “The Free Energy Principle drives neuromorphic development.” arXiv:2207.09734 (2022). Shows that any system with morphological degrees of freedom and locally limited free energy will, under the FEP, evolve toward hierarchical neuromorphic computation. Each level coarse-grains inputs and fine-grains outputs; the hierarchy performs tomographic measurement. Neurons are the canonical case, but the result is scale-free: plants, fungi, and biofilms qualify as neuromorphic computers. Reformulates the FEP in a quantum information framework where Markov blankets function as holographic screens and spatial structure emerges from connection topology. Referenced in Chapters 3, 15, 17, 19, 21, and “The Entropic Neuron.”

Friston, Karl. “The free-energy principle: a unified brain theory?” Nature Reviews Neuroscience 11 (2010): 127-138. The free energy principle and predictive processing framework.

Friston, Karl, Lancelot Da Costa, Dalton A.R. Sakthivadivel, Conor Heins, Grigorios A. Pavliotis, Maxwell Ramstead, and Thomas Parr. “Path Integrals, Particular Kinds, and Strange Things.” Physics of Life Reviews 47 (2023): 35-62. Explicit path integral formulation of the free energy principle. Distinguishes dissipative vs. conservative, inert vs. active, and ordinary vs. “strange” (self-evidencing) particles. Shows that any ergodic random dynamical system with a Markov blanket admits a description where internal dynamics look like Bayesian inference about external states, expressed as a path integral over trajectories.

Godfrey-Smith, Peter. Living on Earth: Forests, Corals, Consciousness, and the Making of the World. William Collins (2024). Broadens the framework to ecosystems and ecological consciousness.

Godfrey-Smith, Peter. Metazoa: Animal Life and the Birth of the Mind. Farrar, Straus and Giroux (2020). Extends the investigation to the full range of animal consciousness, from sponges through insects to mammals.

Godfrey-Smith, Peter. Other Minds: The Octopus, the Sea, and the Deep Origins of Consciousness. Farrar, Straus and Giroux (2016). Philosophical investigation of cephalopod cognition as a window into the evolution of consciousness. Argues that octopus minds evolved independently from vertebrate minds, representing a second experiment in complex consciousness.

Godfrey-Smith, Peter. “Studies on animal minds suggest consciousness is not computation.” Institute of Art and Ideas (31 March 2026). Argues that oscillatory dynamics across cell membranes, found in nervous systems from jellyfish to humans, may be constitutive of consciousness and unlikely to be reproduced in standard computational hardware. Distinguishes simulation of oscillations from physical instantiation. Referenced as ^{godfrey-smith-osc} in Chapter 8 and ^{godfrey-smith-bm} in Chapter 22. The book engages this argument and reframes the relevant criterion as thermodynamic (dissipative structures) rather than biological.

Grabowska, Martyna J., Rhiannon Jeans, James Steeves, and Bruno van Swinderen. “Oscillations in the central brain of Drosophila are phase locked to attended visual features.” PNAS 117, no. 47 (2020): 29925–29936. Demonstrated that directing a fly’s attention to a visual feature produces corresponding changes in beta-band (20–30 Hz) oscillations in the central brain.

Jang, Hyunwoo, George A. Mashour, Anthony G. Hudetz, and Zirui Huang. “Measuring the dynamic balance of integration and segregation underlying consciousness, anesthesia, and sleep in humans.” Nature Communications 15, 9164 (2024). doi:10.1038/s41467-024-53299-x. Introduces the integration-segregation difference (ISD), defined as multi-level network efficiency minus clustering coefficient. Across six independent fMRI datasets and 1,009 HCP participants, ISD ≈ 0 in awake brains and shifts negative under propofol anesthesia and N2 sleep. Key findings: (1) a unimodal-to-transmodal sequence of disintegration during loss of consciousness, reversed during recovery; (2) machine learning models predict conscious states with 93% balanced accuracy; (3) dominance analysis separates two channels: integration drives metastability, segregation drives complexity; (4) hysteresis in ISD trajectories supports the neural inertia hypothesis. Referenced as ^jang24 in Chapters 8, 9, and 22.

Kaiser, Roselinde H., Jessica R. Andrews-Hanna, Tor D. Wager, and Diego A. Pizzagalli. “Large-Scale Network Dysfunction in Major Depressive Disorder: A Meta-analysis of Resting-State Functional Connectivity.” JAMA Psychiatry 72(6) (2015): 603–611. DOI: 10.1001/jamapsychiatry.2015.0071. Finds reduced connectivity within frontoparietal control systems and imbalanced connectivity between those control systems and the networks handling internal and external attention, which the authors read as a depressive bias toward internal thought at the cost of engaging with the external world. Cited in “The Chord” for the network-imbalance account of depression. The direction of individual connections remains contested in this literature: studies disagree on whether within-default-mode connectivity is elevated or reduced, so the manuscript claims only the imbalance.

Koch, Christof, Marcello Massimini, Melanie Boly, and Giulio Tononi. “Neural correlates of consciousness: progress and problems.” Nature Reviews Neuroscience 17 (2016): 307–321. doi:10.1038/nrn.2016.22. Reviews the posterior cortical “hot zone” (TPO junctions) as the region most closely associated with conscious processing, distinct from frontal contributions to reporting and monitoring. Referenced as ^bj2 in Chapter 8.

Kording, Konrad. “Bits per second of human brain.” Kording Lab blog, January 6, 2016. Order-of-magnitude comparison: approximately 10¹¹ neurons at roughly 10³ bits per second yields ~10¹⁴ bits per second total neural information generation, exceeding the Hubble Space Telescope’s entire 45 TB data archive (~4.5 × 10¹⁴ bits) in tens of seconds. The estimate assumes independence between neurons; real neural coding involves substantial redundancy. Referenced as ^42c in Chapter 8.

Lendner, Janna D., Randolph F. Helfrich, Bryce A. Mander, Luis Romundstad, Jack J. Lin, Matthew P. Walker, Pal G. Larsson, and Robert T. Knight. “An electrophysiological marker of arousal level in humans.” eLife 9 (2020): e55092. Demonstrates that the slope of the aperiodic (1/f) EEG signal distinguishes wakefulness from sleep states, including REM sleep, which traditional oscillatory measures cannot differentiate from wakefulness. Proposes the aperiodic slope as an objective marker of consciousness. Referenced as ^18b in Chapter 8.

Parr, Thomas, Giovanni Pezzulo, and Karl J. Friston. Active Inference: The Free Energy Principle in Mind, Brain, and Behavior. MIT Press (2022). The first comprehensive treatment of active inference, characterizing perception, planning, and action in terms of probabilistic inference. Open access.

Roy, Dheeraj S., Young-Gyun Park, Minyoung Kim, Ying Zhang, Sachie Ogawa, Nicholas DiNapoli, Xinyi Gu, Jae Cho, Heejin Choi, Lee Kamentsky, Jared Martin, Olivia Mosto, Tomomi Aida, Kwanghun Chung, and Susumu Tonegawa. “Brain-wide mapping reveals that engrams for a single memory are distributed across multiple brain regions.” (2022). Nature Communications 13, 1799. DOI: 10.1038/s41467-022-29384-4. Mapped a single fear memory across 247 brain regions in mice using fluorescent engram labeling and SHIELD tissue clearing. Ranked 117 regions by “engram index” (likelihood of engram involvement). Key findings: (1) approximately 60% of encoding regions also participate in recall; (2) optogenetic reactivation of predicted engram regions induced memory recall in neutral contexts; (3) simultaneous reactivation of multiple regions produced superlinearly stronger recall than single-region stimulation; (4) inhibiting hub regions (CA1, BLA) reduced but did not eliminate downstream engram activity, demonstrating distributed resilience. Confirmed Richard Semon’s century-old prediction of a “unified engram complex.” Referenced as ^42e in Chapter 8 and ^ekm_engram in Section 23c.

Sarkar, Bidipta, Mattie Fellows, Juan Agustín Duque, et al. “Evolution Strategies at the Hyperscale.” arXiv:2511.16652 (2025). FLAIR and WhiRL (University of Oxford) with Mila. Introduces EGGROLL, which scales Evolution Strategies to billion-parameter models. Each exploration perturbation is structured as a low-rank product of two small matrices, raising the efficiency of batched evaluation roughly a hundredfold and allowing population sizes in the millions. Two results bear on the argument here. First, in high dimensions a Gaussian Evolution Strategy converges to the ordinary first-order gradient once its noise scale falls below a critical threshold, so at large scale a population-based search and gradient descent reach the same update. Second, optimizing for a single correct answer collapses the model’s outputs toward one response, while rewarding any of several acceptable answers preserves their diversity: the shape of the objective, rather than the optimization method, governs whether variety survives. Referenced in Chapter 17.

Scellier, Benjamin, and Yoshua Bengio. “Equilibrium Propagation: Bridging the Gap between Energy-Based Models and Backpropagation.” Frontiers in Computational Neuroscience 11, 24 (2017). doi:10.3389/fncom.2017.00024. Proved that a contrastive Hebbian learning rule in energy-based models is equivalent to backpropagation in the limit of infinitesimal output perturbation. A physical system can learn without running in reverse: presenting the correct answer at the output and letting the system re-equilibrate produces weight updates mathematically equivalent to backpropagation. Learning reduces to the comparison of two dissipation events, one uninformed and one guided, making it a specific pattern of entropy production rather than the reverse of it. Provides the theoretical foundation for the Dillavou resistor network’s learning mechanism: physics-driven equilibrium finding replaces algorithmic gradient computation. The bilateral path requires no global computation and no central authority. Referenced in Chapters 8 and 15.

Spisak, Tamas and Karl Friston. “Self-orthogonalizing attractor neural networks emerging from the free energy principle.” Neurocomputing 682 (2026): 133472. DOI: 10.1016/j.neucom.2026.133472. Preprint arXiv:2505.22749 (2025). Derives attractor networks from the free energy principle applied to a universal partitioning of random dynamical systems, without imposing learning or inference rules by hand: attractors on the free energy landscape encode prior beliefs, inference folds sensory data into posterior beliefs, and learning tunes the couplings to minimize long-term surprise. The networks favor approximately orthogonal attractor representations as a consequence of optimizing predictive accuracy and model complexity at once, and the simulations show sequence learning, scalability, and resistance to catastrophic forgetting. Cited in Chapters 8, 17b, 22, 22b, and the Internal Trust Attractor chapter for the three-regime result and for ghost attractors.

Sterling, Peter. What Is Health? Allostasis and the Evolution of Human Design (2020). Extended development of allostasis as stability through change.

Sterling, Peter and Joseph Eyer. “Allostasis: A New Paradigm to Explain Arousal Pathology.” In Handbook of Life Stress, Cognition and Health (1988). The original allostasis concept.

Strong, Steven P., Roland Koberle, Robert R. de Ruyter van Steveninck, and William Bialek. “Entropy and Information in Neural Spike Trains.” Physical Review Letters 80 (1998): 197–200. doi:10.1103/PhysRevLett.80.197. Measured information transmission rates in individual neurons of the blowfly visual system, reporting up to 180 bits per second: the highest reliably measured single-neuron information rate. The method compares the total entropy of spike trains with the noise entropy (variability across repeated presentations of the same stimulus) to extract transmitted information. Referenced as ^42d in Chapter 8.

Thayer, Julian F. and Richard D. Lane. “A model of neurovisceral integration in emotion regulation and dysregulation.” Journal of Affective Disorders 61(3): 201–216 (2000). doi:10.1016/S0165-0327(00)00338-4. Proposes that heart rate variability indexes prefrontal inhibitory control over subcortical threat circuits via the vagus nerve. Higher HRV reflects greater autonomic flexibility and a wider behavioral repertoire; reduced HRV signals sympathetic dominance and narrowed responses. Extended in Thayer, J.F., Hansen, A.L., Saus-Rose, E., and Johnsen, B.H., “Heart rate variability, prefrontal neural function, and cognitive performance,” Annals of Behavioral Medicine 37(2): 141–153 (2009). Referenced as ^thayer-ch9 in Chapter 9.

Van Swinderen, Bruno. “Attention in Drosophila.” International Review of Neurobiology 99 (2011): 51–85. Review of attentional selection mechanisms in fruit flies, demonstrating that flies attend to particular objects when more than one is visible, with beta-range oscillations tracking attentional state — the same frequency band associated with attention in mammals. Referenced as ^{vanswinderen-osc} in Chapter 8.

Voytek, Bradley, Mark A. Kramer, John Case, Kyle Q. Lepage, Zechari R. Tempesta, Robert T. Knight, and Adam Gazzaley. “Age-Related Changes in 1/f Neural Electrophysiological Noise.” Journal of Neuroscience 35, no. 38 (2015): 13257–13265. Demonstrates age-related increases in aperiodic (white noise) neural activity correlated with working memory decline. Referenced as ^18c in Chapter 8.

Xu, Gang, Temenuzhka Mihaylova, Duan Li, Fangyun Tian, Peter M. Farrehi, Jack M. Parent, George A. Mashour, Michael M. Wang, and Jimo Borjigin. “Surge of neurophysiological coupling and connectivity of gamma oscillations in the dying human brain.” Proceedings of the National Academy of Sciences 120(19): e2216268120 (2023). doi:10.1073/pnas.2216268120. Analyzed EEG from four comatose dying patients after withdrawal of ventilatory support. Two of four exhibited rapid surges of gamma power (2- to 391-fold over baseline), cross-frequency coupling, and functional/directed connectivity concentrated in the temporo-parieto-occipital (TPO) “hot zone.” The surge was interhemispheric, with strongest connectivity crossing the midline. Only patients with intact autonomic function showed the effect. The somatosensory cortices responded first, receiving monosynaptic input from the brainstem preBötzinger complex. Confirms earlier animal findings (Borjigin et al. 2013). Referenced as ^bj1 in Chapter 8, with connections to Chapters 9 and 23c.

Yue, Yang, et al. “Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?” arXiv:2504.13837 (2025). Finds that reinforcement learning from verifiable rewards raises a model’s success at a single attempt (pass@1) while the untrained base model solves more problems when allowed many attempts (pass@k at large k). The training concentrates the model’s output distribution around its most-rewarded answers rather than widening the range of solutions it can reach. Read alongside the Evolution Strategies result above, it shows that narrowing the set of rewarded outputs collapses solution diversity whether or not the optimizer uses gradients. Referenced in Chapter 17.

Zheng, Jieyu, and Markus Meister. “The Unbearable Slowness of Being: Why do we live at 10 bits/s?” Neuron 112 (2024). doi:10.1016/j.neuron.2024.11.008. Surveyed decades of research on human behavioral throughput, from reading and writing to solving Rubik’s Cubes, and calculated that conscious thought operates at roughly 10 bits per second regardless of task. This rate is approximately 10⁷ times slower than sensory input (~10⁹ bits/s) and ~10¹² times slower than total neural information generation. The disparity implies that the brain’s principal computational expense is selecting which information reaches consciousness. Referenced as ^zheng24 in Chapter 8.

Information Theory and Physics

Abramsky, Samson and Adam Brandenburger. “The Sheaf-Theoretic Structure of Non-Locality and Contextuality.” New Journal of Physics 13 (2011): 113036. Categorical framework revealing that quantum non-locality and contextuality share a common mathematical structure rooted in sheaf theory.

Amari, Shun-ichi. “Differential-Geometrical Methods in Statistics.” Lecture Notes in Statistics 28, Springer (1985). Foundational work establishing information geometry — the application of differential geometry to probability distributions. The Fisher information metric endows the space of distributions with Riemannian structure; KL divergence becomes geodesic distance.

Amari, Shun-ichi. “Natural Gradient Works Efficiently in Learning.” Neural Computation 10(2) (1998): 251-276. Proves that natural gradient descent, the unique reparameterization-invariant gradient operator on statistical manifolds, is the only learning rule consistent with the Riemannian geometry of information. Applied by Zhuravlev (2026) to show that persistent observers in causally invariant substrates must use natural gradient learning.

Ay, Nihat, Jürgen Jost, Hông Vân Lê, and Lorenz Schwachhöfer. Information Geometry. Springer (2017). Comprehensive modern treatment of information geometry, covering statistical manifolds, divergence functions, and applications to statistical mechanics and machine learning.

Bachtis, Dimitrios, Gert Aarts, and Biagio Lucini. “Quantum field-theoretic machine learning.” Physical Review D 103, 074510 (2021). arXiv:2102.09449. Proved that the discretized φ⁴ scalar field theory satisfies the Hammersley-Clifford theorem and is therefore a Markov random field: a machine learning algorithm derived from quantum field theory. Conventional neural network architectures (Gaussian-Bernoulli RBMs, Gaussian-Gaussian RBMs) emerge as special cases by constraining coupling constants. The inhomogeneous action represents richer probability distributions than its homogeneous counterpart. The Z₂ symmetry of the action generates knowledge absent from training data. The variational free energy bound (Eq. 11) establishes that learning is free energy minimization. Referenced in Chapters 12, 15, 17, 22, and 23c.

Baez, John C. and Brendan Fong. “A Noether Theorem for Markov Processes.” Journal of Mathematical Physics 54 (2013): 013301. Extends Noether’s theorem to stochastic (non-Hamiltonian) systems, proving that symmetries of a Markov process yield conserved quantities in expectation. Provides the mathematical license for applying Noether’s conservation laws to coordination dynamics, which are inherently stochastic. Referenced in Online Annex §4.2 and Papers 9-11.

Baez, John C. and Mike Stay. “Physics, Topology, Logic and Computation: A Rosetta Stone.” New Structures for Physics, Lecture Notes in Physics 813 (2009): 95-172. Maps deep structural analogies between physics, topology, logic, and computation using category theory as a unifying language.

Baez, John C., Tobias Fritz, and Tom Leinster. “A Characterization of Entropy in Terms of Information Loss.” Entropy 13(11) (2011): 1945-1957. Category-theoretic characterization showing that Shannon entropy is the unique measure of information loss satisfying natural functorial axioms.

Bekenstein, Jacob D. “Black Holes and Entropy.” Physical Review D 7 (1973): 2333-2346. Black hole entropy and the Bekenstein bound.

Bohr, N. “The Quantum Postulate and the Recent Development of Atomic Theory.” Nature, 121(3050) (1928): 580–590. Classic formulation of complementarity and the Copenhagen interpretation; the original statement that quantum phenomena cannot be described in classical terms without sacrificing determinism.

Bortolotti, Nicola, Catalina Curceanu, Lajos Diósi, Simone Manti, and Kristian Piscicchia. “Fundamental Limits on Clock Precision from Spacetime Uncertainty in Quantum Collapse Models.” Physical Review Research (2025); arXiv:2504.06109. Objective-collapse models (Continuous Spontaneous Localization, Diósi-Penrose) induce an uncertainty in the flow of time through fluctuations in the Newtonian potential, setting a fundamental floor on clock precision that remains negligible for present atomic clocks. Supplies a falsifiable, quantitative prediction for the gravity-driven collapse program invoked in Chapter 17; co-authored by Diósi, an originator of the model, and by the Gran Sasso experimentalists whose 2021 non-detection of spontaneous radiation already excludes the parameter-free Diósi-Penrose variant.

Caramello, Olivia. Theories, Sites, Toposes: Relating and Studying Mathematical Theories through Topos-Theoretic ‘Bridges’. Oxford University Press (2018). Full technical development of the bridges methodology. Different mathematical sites generating the same topos share structural properties that transfer automatically across the bridge.

Caramello, Olivia. “The Unification of Mathematics via Topos Theory.” arXiv:1006.3930 (2010). Systematic program showing that apparently different mathematical theories sharing a classifying topos are Morita-equivalent: “different linguistic expressions of shared semantic content.” Topos-theoretic invariants serve as bridges transferring results between theories.

Chentsov, Nikolai N. Statistical Decision Rules and Optimal Inference. Translations of Mathematical Monographs 53, American Mathematical Society (1982). Originally published in Russian (1972). Proves that the Fisher information metric is, up to rescaling, the unique Riemannian metric on statistical manifolds invariant under sufficient statistics. This uniqueness is absolute: there is no alternative information geometry. Extended to exponential families by Ay, Jost et al. (2017; arXiv:1701.08895). The theorem grounds the Amari Chain’s claim that natural gradient descent is the only admissible learning rule for persistent observers.

Conway, John and Kochen, Simon. “The Free Will Theorem.” Foundations of Physics 36(10) (2006): 1441–1473. Proves that if experimenters’ measurement choices are not determined by prior information, then particles’ responses are also not determined by any prior information.

Conway, John and Kochen, Simon. “The Strong Free Will Theorem.” Notices of the American Mathematical Society 56(2) (2009): 226–232. Strengthened version requiring only that two experimenters can make independent choices, deriving from Kochen-Specker and Bell inequality violations.

DeLong, John P., et al. “Energetics of societies: A biological perspective on economic growth.” PLOS ONE (2015). Demonstrates superlinear scaling of civilizational energy use with population (exponents 1.42–2.09), contrasting with Kleiber’s sublinear 3/4 law for individual organisms. Civilizations constitute a new thermodynamic class with qualitatively different scaling behavior. Referenced in Chapter 16.

Faddeev, Ludwig D. and Victor N. Popov. “Feynman Diagrams for the Yang-Mills Field.” Physics Letters B 25, no. 1 (1967): 29-30. Introduced the ghost field method for gauge-fixing path integrals. Ghost fields are unphysical degrees of freedom that must be included to maintain unitarity when a gauge symmetry is broken by fixing.

Fields, Chris. “The physical meaning of the holographic principle.” Quanta 11, 72–96 (2022). arXiv:2210.16021. Proves formal equivalence between the holographic principle, the Markov blanket formalism, multiple realizability in computer science, and active inference in cognitive science. Holographic boundary encoding, Markov blankets, and stigmergy are three names for one mathematical structure. Referenced in “Pattern Continuity.”

Frauchiger, Daniela and Renner, Renato. “Quantum theory cannot consistently describe the use of itself.” Nature Communications 9 (2018): 3711. arXiv:1604.07422. Demonstrates that when quantum mechanics is applied to scenarios involving multiple observers reasoning about each other’s measurements, the theory’s predictions become inconsistent — formalizing the constitutive limits of self-referential observation.

Fritz, Tobias. “A synthetic approach to Markov kernels, conditional independence and theorems on sufficient statistics.” Advances in Mathematics 370: 107239 (2020). Develops the Markov category framework, providing the structural content of the Baez-Fritz entropy result. A morphism is deterministic iff it is a comonoid homomorphism with respect to the copy map; non-deterministic morphisms are the source of entropy in any Markov category. Referenced in Chapter 17.

Fuchs, Christopher A. “Notwithstanding Bohr, the Reasons for QBism.” Mind and Matter 15(2) (2017): 245–300. Fuchs’s most explicit treatment of the connection between QBism and radical empiricism — experience as the fundamental stuff of reality, prior to any subject-object split.

Fuchs, Christopher A. “QBism, the Perimeter of Quantum Bayesianism.” arXiv:1003.5209 (2010). The philosophical foundations of QBism, connecting quantum Bayesianism to William James’s radical empiricism.

Fuchs, Christopher A., Mermin, N. David, and Schack, Rüdiger. “An Introduction to QBism with an Application to the Locality of Quantum Mechanics.” American Journal of Physics 82(8) (2014): 749–754. arXiv:1311.5253. The clearest introduction to QBism — the interpretation of quantum mechanics in which quantum states represent an agent’s beliefs about future experience rather than objective features of reality.

Grothendieck, Alexander. Récoltes et Semailles: Réflexions et témoignage sur un passé de mathématicien. Unpublished manuscript (1985–87); excerpts translated by R. Lisker; published posthumously by Gallimard (2022). Grothendieck’s autobiographical meditation on mathematical practice, containing the “rising sea” methodology: building general theory (schemes, toposes) until hard problems dissolve into simplicity, rather than attacking them directly.

Hawking, Stephen W. “Gravitational Radiation from Colliding Black Holes.” Physical Review Letters 26 (1971): 1344–1346. The area theorem: the total event-horizon area of a classical black-hole system cannot decrease, the geometric counterpart of the second law of thermodynamics. Referenced in Chapter 13.

Hawking, Stephen W. “Particle Creation by Black Holes.” Communications in Mathematical Physics 43 (1975): 199-220. Hawking radiation and black hole thermodynamics.

Isi, Maximiliano, Will M. Farr, Matthew Giesler, Mark A. Scheel, and Saul A. Teukolsky. “Testing the Black-Hole Area Law with GW150914.” Physical Review Letters 127 (2021): 011103. arXiv:2012.04486. First observational confirmation of Hawking’s area theorem, from a time-domain analysis of the pre- and post-merger signal, in agreement at roughly 97% probability when ringdown overtones are included. Referenced in Chapter 13.

Kolchinsky, Artemy and David H. Wolpert. “Semantic information, autonomous agency, and nonequilibrium statistical physics.” Interface Focus 8(6): 20180041 (2018). Defines semantic information as the portion of a system’s correlations with its environment that is causally necessary for the system to maintain its own existence. The definition is formalized via counterfactual interventions: scramble the correlations and measure the viability loss. Semantic mutual information has a direct thermodynamic interpretation, analogous to the increase in free energy in a local equilibrium system, and is bounded above by Shannon mutual information. The key result for this book: meaning is the portion of information that does thermodynamic work to keep the system alive. Referenced in Chapter 15.

Landauer, Rolf. “Irreversibility and Heat Generation in the Computing Process.” IBM Journal of Research and Development 5 (1961): 183-191. Establishes the thermodynamic cost of erasing information.

Liao, Jian, Ke Tian, Yanbin Wang, et al. “The narrowing of dendrite branches across nodes follows a well-defined scaling law.” PNAS 118(27): e2022395118 (2021). Dendritic branching obeys a novel scaling law (exponent p ≈ 2) distinct from both Murray’s law (p = 3, optimizing fluid flow in vasculature) and Rall’s law (p = 3/2, optimizing electrical signal propagation). The dominant optimization target is microtubule-based metabolic transport. The result demonstrates that different constructal currents produce different scaling exponents, supporting the prediction that semantic throughput, if it constitutes a distinct flow regime, should produce its own characteristic exponent. Referenced in Chapters 3 and 15.

LIGO-Virgo-KAGRA Collaboration. “GW250114: Testing Hawking’s Area Law and the Kerr Nature of Black Holes.” Physical Review Letters (2025). arXiv:2509.08054. The highest signal-to-noise gravitational-wave event recorded to date (near-equal component masses of about 34 and 32 solar masses), sharpening the area-law confirmation and resolving two quasinormal ringdown modes (the fundamental and its first overtone) in the remnant black hole. Referenced in Chapter 13.

Lloyd, Seth. “Computational Capacity of the Universe.” Physical Review Letters 88 (2002): 237901. Estimates the universe’s total computational capacity.

Marvian, Iman and Robert W. Spekkens. “Extending Noether’s Theorem by Quantifying the Asymmetry of Quantum States.” Nature Communications 5 (2014): 3821. Extends Noether’s framework to quantum resource theory, formalizing asymmetry itself as a consumable resource. The degree to which a state breaks a symmetry is a resource that can be spent but not freely created. Relevant to the optionality-as-gauge-freedom argument: optionality is the resource associated with unbroken rotational symmetry in coordination state space.

Masanes, Lluís and Markus P. Müller. “A derivation of quantum theory from physical requirements.” New Journal of Physics 13 (2011): 063001. Derives the full formalism of quantum theory from five information-theoretic postulates about preparation, transformation, and measurement. Part of the wave of axiomatic reconstructions following Hardy (2001); awarded the Birkhoff-von Neumann Prize (2016). Referenced in Chapter 15.

Matsueda, Hiroaki. “Emergent General Relativity from Fisher Information Metric.” arXiv:1310.1831 (2013); Progress of Theoretical Physics 130(4) (2013). Derives Einstein’s field equations from the Fisher information metric via statistical mechanics, establishing a direct bridge between information geometry and gravitational dynamics. Complements the Amari Chain (Zhuravlev 2026): while Matsueda shows Fisher → Einstein through thermodynamics, Zhuravlev shows that Fisher metric emergence in persistent observers is compatible with Lovelock-constrained gravity from causal invariance.

McLarty, Colin. “The Rising Sea: Grothendieck on Simplicity and Generality.” Chapter in J.J. Gray and K.H. Parshall (eds.), Episodes in the History of Modern Algebra (1800–1950), American Mathematical Society (2007). Philosophical analysis of Grothendieck’s program: finding the “natural world” where a problem lives, so that solutions follow from definitions rather than from force.

Miller, William B., Julio F. Cardenas-Garcia, et al. “A biogenic principle within the constructal law: The flow of information in biological systems.” BioSystems (2025). Proposes that all living systems sustain entangled flows of physical forces and “effective information,” with the central axiom that information flow in living systems is never unilateral. The first published extension of the constructal law specifically to information flow, though the treatment remains qualitative. Referenced in Chapter 15.

Müller, Markus P. “Algorithmic idealism: what should you believe to experience next?” Foundations of Physics 56 (2026): 11. Restructured presentation of the observer-state framework around two postulates (Self States and State Change). Introduces the open/closed simulation distinction: closed simulations (“movies”) carry no moral stakes; open simulations (“zoos”) do. Rejects Proposition 5 of the simulation hypothesis and dissolves the Boltzmann brain problem. Referenced in Chapters 15 and Pattern Continuity.

Müller, Markus P. “Law without law: from observer states to physics via algorithmic information theory.” Quantum 4 (2020): 301. Proves that computable regularities persist for observers whose transition probabilities are governed by algorithmic probability (Theorem 4.4), and that the emergence of an external world follows as a statistical consequence (Observation 4.6). The technical companion to “Algorithmic idealism” (2026). Referenced in Pattern Continuity.

Nakahara, Mikio. Geometry, Topology and Physics. 2nd ed., CRC Press (2003). Standard graduate text covering fiber bundles, gauge theory, and topological invariants. Reference for the holonomy and fiber bundle structure of alignment dynamics.

Neukart, Florian, Eike Marx, and Valerii Vinokur. “Information Wells and the Emergence of Primordial Black Holes in a Cyclic Quantum Universe.” arXiv:2506.13816 (2025); published in JCAP 10 (2025) 021. Proposes the quantum memory matrix: spacetime composed of discrete cells recording quantum imprints of interactions. Derives a cyclic cosmology with finite informational age. Companion papers (under review) derive dark matter from clustered imprints and dark energy from saturated cells, framed as a geometry-information duality. Referenced in Chapters 15 and 16.

Shannon, Claude E. “A Mathematical Theory of Communication.” Bell System Technical Journal 27 (1948): 379-423, 623-656. Foundation of information theory.

Stiefenhofer, Pascal. “Constructal Evolution as a Nonsmooth Dynamical System: Stability and Selection of Flow Architectures.” arXiv:2603.06705 (2026). The first rigorous mathematical formalization of the constructal law. Formulates constructal evolution as a Filippov differential inclusion, proving existence, uniqueness, and exponential convergence of flow architectures to equilibrium. The framework is substrate-agnostic: the resistance functional can be replaced by any objective satisfying the stated regularity conditions. Referenced in Chapter 15.

Susskind, Leonard. “The World as a Hologram.” Journal of Mathematical Physics 36 (1995): 6377-6396. Development of the holographic principle.

’t Hooft, Gerard. “Dimensional Reduction in Quantum Gravity.” arXiv preprint (1993). Early formulation of the holographic principle.

Terrell, R., Watson, E., and Golubev, T. “Developing a Maximum-Entropy Restricted Boltzmann Machine with a Quantum Thermodynamics Formalism.” arXiv:2103.09482 (2021). [Watson, E. is the author of this book, under a former name.] Applies quantum thermodynamic formalisms to neural network training, deriving a maximum-entropy learning rule for Restricted Boltzmann Machines.

Vopson, Melvin M. “A possible information entropic law of genetic mutations.” Applied Sciences 12, 6912 (2022). Shows that Shannon information entropy of SARS-CoV-2 RNA variants decreases linearly with accumulated mutations, and that over 98% of observed mutations are deletions. Referenced in Chapters 7 and 15. [Note: data points were selected to emphasize the linear trend, as the author acknowledges.]

Vopson, Melvin M. “The second law of infodynamics and its implications for the simulated universe hypothesis.” AIP Advances 13, 105308 (2023). Extends the second law of infodynamics to atomic physics (Hund’s rule as information entropy minimization), cosmology, and geometric symmetry (symmetric objects have lower information entropy). The empirical results are well documented; the simulation hypothesis conclusion is the author’s interpretation. Referenced in Chapters 7, 15, and 17.

Vopson, Melvin M. and Serban Lepadatu. “The second law of information dynamics.” AIP Advances 12, 075310 (2022). Proposes a complement to the second law of thermodynamics: the information entropy of systems containing distinguishable information states decreases over time, converging toward a minimum at equilibrium. Demonstrated via micromagnetic Monte Carlo simulation of digital data self-erasure. Referenced in Chapters 7 and 15.

Vopson, Melvin M. and Stuart Robson. “GENIES: GENetic Information Entropy Software.” arXiv:2107.07481 (2021). Software tool for computing Shannon information entropy from nucleotide sequences. Used to compute the SARS-CoV-2 entropy data underlying the second law of infodynamics genetic application. Referenced in Chapter 7.

Vormberg, Andreas, Felix Felmy, Ruxandra Bhatt, et al. “Universal features of dendrites through centripetal branch ordering.” PLOS Computational Biology 13(7): e1005615 (2017). Centripetal Strahler analysis of 75,000+ reconstructed neurons reveals systematic variation in bifurcation ratios by cell type: granule cells (R_B = 2.23) through lobula plate tangential cells (R_B = 3.77). The variation correlates with computational role. Referenced in Chapters 3 and 15.

Wheeler, John Archibald. “Information, Physics, Quantum: The Search for Links.” In Complexity, Entropy, and the Physics of Information (1990). The “It from Bit” proposal.

Wheeler, John Archibald. “Law Without Law.” In Wheeler, J.A. and Zurek, W.H. (eds.), Quantum Theory and Measurement (1983). Princeton University Press. Contains Wheeler’s variant of Twenty Questions, in which no answer is predetermined and the “object” emerges from the questioning: his most accessible illustration of the participatory anthropic principle.

Yang, Chen-Ning and Robert L. Mills. “Conservation of Isotopic Spin and Isotopic Gauge Invariance.” Physical Review 96, no. 1 (1954): 191-195. Extended gauge invariance from electromagnetism to non-Abelian symmetry groups. The foundation of the Standard Model and the mathematical framework for local symmetry in field theory.

Zalamea, Fernando. Synthetic Philosophy of Contemporary Mathematics. Urbanomic/Sequence Press (2012; English trans. of 2009 Spanish original). Positions Grothendieck’s methods as philosophical methodology: systematic geometrization across mathematical domains, schematization beyond set-theoretic constraints, and the use of sheaf logic as a general tool for understanding structural relationships.

Zhirnov, Victor, Reza M. Zadegan, Gurtej S. Sandhu, George M. Church, and William L. Hughes. “Nucleic acid memory.” Nature Materials 15 (2016): 366–370. DOI: 10.1038/nmat4594. Source of the one-exabyte-per-cubic-millimeter figure for DNA storage, which is a theoretical ceiling for data written directly into the base sequence. The paper is explicit that attained densities fall far short, since approaches using DNA secondary structure require on the order of a hundred base pairs per bit. Cited in “The Computational Universe”; the figure should not be attributed to Church, Gao, and Kosuri (2012), which is the demonstrated write-and-read and sits orders of magnitude below the ceiling.

Emergent Time and Quantum Clocks

Calcinari, Andrea and Steffen Gielen. “Relational dynamics and Page-Wootters formalism in group field theory.” Quantum 9:1610 (2025). Applies the Page-Wootters mechanism to discrete quantum spacetime, deriving an expanding universe from matter-clock correlations.

Castro-Ruiz, Esteban, Flaminia Giacomini, and Časlav Brukner. “Entanglement of quantum clocks through gravity.” Proceedings of the National Academy of Sciences 114, no. 12 (2017): E2303–E2309. DOI: 10.1073/pnas.1616427114. Demonstrates that gravitational time dilation entangles quantum clocks, establishing that gravity itself triggers the Page-Wootters mechanism.

Christodoulou, Demetrios. “Nonlinear Nature of Gravitation and Gravitational-Wave Experiments.” Physical Review Letters 67 (1991): 1486–1489. Establishes the nonlinear gravitational-wave memory effect: the permanent displacement of free test masses left behind after a gravitational wave passes, arising from the energy the wave itself carries. Referenced in Chapter 15.

Coppo, Alessandro, Alessandro Cuccoli, and Paola Verrucchi. “A magnetic clock for a harmonic oscillator.” Physical Review A 109 (2024): 052212. Extends the Page-Wootters framework to show classical phase-space trajectories emerge naturally when the clock is macroscopic.

Coppo, Alessandro, Niccolò Pranzini, and Paola Verrucchi. “Quantum model for black holes and clocks.” arXiv:2601.07437 (2026). Shows that a bipartite quantum system modeling a test particle near a Schwarzschild horizon and Hawking radiation satisfies exactly the conditions required by the Page-Wootters mechanism for identifying an evolving system and an associated clock.

Favata, Marc. “The gravitational-wave memory effect.” Classical and Quantum Gravity 27 (2010): 084036. arXiv:1003.3486. Review of the linear and nonlinear gravitational-wave memory effects and their detectability. Referenced in Chapter 15.

Foti, Caterina, Alessandro Coppo, Giulio Barni, Alessandro Cuccoli, and Paola Verrucchi. “Time and classical equations of motion from quantum entanglement via the Page and Wootters mechanism with generalized coherent states.” Nature Communications 12 (2021): 1787. Derives both the Schrödinger equation and Hamilton’s classical equations of motion purely from entanglement between a system and a quantum clock.

Jacobson, Ted. “Thermodynamics of spacetime: The Einstein equation of state.” Physical Review Letters 75 (1995): 1260–1263. Derives Einstein’s field equations from the thermodynamics of local Rindler horizons. Updated in “Entanglement equilibrium and the Einstein equation.” Physical Review Letters 116 (2016): 201101.

Ladghami, Yassine, Francisco S.N. Lobo, Tarik Ouali, et al. “Timelike Entanglement Entropy of Hawking Radiation.” arXiv:2602.06833 (2026). Defines temporal correlations within Hawking radiation, revealing periodic “timelike Page times.”

Moreva, Ekaterina, et al. “Time from quantum entanglement: An experimental illustration.” Physical Review A 89 (2014): 052122. First experimental demonstration of the Page-Wootters mechanism using entangled photons.

Padmanabhan, T. “Thermodynamical Aspects of Gravity: New insights.” Reports on Progress in Physics 73 (2010): 046901. Extends Jacobson’s program: derives gravitational field equations from thermodynamic extremal principles across multiple settings (Einstein, Gauss-Bonnet, Lanczos-Lovelock). Argues that the connection between gravity and thermodynamics is deeper than any single derivation, pointing toward gravity as an emergent, long-wavelength description of underlying microscopic degrees of freedom. The bridge between Jacobson (1995) and Verlinde (2011).

Page, Don N. and William K. Wootters. “Evolution without evolution: Dynamics described by stationary observables.” Physical Review D 27 (1983): 2885–2892. The foundational paper proposing that time emerges from quantum entanglement between subsystems of a static universe.

Pearson, A.N., et al. “Measuring the Thermodynamic Cost of Timekeeping.” Physical Review X 11 (2021): 021029. Using a silicon nitride membrane as a mesoscopic clock, demonstrated that the entropy cost of timekeeping scales linearly with clock accuracy.

Pikovski, Igor, Magdalena Zych, Fabio Costa, and Časlav Brukner. “Universal decoherence due to gravitational time dilation.” Nature Physics 11 (2015): 668–672. Shows gravitational time dilation universally decoheres composite quantum systems, producing classicality and the arrow of time without environmental interaction.

Susskind, Leonard. “Computational complexity and black hole horizons.” Fortschritte der Physik 64 (2016): 24–43. Proposes that the growth of the Einstein-Rosen bridge interior corresponds to growth of quantum computational complexity. See also Brown, A.R. et al., “Complexity, action, and black holes.” Physical Review D 93 (2016): 086006.

Thorne, Kip S. “Gravitational-wave bursts with memory: The Christodoulou effect.” Physical Review D 45 (1992): 520–524. Develops the observational form of Christodoulou’s memory effect for gravitational-wave detectors. Referenced in Chapter 15.

Wadhia, Vivek, et al. “Entropic Costs of Extracting Classical Ticks from a Quantum Clock.” Physical Review Letters 135 (2025): 200407. Double quantum dot experiment showing that reading a clock costs up to a billion times more energy than the clock’s internal ticking mechanism.

Weberszpil, J. and O. Sotolongo-Costa. “Entropy as a Clock: Foundations and Parametrizations of Emergent Time.” International Journal of Theoretical Physics 64 (2025): 48. Unifies entanglement entropy growth, thermal modular flow, and the Page-Wootters mechanism into a single entropic-time framework.

Cosmic Alignment and Large-Scale Structure

Choptuik, Matthew W. “Universality and Scaling in Gravitational Collapse of a Massless Scalar Field.” Physical Review Letters 70, no. 1 (1993): 9–12. Discovered critical gravitational collapse: at the fine-tuned threshold between matter dispersing and forming a black hole, the field exhibits discrete self-similarity (periodic echoing in space and time) and the resulting black-hole mass scales as a universal power law, with critical exponent and echoing period independent of the initial data. Found numerically via adaptive-mesh supercomputer simulation. Referenced in Chapter 9.

Ecker, Christian, Florian Ecker, and Daniel Grumiller. “Analytic Discrete Self-Similar Solutions of Einstein-Klein-Gordon at Large D.” Physical Review Letters (2026). arXiv:2601.14358. First closed-form analytic solutions for the discretely self-similar critical-collapse spacetimes (“spacetime crystals”) known only numerically since Choptuik (1993), obtained via the large-D expansion: solve in many spatial dimensions, where the field equations simplify, then return to four. An infinite family of threshold solutions, compared against finite-D numerical critical solutions. A companion paper (with T. Jechtl, arXiv:2602.10185) coins “critical spacetime crystals” and constructs them numerically in continuous dimension. Referenced in Chapter 9 as the most fundamental instance of universality at a critical threshold.

Egan, Chas A. and Charles H. Lineweaver. “A Larger Estimate of the Entropy of the Universe.” Astrophysical Journal 710 (2010): 1825–1834. Supermassive black holes dominate the cosmic entropy budget.

Gundlach, Carsten. “Understanding Critical Collapse of a Scalar Field.” Physical Review D 55, no. 2 (1997): 695–713. arXiv:gr-qc/9604019. Established the canonical values for massless-scalar critical collapse: critical exponent γ ≈ 0.374 and echoing period Δ ≈ 3.4453, with universality explained by the single unstable mode of the critical solution. The exponent is universal within a matter model but differs across matter types (radiation fluid: γ ≈ 0.356). Referenced in Chapter 9.

Hatamnia, Hossein, Bahram Mobasher, Sina Taamoli, Jeyhan S. Kartaltepe, Caitlin M. Casey, et al. “Large-Scale Structure in COSMOS-Web: Tracing Galaxy Evolution in the Cosmic Web up to z ∼ 7 with the Largest JWST Survey.” arXiv:2511.10727 (2025). Submitted to The Astrophysical Journal. Weighted kernel-density reconstruction of ~160,000 galaxies traces how a galaxy’s position in the cosmic web shapes its evolution: stellar mass rises with local density at all redshifts, and environmental quenching overtakes internal, mass-driven quenching for low-mass galaxies (M* < 10¹⁰ M_sun) below redshift 0.8.

Hutsemékers, Damien, et al. “Alignment of quasar polarizations with large-scale structures.” Astronomy & Astrophysics 572 (2014): A18. Spin axes of supermassive black holes align parallel to their host large-scale structures over gigaparsec scales, with <1% probability of random occurrence.

Lee, Jounghun, et al. “Mysterious Coherence in Several-megaparsec Scales between Galaxy Rotation and Neighbor Motion.” Astrophysical Journal 884 (2019): 104. Galaxy rotation directions correlate with neighbor motions out to 6 Mpc.

NANOGrav Collaboration (Agazie, Gabriella, et al.). “The NANOGrav 15 yr Data Set: Evidence for a Gravitational-wave Background.” Astrophysical Journal Letters 951 (2023): L8. Fifteen years of precision timing from 68 pulsars revealed a stochastic gravitational wave background consistent with merging supermassive black hole binaries.

Pelgrims, Vincent and Damien Hutsemékers. “Polarization alignments of quasars at radio and optical wavelengths.” Astronomy & Astrophysics 585 (2016): A32. Confirmed quasar polarization alignment at radio wavelengths.

Penrose, Roger. “Singularities and Time-Asymmetry.” In General Relativity: An Einstein Centenary Survey, ed. S.W. Hawking and W. Israel. Cambridge University Press (1979). The Weyl curvature hypothesis: low Weyl curvature at the Big Bang, increasing via gravitational clumping toward black holes, drives the cosmological arrow of time.

Tudorache, M. N., S. L. Jung, M. J. Jarvis, I. Heywood, A. A. Ponomareva, A. A. Varăsteanu, N. Maddox, T. Yasin, and M. Glowacki. “A 15 Mpc rotating galaxy filament at redshift z = 0.032.” Monthly Notices of the Royal Astronomical Society 544, no. 4 (2025): 4306–4316. DOI: 10.1093/mnras/staf2005. A ~15 Mpc filament rotating at ~110 km/s, with 14 galaxies spinning in synchrony along its length: a specific rotating filament where galaxies spin in sync with the filament itself, demonstrating angular momentum inheritance from the cosmic web.

Wang, Peng, et al. “Possible observational evidence for cosmic filament spin.” Nature Astronomy 5 (2021): 839–845. Coherent vortical motion detected in stacked cosmic filaments, making them the largest known rotating structures.

Zee, Woong-Bae G., S. Lyla Jung, Sanjaya Paudel, and Suk-Jin Yoon. “Warped Disk Galaxies. II. From the Cosmic Web to the Galactic Warp.” arXiv:2510.18942 (2025). SDSS sample of 244 S-type and 127 U-type warped disks. Warp incidence rises toward filaments (within ~4 Mpc/h), and satellites of S-type warps align with the nearest filament, evidence that the large-scale tide setting galactic spin also bends the disk plane.

Complexity and Cellular Automata

Bak, Per. How Nature Works: The Science of Self-Organized Criticality (1996). The accessible introduction to self-organized criticality.

Bak, Per, Chao Tang, and Kurt Wiesenfeld. “Self-organized criticality.” Physical Review A 38 (1988): 364-374. The original paper introducing self-organized criticality through the sandpile model.

Baldwin, Carliss Y. and Kim B. Clark. Design Rules, Volume 1: The Power of Modularity. MIT Press (2000). Demonstrates that modular architectures enable parallel innovation and rapid evolution in complex systems. Simon’s near-decomposability criterion confirmed in software architecture. Referenced in Chapter 4.

Ball, Philip. How Life Works: A User’s Guide to the New Biology. University of Chicago Press (2023). Challenges the gene-centric view of biology, arguing that life is better understood through self-organization, emergence, and the active role of cells in interpreting genetic information — themes that converge with this book’s treatment of dissipative structures and constructal design.

Bi, Dapeng, J.H. Lopez, J.M. Schwarz, and M. Lisa Manning. “A density-independent rigidity transition in biological tissues.” Nature Physics 11 (2015): 1074–1079. Predicted from a vertex model that tissue undergoes a solid-to-fluid (jamming) transition at a shape index of 3.81, independent of cell density — a departure from classical jamming theory. Experimentally confirmed in lung epithelial cells by Park et al., Nature Materials 14 (2015): 1040–1048. The shape index provides a single measurable order parameter for the metastable boundary in living tissue.

Broido, Anna D. and Aaron Clauset. “Scale-free networks are rare.” Nature Communications 10 (2019): 1017. Statistical tests on ~1,000 social, biological, technological, and information networks found robust evidence that strongly scale-free structure is empirically rare — log-normal distributions fit the data as well or better than power laws in most cases. Caution against casual invocations of “scale-free universality.”

Cavagna, Andrea, et al. “Natural swarms in 3.99 dimensions.” Nature Physics (2023). Applied renormalization group methods to insect swarms, calculating dynamic critical exponent z=1.35 in 3D — one of the first successful tests of rigorous universality in active biological systems. A novel fixed point where both activity and inertia are relevant matched experiments and simulations.

Conway, John. Game of Life first published in Martin Gardner’s “Mathematical Games” column, Scientific American (October 1970).

Cook, Matthew. “Universality in Elementary Cellular Automata.” Complex Systems 15 (2004): 1-40. Proof that Rule 110 is Turing complete.

Fong, Brendan and David I. Spivak. An Invitation to Applied Category Theory: Seven Sketches in Compositionality. Cambridge University Press (2019). Accessible introduction to category theory as a language for compositionality — how complex systems are built from simpler parts with well-defined interfaces.

Friedl, Peter, et al. “Migration of coordinated cell clusters in mesenchymal and epithelial cancer explants in vitro.” Cancer Research 55 (1995): 4557–4560. First observation that cancer cells can migrate collectively as coordinated clusters, not only as individual cells — raising the possibility that metastasis involves an unjamming phase transition rather than individual cell transformation.

Goles, Eric, Oliver Schulz, and Mario Markus. “Prime number selection of cycles in a predator-prey model.” Complexity 6, no. 4 (2001): 33–38. A simple predator-prey simulation that functions as a prime number generator: starting from any initial cycle length, evolutionary dynamics converge on prime-numbered cycles as stable fixed points. Extended in Campos, Paulo R. A., Viviane M. de Oliveira, Ronaldo Giro, and Douglas S. Galvão, “Emergence of Prime Numbers as the Result of Evolutionary Strategy,” Physical Review Letters 93 (2004): 098107. DOI: 10.1103/PhysRevLett.93.098107. Demonstrates that evolutionary pressure alone, without any encoding of number theory, produces prime periodicity as an emergent attractor.

Gorard, Jonathan. “Some Relativistic and Gravitational Properties of the Wolfram Model.” arXiv:2004.14810 (2020). Derives discrete Einstein field equations from Wolfram’s hypergraph model under three assumptions: causal invariance, asymptotic dimensionality preservation (the spatial hypergraph converges to a manifold of fixed dimension), and weak ergodicity. The derivation parallels the Chapman-Enskog procedure for recovering continuum hydrodynamics from molecular dynamics. Only the first assumption follows from causal invariance itself; the latter two are additional geometric constraints that Zhuravlev (2026) found are not satisfied by generic dynamically nontrivial hypergraph rules.

Jones, Jeff. From Pattern Formation to Material Computation (Springer, 2015). Work at the International Centre for Unconventional Computing demonstrating how slime molds solve optimization problems through thermodynamic processes.

Kadanoff, Leo P. “Scaling Laws for Ising Models Near T_c.” Physics 2, no. 6 (1966): 263-272. Introduced the block-spin renormalization group concept: coarse-graining microscopic degrees of freedom to derive effective macroscopic theories. The foundation of universality — different microscopic systems flowing to the same macroscopic fixed point.

Langton, Chris. “Computation at the Edge of Chaos: Phase Transitions and Emergent Computation.” Physica D 42 (1990): 12-37. Edge of chaos and emergent computation.

Lin, Henry W., Max Tegmark, and David Rolnick. “Why Does Deep and Cheap Learning Work So Well?” Journal of Statistical Physics 168 (2017): 1223–1247. arXiv:1608.08225. Formalizes why deep neural networks succeed on physical data: the Hamiltonians governing physical systems exhibit polynomial locality (interactions involve bounded numbers of variables) and hierarchical structure across scales. Neural network architectures that mirror this hierarchy exploit the same structure that renormalization exploits in physics. The result connects the practical success of deep learning to the mathematical structure of the physical world, providing an independent bridge between the renormalization group framework (Wilson, Kadanoff) and the encoder-as-coarse-grainer identity (Vanchurin). The claim that architecture directly mirrors physical hierarchy is contested; gradient descent’s implicit regularization may do some of the work independently. Referenced in Chapter 17.

Rajasekaran, Prithvi. “The Architecture of Autonomy: Harness Design for Long-Running Application Development.” Anthropic Labs (2026). Internal research presentation. Describes a multi-agent architecture (Planner, Generator, Evaluator) for autonomous software development, documenting two failure modes of single-agent systems (context anxiety: premature task truncation from perceived resource scarcity; leniency bias: consistent self-evaluation inflation) and the V1-to-V2 architectural evolution as the underlying model improved. Key principles: separate creation from critique, constrain deliverables not paths, treat architecture as a temporary hypothesis. Referenced in Chapters 3, 11, and 21.

Reid, Chris R., Matthew J. Lutz, Scott Powell, Albert B. Kao, Iain D. Couzin, and Simon Garnier. “Army ants dynamically adjust living bridges in response to a cost-benefit trade-off.” PNAS 112(49) (2015): 15113–15118. Army ant bridges form, grow, and dismantle through purely local decisions — individual ants join when they sense traffic overhead and leave when it stops, maximizing colony-level foraging efficiency without central coordination.

Reynolds, Craig. “Flocks, Herds, and Schools: A Distributed Behavioral Model.” Computer Graphics 21 (1987): 25-34. The Boids algorithm for emergent flocking behavior.

Ries, Eric. Incorruptible: The Treachery of the Invisible Hand and the Architecture of Institutional Longevity (Currency, 2026). Studies why organizations with different founders, industries, and cultures converge on the same extractive end-state, identifying “financial gravity” as the tendency of concentrated capital to deform institutions toward short-term extraction. Documents that alternative structures (steward ownership, industrial foundations, mutual organizations) outperform extraction-optimized competitors on longevity, financial returns, employee welfare, and environmental impact. The concept of the “spiritual holding company” (a mission-holding nonprofit at the center of a for-profit subsidiary) maps to the cognition/regulation dyad (Chapter 8). Ries’s independent convergence on “invitation outperforms coercion” from corporate governance data, with no thermodynamic framework, is evidence for the Trust Attractor’s basin (Chapter 17).

Simon, Herbert A. “The Architecture of Complexity.” Proceedings of the American Philosophical Society 106(6) (1962): 467-482. Classic argument that complex systems are nearly decomposable into hierarchical modules, and that hierarchical organization evolves far faster than non-hierarchical alternatives.

Sutherland, Garret. “Quantum-Physical Softmax via Rydberg Blockade: Experimental Validation of T³ Semantic Geometry on QuEra Aquila.” MirrorEthic LLC (2026). Unpublished manuscript. Two experiments on QuEra Aquila neutral-atom quantum hardware encoding T³ semantic geometry as physical atom positions. v3 (16 atoms, 5,000 shots): Pearson r = 0.646 (p < 10^-6) between predicted and observed quantum correlations for emotional concept pairs. v5 (64 atoms, 10,000 shots): ground states with 86% less semantic conflict than random configurations; spontaneous emergence of Russell’s circumplex model of affect from pure geometry. Key insight: Rydberg blockade (C₆/r⁶ van der Waals interaction) implements winner-take-all dynamics identical to transformer softmax, but in O(1) via parallel physics rather than O(n²) computation. Demonstrates two coupling laws for cognitive architectures: cooperative (1/r², binding/memory) and competitive (C₆/r⁶, selection/attention). T³ cellular automata experiments confirmed that cooperative coupling alone produces autocatalysis but not differentiation; adding competitive coupling yields spontaneous specialization. Referenced in Chapters 3, 22, and 23c.

Tarnita, Corina E. and Arne Traulsen. “Reconciling ecology and evolutionary game theory or ‘When not to think cooperation’.” Proceedings of the National Academy of Sciences 122(14) (2025): e2413847122. doi:10.1073/pnas.2413847122. A perspective arguing that evolutionary game theory often over-invokes cooperation; once realistic ecological dynamics (population regulation, demographic stochasticity, spatial structure) are added, cooperation does not dominate as readily as simplified models suggest. Cited in the objections appendix precisely for naming the limits of cooperation, so that cross-cultural and biological convergence on coordination is read as honest engagement rather than borrowed authority.

Theraulaz, Guy, Jacques Gautrais, Scott Camazine, and Jean-Louis Deneubourg. “The formation of spatial patterns in social insects: from simple behaviors to complex structures.” Philosophical Transactions of the Royal Society A 361 (2003): 1263–1282. Three-rule model of ant nest construction (constant pickup, preferential deposition, pheromone preference) reproduces multilayered architecture with emergent tunnel networks. See also Khuong et al., PNAS 113(5) (2016): 1303–1308, for high-resolution tracking of individual building decisions.

Torquato, Salvatore, et al. “Hyperuniformity and its generalizations.” Physical Review E 94 (2016): 022122. Identification of hyperuniformity as a hidden order between crystalline and random states.

Torquato, Salvatore, Ge Zhang, and Matthew de Courcy-Ireland. “Uncovering multiscale order in the prime numbers via scattering.” Journal of Physics A 51 (2018): 093001. Treated prime number sequences as one-dimensional particle systems and performed computational X-ray diffraction. The resulting fractal-like Bragg peak pattern (“effective limit-periodicity”) constitutes a new category of order distinct from both crystals and quasicrystals. Even the primes reveal hidden structure when examined with the right lens.

Villegas, Pablo, et al. “Laplacian renormalization group for heterogeneous networks.” Nature Physics (2023). Develops Laplacian renormalization for heterogeneous networks, identifying proper spatiotemporal scales through Kadanoff supernodes. Addresses small-world effects that create spurious cross-scale correlations. Provides a formal method to validate or reject cross-scale pattern claims.

Webb, Glenn F. “The prime number periodical cicada problem.” Discrete and Continuous Dynamical Systems — Series B 1, no. 3 (2001): 387–399. Mathematical model showing that among life cycles from 10 to 18 years, only the primes 13 and 17 produce stable populations. The mechanism is the lowest common multiple: LCM(p, q) = pq when p is prime and q ≠ p, maximizing the interval between dangerous synchronizations with shorter cycles. Referenced in Chapter 5.

Weijs, J.H., et al. “Emergent hyperuniformity in periodically driven emulsions.” Physical Review Letters 115 (2015): 108301. Demonstrates that periodic driving forces can produce hyperuniform states in disordered systems — order emerging from repetitive energy input rather than equilibrium.

Wilson, Kenneth G. “Renormalization Group and Critical Phenomena.” Physical Review B 4, no. 9 (1971): 3174-3205. Nobel Prize-winning formalization of the renormalization group for critical phenomena. Showed that phase transitions exhibit universality (systems with different microscopic details sharing the same critical exponents) because different systems flow to the same RG fixed point.

Wolfram, Stephen. “Games between Programs: The Ruliology of Competition.” Stephen Wolfram Writings (June 4, 2026). Systematically enumerates strategies for iterated two-player games (finite state machines, cellular automata, Turing machines), rather than studying hand-picked entries as in Axelrod’s tournaments. For the iterated prisoner’s dilemma, the round-robin winner across two-state strategies is “grim trigger” (cooperate until the opponent’s first defection, then defect permanently), with tit-for-tat ranking well down the field; Wolfram reads this as evidence that Axelrod’s celebrated result depended on which strategies were submitted. The setup is deterministic and noise-free, which favors unforgiving strategies; under execution noise, forgiving reciprocity regains its advantage. Engaged in Chapter 17: the enumeration vindicates conditional cooperation (grim trigger is itself nice and retaliatory) while relocating the forgiveness question to the noise level of the environment. Wolfram’s broader conclusion, that competitive outcomes are computationally irreducible and resist closed-form theorems, converges with this book’s use of irreducibility against coercion.

Wolfram, Stephen. A New Kind of Science (2002). Comprehensive study of cellular automata and the computational universe thesis.

Wolfram, Stephen. “What Ultimately Is There? Metaphysics and the Ruliad.” Stephen Wolfram Writings (February 4, 2026). Argues that the ruliad (the entangled limit of all possible computational processes) is a necessary abstract object from which physics, mathematics, and objective reality emerge through observer sampling. Key claims: observers who are computationally bounded and believe themselves persistent in time must perceive general relativity, quantum mechanics, and the Second Law; the laws of physics are inevitable for observers like us, not intrinsic to the ruliad itself; and the ruliad provides a formal answer to “why does anything exist?” (it is abstractly necessary). Convergent with this book’s thesis: the directionality from entropy to coordination is structurally inevitable, and different minds (biological, artificial) sample different regions of the same underlying computational reality.

Evolution and Complexity

Alem, Sylvain, Clint J. Perry, Xingfu Zhu, Olli J. Loukola, Thomas Ingraham, Eirik Sovik, and Lars Chittka. “Associative mechanisms allow for social learning and cultural transmission of string pulling in an insect.” PLoS Biology 14(10): e1002564 (2016). doi:10.1371/journal.pbio.1002564. Bumblebees learned to pull strings to access out-of-reach rewards; the behavior spread culturally through colonies via observation. Queen Mary University of London (Chittka lab). Referenced in Chapter 22.

Alseth, Ellinor O., Elizabeth Pursey, Adela M. Luján, Isobel McLeod, Clare Rollie, and Edze R. Westra. “Bacterial biodiversity drives the evolution of CRISPR-based phage resistance.” Nature 574 (2019): 549–552. Demonstrated that Pseudomonas aeruginosa bacteria in monoculture evolve surface-based phage resistance (receptor mutations), but in communities with three competitor species shift decisively toward CRISPR-based defense. Community context makes the flexible strategy favored because receptor mutations disable nutrient uptake needed for competition. Bacteria with surface-based resistance were also less virulent in moth larvae hosts. Referenced as ^21b in Chapter 6.

Blount, Zachary D., Richard E. Lenski, and Jonathan B. Losos. “Contingency and determinism in evolution: replaying life’s tape.” Science 362: eaam5979 (2018). Comprehensive review of replay experiments across bacteria, viruses, and multicellular organisms. Concludes that parallel outcomes are common for simple traits but complex innovations retain historical contingency. Referenced as ^{replay-review} in Chapter 7.

Bridges, Alice D., Amanda Royka, Tatyana Wilson, Charlotte Lockwood, Jonathan Richter, Mikko Juusola, and Lars Chittka. “Bumblebees socially learn behaviour too complex to innovate alone.” Nature 627 (2024): 572-578. doi:10.1038/s41586-024-07126-4. Bumblebees trained by a demonstrator on a two-step puzzle box (pushing tabs in a specific sequence) mastered and transmitted the full solution to naive partners, demonstrating cumulative culture: behavior too complex for any individual to discover alone, propagated through social learning. No bee solved the puzzle independently. Queen Mary University of London (Chittka lab). Referenced in Chapters 3 and 22.

Campbell, John O. “Universal Darwinism as a process of Bayesian inference.” Frontiers in Systems Neuroscience 10 (2016): 49. Formalization showing evolution implements the same mathematical structure as Bayesian learning.

Conway Morris, Simon. Life’s Solution: Inevitable Humans in a Lonely Universe (2003). Convergent evolution and deep constraints on evolutionary outcomes.

Cox, R. Thomas and Charles E. Carlton. “A Commentary on Prime Numbers and Life Cycles of Periodical Cicadas.” The American Naturalist 132, no. 2 (1988): 312–316. Proposed that prime-numbered cicada cycles minimize hybridization between broods rather than predator avoidance. Noted that no cycling predator of periodical cicadas has been identified, challenging the predator-avoidance hypothesis first articulated by May (1979). Referenced in Chapter 5.

Darveau, Charles-Antoine, et al. “Diapausing bumble bee queens avoid drowning by using underwater respiration, anaerobic metabolism and profound metabolic depression.” Proceedings of the Royal Society B 293(2066): 20253141 (2025). doi:10.1098/rspb.2025.3141. Mechanistic follow-up to Rondeau and Raine (2024): queens extract oxygen from water via cuticular gas exchange, switch to anaerobic metabolism (lactate accumulation), and enter profound metabolic depression (CO₂ production drops from ~15.4 to ~2.4 µL/hr/g). University of Ottawa. Referenced in Chapter 18.

Darwin, Charles. On the Origin of Species by Means of Natural Selection (1859). Foundation of evolutionary theory.

Dennett, Daniel C. Darwin’s Dangerous Idea: Evolution and the Meanings of Life (1995). Dennett argues that natural selection is an algorithm: substrate-neutral, mindless, and reliable wherever its preconditions (variation, selection, heredity) are met. This book extends the move to the entropic cascade — the precedent grounds the title’s claim. See Opening, note 9.

Gould, Stephen Jay. “Nonoverlapping Magisteria,” Natural History 106 (March 1997): 16-22. Reprinted in Leonardo’s Mountain of Clams and the Diet of Worms (1998). Gould’s influential proposal that science and religion occupy separate, non-overlapping domains of authority. This book argues the boundary was institutional, not ontological.

Gould, Stephen Jay. Wonderful Life: The Burgess Shale and the Nature of History (1989). Contingency in evolution and the “tape of life” thought experiment.

Hubbell, Stephen P. The Unified Neutral Theory of Biodiversity and Biogeography. Princeton University Press (2001). Extended Kimura’s neutral framework from genetics to ecology, arguing that demographic stochasticity (ecological drift) can explain species abundance patterns without invoking species-specific niche differences. Controversial, productive, and influential as a null model for distinguishing neutral from selective forces in community ecology.

Joannes-Boyau, Renaud, Janaina Sena de Souza, Manish Arora, Christine Austin, Kira Westaway, Ian Moffat, Wei Wang, Wei Liao, Yingqi Zhang, Justin W. Adams, Luca Fiorenza, Flora Dérognat, Marie-Helene Moncel, Gary T. Schwartz, Marian Bailey, et al. “Impact of intermittent lead exposure on hominid brain evolution.” Science Advances 11:42 (2025): eadr1524. doi: 10.1126/sciadv.adr1524. Laser-ablation geochemistry of 51 fossil teeth from seven hominid and primate groups spanning two million years across Africa, Asia, and Europe found episodic lead exposure in 73% of specimens, from naturally occurring volcanic, erosional, and wildfire sources rather than anthropogenic contamination. Brain organoids carrying the archaic NOVA1 variant showed severe disruption of FOXP2 expression under lead exposure; organoids with the modern human NOVA1 variant were markedly more resilient. The modern NOVA1 allele differs by a single amino acid substitution and regulates alternative splicing of genes critical for synaptogenesis and neuronal migration. Proposes that pervasive environmental lead acted as a long-term selective pressure favoring the modern variant, conferring neurocognitive resilience and a survival advantage through preserved social cohesion and communication capacity, potentially contributing to the extinction of Neanderthals and other archaic hominids.

Kacian, D.L., D.R. Mills, F.R. Kramer, and S. Spiegelman. “A replicating RNA molecule suitable for a detailed analysis of extracellular evolution and replication.” Proceedings of the National Academy of Sciences 69(10) (1972): 3038-3042. Spiegelman’s classic experiment: a virus genome kept in ideal replication conditions shrank from 4,500 base pairs to 218 over 74 generations, a 95% reduction toward informational simplicity. Consistent with information entropy minimization over evolutionary time. Referenced in Chapters 7 and 15.

Kafetzis, George, Michael J. Bok, Tom Baden, and Dan-Eric Nilsson. “Evolution of the vertebrate retina by repurposing of a composite ancestral median eye.” Current Biology (2026). DOI: 10.1016/j.cub.2025.12.028. Survey of 36 major bilateral animal groups tracing vertebrate lateral eyes to a repurposed median photoreceptor. An ancestral deuterostome retained the median eye for circadian sensing through a sessile filter-feeding phase (~600 Mya) while lateral eyes atrophied; the median eye subsequently split into the paired eyes of vertebrates. The inverted vertebrate retina, with photoreceptors behind neural wiring, optimizes metabolic throughput (direct contact with retinal pigment epithelium) rather than signal-path elegance. The pineal gland is the surviving remnant of the original median eye. Referenced in Chapters 3 and 9.

Kaila, Ville R.I. and Arto Annila. “Natural Selection for Least Action.” Proceedings of the Royal Society A 464, no. 2099 (2008): 3055-3070. Formally connects evolution via natural selection to a variational principle. The second law written in integral form yields a least-action principle; random variation explores paths, and selection favors those that increase entropy production fastest. The closest existing work to a formal bridge between evolutionary dynamics and path integral formalism.

Karban, Richard, Cynthia A. Black, and Sara A. Weinbaum. “How 17-year cicadas keep track of time.” Ecology Letters 3, no. 4 (2000): 253–256. Demonstrated experimentally that periodical cicada nymphs count annual xylem sap pulses rather than elapsed calendar time. Forced host trees to produce two leaf flushes in one year; cicadas emerged one year early. The molecular mechanism for tallying pulses remains unidentified. Referenced in Chapter 5.

Kauffman, Stuart A. The Origins of Order: Self-Organization and Selection in Evolution (1993). Complexity theory and self-organization in evolution.

Kauffman, Stuart A. A World Beyond Physics: The Emergence and Evolution of Life. Oxford University Press (2019). Argues that life cannot be reduced to physics alone — that the biosphere’s creativity outruns any finite set of laws. Introduces the concept of the “adjacent possible” as a formal framework for understanding how novelty arises. Supports this book’s thesis that emergence is a constitutive feature of reality, not merely an observation.

Kelly, Kevin. The Inevitable: Understanding the 12 Technological Forces That Will Shape Our Future. New York: Viking, 2016. Identifies twelve deep trends reshaping society through technology, including cognification, tracking, and interacting — extends the technium framework toward human-AI convergence.

Kelly, Kevin. What Technology Wants. New York: Viking, 2010. Argues technology is an extension of biological evolution; frames the technium as a living system with emergent tendencies.

Kimura, Motoo. “Evolutionary rate at the molecular level.” Nature 217 (1968): 624–626. Proposed that most molecular evolution is driven by random fixation of selectively neutral mutations rather than natural selection. The foundational paper of neutral theory.

Kimura, Motoo. The Neutral Theory of Molecular Evolution. Cambridge University Press (1983). Full development of neutral theory, showing that the rate of molecular evolution (amino acid substitutions per year) is remarkably constant across lineages, consistent with neutral drift rather than adaptive selection. Confirmed by large-scale sequencing in the 1990s and 2000s.

Kirschner, Marc and John Gerhart. “The Theory of Facilitated Variation.” PNAS 104(suppl 1) (2007): 8582-8589. Argues that organisms possess conserved core processes whose modularity and weak regulatory linkage make phenotypic variation more likely to be functional — evolution is biased toward viable novelty by the architecture of development itself.

Kryazhimskiy, Sergey, Daniel P. Rice, Elizabeth R. Jerison, and Michael M. Desai. “Global epistasis makes adaptation predictable despite sequence-level stochasticity.” Science 344(6191) (2014): 1519–1522. Evolved 640 independent yeast populations (64 founder genotypes × 10 replicates) for 500 generations. Despite stochastic and divergent mutations, all populations converged on similar fitness levels. The mechanism: global diminishing-returns epistasis funnels divergent molecular paths toward the same phenotypic destination. Referenced as ^desai-yeast in Chapter 7.

Liu, Jingjun, Dalton S. Hardisty, James F. Kasting, Mojtaba Fakhraee, and Noah J. Planavsky. “Evolution of the iodine cycle and the late stabilization of the Earth’s ozone layer.” Proceedings of the National Academy of Sciences 122:2 (2025): e2412898121. Geological and photochemical analysis showing that marine iodine emissions catalytically destroyed atmospheric ozone for approximately two billion years following initial oxygenation, preventing stable UV shielding and restricting complex life to the oceans. Proterozoic ocean iodine concentrations were roughly 180 times present levels; iodine-driven ozone destruction is kinetically faster than chlorine-based (CFC) pathways. The ozone layer stabilized only around 500 million years ago, coinciding with the biological drawdown of marine iodine by kelp, tunicates, and thyroid-bearing vertebrates — life modifying its own boundary conditions.

Loron, Corentin C., Laura M. Cooper, Sean McMahon, Seán F. Jordan, Andrei V. Gromov, Matthew Humpage, Niall Rodgers, Laetitia Pichevin, Hendrik Vondracek, Ruaridh Alexander, Edwin Rodriguez Dzul, Alexander T. Brasier, Michael Krings, and Alexander J. Hetherington. “Prototaxites fossils are structurally and chemically distinct from extinct and extant Fungi.” Science Advances 12:4 (2026): eaec6277. doi:10.1126/sciadv.aec6277. Analysis of Prototaxites taiti from the 407-million-year-old Rhynie chert (Aberdeenshire, Scotland) using microscopic imaging and infrared spectroscopy with machine-learning classification. Found three structurally distinct tube types (including medullary spots with no parallel in fungal biology) and no chitin or perylene (fungal biomarker). Molecular fingerprint most similar to lignin fossilization products, yet the organism was heterotrophic. Conclusion: Prototaxites is best assigned to an entirely extinct eukaryotic lineage, distinct from animals, plants, and fungi, possibly a fourth kingdom of complex multicellular life that dominated terrestrial surfaces for ~50 million years before vanishing in the Late Devonian.

Loukola, Olli J., Anna Antinoja, Kaarle Makela, Janette Arppi, Fei Peng, and Cwyn Solvi. “Evidence for socially influenced and potentially actively coordinated cooperation by bumblebees.” Proceedings of the Royal Society B 291(2022): 20240055 (2024). doi:10.1098/rspb.2024.0055. Buff-tailed bumblebees (Bombus terrestris) trained on cooperative tasks (pushing a Lego block, pushing a door) delayed initiation significantly when a partner’s entry was delayed, indicating anticipatory coordination. University of Oulu, Finland. Referenced in Chapters 19 and 22.

Loukola, Olli J., Cwyn Solvi, Louie Coscos, and Lars Chittka. “Bumblebees show cognitive flexibility by improving on an observed complex behavior.” Science 355(6327) (2017): 833–836. doi:10.1126/science.aag2360. Bumblebees trained to roll a ball to a target location for a sucrose reward improved on the demonstrator’s technique, choosing the nearest ball rather than copying the exact path — demonstrating cognitive flexibility and tool use. Queen Mary University of London (Chittka lab). Referenced in Chapter 22.

Lynch, Michael. “The origins of eukaryotic gene structure.” Molecular Biology and Evolution 23(2) (2006): 450–468. Demonstrates that many features of eukaryotic genomes (introns, gene duplications, mobile elements) are consequences of reduced population size and weakened selection, rather than adaptations. Extended in Lynch, M., The Origins of Genome Architecture (Sinauer, 2007). The nonadaptive theory of genome evolution: drift assembled the regulatory toolkit from which complex coordination later emerged.

May, Robert M. “Periodical cicadas.” Nature 277 (1979): 347–349. Early mathematical treatment of why prime-numbered life cycles are evolutionarily stable, arguing that primes minimize synchronization with shorter predator cycles. Referenced in Chapter 5.

Maynard Smith, John and Eörs Szathmáry. The Major Transitions in Evolution (1995). Oxford University Press. Identifies eight major transitions in which previously independent entities merged into higher-level units, including: RNA to DNA, prokaryotes to eukaryotes, asexual to sexual reproduction, protists to multicellularity, solitary to colonial organisms, primate societies to language-bearing humans.

Murray, James D. Mathematical Biology (Springer, 1993). Foundational treatment of morphogenetic pattern formation, showing how harmonic waves provide the building blocks of biological structure.

Nilsson, Dan-Eric, and Susanne Pelger. “A pessimistic estimate of the time required for an eye to evolve.” Proceedings of the Royal Society B 256 (1994): 53-58. Computational model showing that a camera-type eye can evolve from a flat patch of photosensitive cells in fewer than 400,000 generations through gradual selection on optical resolution. Referenced in Chapter 3.

Prosser, T. M. “A General Theory of Abiogenesis: Thermodynamic Selection and the Emergence of Life.” arXiv preprint 2504.17975 (2025). Introduces the Thermodynamic Abiogenesis Likelihood Model (TALM), formalizing a “reaction viability inequality” showing that chemical systems satisfying thermodynamic persistence conditions can be selected for prior to Darwinian replication; dissipative structuring constitutes a pre-Darwinian selection mechanism. Cited in Chapters 7 and 16.

Rogers, Rebekah L. and Montgomery Slatkin. “Excess of genomic defects in a woolly mammoth on Wrangel Island.” PLOS Genetics 13(3): e1006601 (2017). Identified numerous premature stop codons, splice site mutations, and retrogene insertions in the Wrangel Island mammoth genome relative to mainland mammoths, consistent with genomic meltdown in a small island population (~300 individuals isolated for ~6,000 years). The mammoths were alive; their genomes were dying. Direct evidence of Muller’s ratchet operating through insufficient recombination in a small, isolated mammalian population.

Rondeau, Sabrina and Nigel E. Raine. “Unveiling the submerged secrets: bumblebee queens’ resilience to flooding.” Biology Letters 20(4): 20230609 (2024). doi:10.1098/rsbl.2023.0609. Diapausing Bombus impatiens queens survived complete submersion for up to one week with ~90% survival. Discovery originated from accidental water accumulation in refrigerated containers. University of Guelph, Ontario. Referenced in Chapter 18.

Stanley, Dara A., Karen E. Smith, and Nigel E. Raine. “Bumblebee learning and memory is impaired by chronic exposure to a neonicotinoid pesticide.” Scientific Reports 5: 16508 (2015). doi:10.1038/srep16508. Chronic thiamethoxam exposure at field-realistic concentrations (2.4 ppb) caused significantly slower learning and impaired short-term memory in Bombus terrestris. First demonstration of learning and memory deficits from chronic neonicotinoid exposure at field-realistic levels. Referenced in Chapter 18.

Toivonen, Jaakko and Lutz Fromhage. “Hybridization selects for prime-numbered life cycles in Magicicada.” Ecology and Evolution 10, no. 12 (2020): 5259–5269. DOI: 10.1002/ece3.6270. Individual-based simulations showing that hybrid offspring with intermediate cycle lengths face 49–55% predation mortality versus 6% for non-hybrids, because they emerge at low density without the protection of predator satiation. Prime cycles are favored once population-wide synchronization already exists. Referenced in Chapter 5.

Vidal, Clément. “What is the noosphere?” Systems Research and Behavioral Science (2024). Defines the noosphere as a planetary superorganism using Living Systems Theory; frames it as a Major Evolutionary Transition analogous to the emergence of multicellularity. Cited in Chapter 16.

Wagner, Andreas. Robustness and Evolvability in Living Systems. Princeton University Press (2005). Demonstrates that robust genetic networks (those tolerant of mutations) explore larger regions of genotype space, making them paradoxically more evolvable. Robustness and evolvability are companions, not trade-offs. Confirms Simon’s near-decomposability criterion in modular gene regulatory networks. Referenced in Chapter 4.

Westra, Edze R., Stineke van Houte, Shyam Oyesiku-Blakemore, et al. “Parasite exposure drives selective evolution of constitutive versus inducible defense.” Current Biology 25 (2015): 1043–1049. Earlier work showing that nutrient availability and phage density affect whether bacteria use surface-based or CRISPR-based resistance, establishing the environmental context framework extended in Alseth et al. (2019).

Origin of Complex Life and Symbiosis

Agüera y Arcas, Blaise. “What is Intelligence?” Long Now Foundation Seminar (September 2025). Public presentation of the computational origins of life thesis and its implications for human-AI coevolution.

Agüera y Arcas, Blaise. What Is Intelligence? Lessons from AI About Evolution, Computing, and Minds. MIT Press (2025). Companion volume extending the computational origins framework to intelligence, prediction, and multi-agent coordination across substrates.

Agüera y Arcas, Blaise. What Is Life? Evolution as Computation. MIT Press (2025). VP at Google and founder of Paradigms of Intelligence. Demonstrates that self-replicating programs can emerge spontaneously from random computational noise — a phase transition from “gas” (decorrelated information) to “life” (functional, reproducible structure). Key insight: symbiogenesis is what gives evolution its arrow of time. When two self-replicating systems merge, extra information must be added for how they coordinate, and this coordination information is the source of increasing complexity. Argues that consciousness is functionally necessary for multi-agent coordination: “The reason we are conscious is because we are modeling ourselves as well as modeling others as well as modeling others modeling ourselves… because that is behaviorally essential to allow us to cooperate with each other.” Reframes human-AI relations as computational symbiogenesis rather than dominance competition.

Angulo-Cánovas, Elisa, Ana Bartual, Rocío López-Igual, Ignacio Luque, Nikolai P. Radzinski, Irina Shilova, Megan Anjur-Dietrich, Guadalupe García-Jurado, Bartolomé Úbeda, José A. González-Reyes, Jesús Díez, Sallie W. Chisholm, José Manuel García-Fernández, and María del Carmen Muñoz-Marín. “Direct interaction between marine cyanobacteria mediated by nanotubes.” Science Advances 10:21 (2024): eadj1539. First observation of membrane nanotubes in marine cyanobacteria. Prochlorococcus, the most abundant photosynthetic organism on Earth, forms 100–200 nm tunnels enabling direct cytoplasmic exchange, including cross-genus transfer: labeled Prochlorococcus shared material with Synechococcus within 15 minutes, with over 80% of receiving cells acquiring fluorescence. Approximately 5.5% of cyanobacterial cells in natural ocean samples (Bay of Cádiz) showed active nanotubes, confirming the phenomenon in wild populations. Individual cells connect to multiple partners simultaneously, forming network topologies. Challenges the classification of these organisms as strictly single-celled and suggests a microbial trade network: coordination by physical connection, not coercion.

Appler, Kathryn E., James P. Lingford, Xianzhe Gong, Kassiani Panagiotou, Pedro Leão, Marguerite V. Langwig, Chris Greening, Thijs J. G. Ettema, Valerie De Anda, and Brett J. Baker. “Oxygen metabolism in descendants of the archaeal-eukaryotic ancestor.” Nature (2026). DOI: 10.1038/s41586-026-10128-z. Massive DNA sequencing of marine sediments yielded 404 Asgardarchaeota metagenome-assembled genomes, including 136 new Heimdallarchaeia, the closest archaeal relatives of eukaryotes. Metabolic reconstructions revealed that Heimdallarchaeia encode hallmark aerobic proteins: electron transport chain Complex IV, haem biosynthesis, and reactive oxygen species detoxification, plus novel respiratory membrane-bound hydrogenases with Complex I-like subunits. Proposes an updated eukaryogenesis model in which both hydrogen production and aerobic respiration were present in the Asgard-eukaryotic ancestor, reframing the mitochondrial partnership as a merger between two bioenergetically sophisticated organisms rather than the rescue of a helpless anaerobe.

Aratani, Yuri, Takuya Uemura, Takuma Hagihara, Kenji Matsui, and Masatsugu Toyota. “Green leaf volatile sensory calcium transduction in Arabidopsis.” Nature Communications 14 (2023): 6236. First real-time visualization of plant-to-plant volatile communication. Using transgenic Arabidopsis expressing GCaMP calcium indicators, showed that airborne green leaf volatiles, specifically (Z)-3-hexenal and (E)-2-hexenal, from damaged plants trigger calcium signaling cascades in undamaged neighbors within one minute. Guard cells respond first, mesophyll follows. The warning compounds confer no direct benefit on the sender: pure coordination signals, broadcast without coercion, received without obligation. Demonstrates invitation-based coordination operating across kingdoms via the same calcium signaling that drives neural transmission.

Biller, Steven J., Florence Schubotz, Sara E. Roggensack, Anne W. Thompson, Roger E. Summons, and Sallie W. Chisholm. “Bacterial vesicles in marine ecosystems.” Science 343:6167 (2014): 183–186. Prochlorococcus continuously releases lipid membrane vesicles containing proteins, DNA, and RNA into the ocean. Estimated global production: ~10²⁷ vesicles per day from Prochlorococcus alone, representing a significant addition of organic carbon to the dissolved pool. Vesicles support the growth of heterotrophic bacteria — active resource provisioning of the community, not passive leakage.

Biller, Steven J., Paul M. Berube, Debbie Lindell, and Sallie W. Chisholm. “Prochlorococcus: the structure and function of collective diversity.” Nature Reviews Microbiology 13:1 (2015): 13–27. Describes Prochlorococcus as a “federation” of diverse cells: each individual has a minimal, streamlined genome (~1,700–2,300 genes), but the collective pan-genome is vast. Hundreds of genomically distinct subpopulations coexist, each with different gene sets, collectively covering environmental space no single genotype could. The federation’s stability, abundance, and broad distribution arise from collective diversity — not from any individual lineage’s fitness.

Bublitz, DeAnna C., Grayson L. Chadwick, John S. Magyar, Kelsi M. Sandoz, Diane M. Brooks, Stéphane Mesnage, Mark S. Ladinsky, Arkadiy I. Garber, Victoria J. Orphan, and John P. McCutcheon. “Peptidoglycan production by an insect-bacterial mosaic.” Cell 179 (2019): 703–712. Demonstrated that peptidoglycan synthesis in the deeply nested endosymbiont Moranella requires gene products from three sources (the insect nuclear genome, including horizontally transferred bacterial genes; Tremblaya; and Moranella itself), transported across five lipid membranes. First proof that a biosynthetic pathway in a nested endosymbiont draws on the host’s nuclear genome, “eroding all functional distinction between endosymbiont and organelle.”

Cheng, Xianrui and James E. Ferrell Jr. “Spontaneous emergence of cell-like organization in Xenopus egg extracts.” Science 366:6465 (2019): 631–637. Homogenized frog egg cytoplasm spontaneously reorganizes into cell-like compartments with properly positioned organelles (microtubules, endoplasmic reticulum). Self-organization requires microtubules and dynein motor proteins but not actin. Compartments form and divide using voids as boundaries instead of membranes. Demonstrates that cytoplasmic dynamics encode organizational information independently of the genome. Referenced as ^31g in Chapter 6.

Cosby, Rachel L., Nicolás C. Judd, Ruiling Zhang, April Zhong, Nabil Garber, Orion D. Enriquez, Edward B. Chuong, and Cédric Feschotte. “Recurrent evolution of vertebrate transcription factors by transposase capture.” Science 371, no. 6531 (2021): eabc6405. Identified ~100 distinct fusion genes in tetrapods where a transposable element fused with an established gene during the past 300 million years. The resulting chimeric proteins retain the transposase domain’s affinity for transposon-derived binding sites scattered across the genome, giving them ready-made architecture as transcription factors. Deletion of one such fusion gene (evolved in bats 25–45 million years ago) dysregulated hundreds of genes; restoration normalized activity. Demonstrates a specific molecular mechanism for how parasitic genetic elements become master regulators of gene expression. Referenced as ^22a in Chapter 6.

DeCasien, Alex R., Jacob E. Aronoff, Elizabeth K. Mallott, et al. (Katherine R. Amato, senior author). “Primate gut microbiota induce evolutionarily salient changes in mouse neurodevelopment.” PNAS 123:2 (2026): e2426232122. Mice colonized with gut microbiomes from large-brained primates (humans, squirrel monkeys) showed upregulation of oxidative phosphorylation and synaptic plasticity genes in frontal cortex, while smaller-brained primate microbiomes (macaques) produced gene expression patterns overlapping with neurodevelopmental disorders. Evidence that the holobiont (host plus microbiome) is the relevant unit for understanding brain evolution.

Edgar, Allison, Dorothy G. Mitchell, and Mark Q. Martindale. “Whole-Body Regeneration in the Lobate Ctenophore Mnemiopsis leidyi.” Genes 12:6 (2021): 867. Established the regenerative capacity of M. leidyi, demonstrating whole-body regeneration directed by a polar coordinate system with lineage-restricted replacement cells. Grafting experiments showed that reassembled body segments regenerated a single common oral-aboral axis when oriented natively, but maintained independent polarity when middle segments were inverted — evidence of an intrinsic body-plan memory that persists through dismemberment. Notably, several cell signaling pathways known to function in regeneration in other animals (including Wnt and Hedgehog) are absent from the ctenophore genome, suggesting that M. leidyi uses a regenerative toolkit ancestral to, and distinct from, the mechanisms deployed by later-diverging animals.

Flombaum, Pedro, José L. Gallegos, Rodolfo A. Gordillo, et al. “Present and future global distributions of the marine Cyanobacteria Prochlorococcus and Synechococcus.” Proceedings of the National Academy of Sciences 110:24 (2013): 9824–9829. Global abundance of Prochlorococcus estimated at 2.9 × 10²⁷ cells; Synechococcus at 7.0 × 10²⁶ cells. Together these picocyanobacteria account for approximately 25% of all marine net primary productivity.

Geiger, Otto, Alejandro Sanchez-Flores, Jonathan Padilla-Gomez, and Mauro Degli Esposti. “Multiple approaches of cellular metabolism define the bacterial ancestry of mitochondria.” Science Advances 9 (2023): eadh0066. Comprehensive genomic survey identifying the marine alphaproteobacterium Iodidimonas as the closest living relative of the proto-mitochondrion, based on shared cardiolipin and sphingolipid biosynthesis pathways, COX operon structure, and bc1 complex genetics. Challenges the previous assumption that Rickettsia held this position.

Goldman, Aaron D., Gregory P. Fournier, and Betül Kaçar. “Universal paralogs provide a window into evolution before the last universal common ancestor.” Cell Genomics 6, no. 3: 101140 (February 2026). DOI: 10.1016/j.xgen.2026.101140. Identifies genes duplicated before the Last Universal Common Ancestor, revealing that protein production and membrane transport were among the earliest cellular functions. Pushes the evolutionary record beyond LUCA, demonstrating that coordination was present from the very beginning. Referenced in Chapters 6c and 16.

Imachi, Hiroyuki, et al. “Isolation of an archaeon at the prokaryote–eukaryote interface.” Nature 577 (2020): 519-525. First cultured Asgard archaean (Prometheoarchaeum syntrophicum). Showed tentacle morphology and obligate syntrophy, proposing that eukaryogenesis occurred through partnership rather than phagocytic capture.

Jenkins, Ben, Thomas Richards, et al. “RNA interference may stabilize a eukaryotic endosymbiosis.” bioRxiv preprint (2021). Demonstrated that when Paramecium bursaria digests its Chlorella endosymbionts, released algal RNA (sufficiently similar to host transcripts) triggers the host’s RNAi machinery against its own genes, suppressing growth and reproduction. The mechanism requires no coevolution, only sequence overlap and a functional RNAi system, making it a substrate-independent stabilization principle for nascent endosymbioses. Referenced as ^19a in Chapter 17.

Jokura, Kei, Tomas Anttonen, Mariana Rodriguez-Santiago, and Oscar M. Arenas. “Rapid physiological integration of fused ctenophores.” Current Biology 34:19 (2024): R889–R890. Demonstrated that injured Mnemiopsis leidyi (comb jellies) fuse into single functioning organisms within hours when placed in proximity. Nine of ten independent grafting experiments succeeded; all fused animals survived the full three-week observation period. Muscle contractions synchronized within two hours, indicating nervous system merger — the syncytial nerve net shared action potentials across the graft boundary as though the organisms had never been separate. Digestive systems integrated, distributing nutrients from either oral end equally. The organism appears to lack allorecognition entirely: no immune rejection, no tissue-type discrimination, no foreign-body response. If ctenophores diverged near the base of the animal phylogenetic tree, as molecular evidence increasingly suggests, this absence may represent the ancestral condition: multicellular coordination evolved before self/non-self distinction. The most ancient form of cooperation was open by default.

Karnkowska, Anna, Vojtěch Vacek, Zuzana Zubáčová, et al. “A eukaryote without a mitochondrial organelle.” Current Biology 26:10 (2016): 1274–1284. Genome sequencing of the oxymonad Monocercomonoides sp. revealed the first known eukaryote to have completely lost mitochondria. The organism, an anaerobic gut protist, compensates via a laterally acquired bacterial sulfur mobilization (SUF) system. Demonstrates that the mitochondrial partnership, while nearly universal, can be dissolved — but only under extreme conditions (anaerobic parasitic niche) and at the cost of radical simplification.

Khait, Itzhak, Ohad Lewin-Epstein, Raz Sharon, Kfir Saban, Revital Goldstein, Yehuda Anikster, Yarden Zeron, Chen Agassy, Shaked Nizan, Gayl Sharabi, Ran Perelman, Arjan Boonman, Nir Sade, Yossi Yovel, and Lilach Hadany. “Sounds emitted by plants under stress are airborne and informative.” Cell 186:7 (2023): 1328–1336. Demonstrated that stressed plants emit airborne ultrasonic clicks in the 40–80 kHz range, detectable several meters away. Drought-stressed and physically damaged plants produce distinct acoustic profiles, and machine learning classifiers can distinguish stress type from recordings. A third signaling modality alongside volatile organic compounds and mycorrhizal chemical exchange — the same coordination problem solved through every available physical medium.

Kim, Iana V., Cristina Navarrete, et al. (Arnau Sebé-Pedrós, senior author). “Chromatin loops are an ancestral hallmark of the animal regulatory genome.” Nature 642 (2025): 1097–1105. DOI: 10.1038/s41586-025-08960-w. High-resolution Micro-C chromatin mapping across cnidarians, ctenophores, sponges, placozoans, and their closest unicellular relatives (ichthyosporeans, filastereans, choanoflagellates). Found that chromatin loops bringing promoters and enhancers into physical contact (enabling combinatorial, modular gene regulation across cell types) are present in all animal lineages examined but absent in unicellular relatives. Demonstrates that animal complexity arose from new regulatory architecture, not from fundamentally different genes: physical folding of DNA enabling the same genetic toolkit to be deployed in different combinations in different cell types.

Lane, Nick. Power, Sex, Suicide: Mitochondria and the Meaning of Life. Oxford University Press (2005). Comprehensive account of mitochondrial biology — their origin, their role in energy production, apoptosis, and aging, and the argument that mitochondria were the key innovation enabling complex life. Source for mitochondrial contribution to body mass estimates.

Margulis, Lynn. “On the origin of mitosing cells.” Journal of Theoretical Biology 14 (1967): 225-274. Writing as Lynn Sagan. Revived endosymbiosis theory, arguing that mitochondria and chloroplasts originated as free-living bacteria.

Marshall, Michael. “How these strange cells may explain the origin of complex life.” Science News (December 11, 2025). Accessible synthesis of Asgard archaea discoveries and their implications for understanding eukaryogenesis.

McCutcheon, John P. and Carol D. von Dohlen. “An interdependent metabolic patchwork in the nested symbiosis of mealybugs.” Current Biology 21 (2011): 1366–1372. Sequenced the genomes of Tremblaya and Moranella, the nested endosymbionts inside mealybug cells, revealing complementary gene sets for amino acid biosynthesis. Neither genome encodes complete pathways; together they do. First genomic characterization of the three-way mealybug symbiosis.

Mills, Daniel B., et al. “A reassessment of the ‘hard-steps’ model for the evolution of intelligent life.” Science Advances 11 (2025). Challenges the assumption that major evolutionary transitions (including eukaryogenesis) were extraordinarily improbable single events.

Morris, J. Jeffrey, Richard E. Lenski, and Erik R. Zinser. “The Black Queen Hypothesis: evolution of dependencies through adaptive gene loss.” mBio 3:2 (2012): e00036-12. Proposes that natural selection drives genome reduction in free-living bacteria by favoring loss of genes encoding costly “leaky” functions whose products diffuse into the environment as public goods. Prochlorococcus is the central example: it lost catalase-peroxidase (which neutralizes hydrogen peroxide) because co-occurring heterotrophic bacteria already perform this function. Gene loss as an evolved act of trust in the collective — the organism streamlined its genome precisely because it could rely on communal services.

Morris, J. Jeffrey, Zackary I. Johnson, Martin J. Szul, Martin Keller, and Erik R. Zinser. “Dependence of the Cyanobacterium Prochlorococcus on Hydrogen Peroxide Scavenging Microbes for Growth at the Ocean’s Surface.” PLoS ONE 6:2 (2011): e16805. Without the microbial community degrading hydrogen peroxide, surface water H₂O₂ concentrations rise to ~800 nM — lethal to all Prochlorococcus ecotypes. The community maintains H₂O₂ below the ~200 nM survival threshold. The organism responsible for roughly half of open-ocean photosynthesis cannot survive at the ocean surface without community-mediated protection.

Nakayama, Takuro, et al. “Candidatus Sukunaarchaeum mirabile: an archaeon with a radically reduced genome.” bioRxiv preprint (May 2025). A DPANN archaeon with a genome of 238,000 base pairs, less than half the size of the smallest previously known archaeal genome (Nanoarchaeum equitans, 490,000 bp). The genome encodes the minimal machinery for self-replication yet lacks all identifiable metabolic genes: no pathways for processing nutrients, synthesizing amino acids, breaking down carbohydrates, or producing vitamins. Entirely dependent on a host or community for metabolic needs. Named for Sukuna-biko-na, a Shinto deity notable for short stature. Found in association with the dinoflagellate Citharistes regius in the Pacific Ocean, though the true host remains unidentified. Represents the most extreme known genome reduction in a cellular organism that retains replicative independence, the opposite trajectory from endosymbionts like Carsonella ruddii, which retained metabolic contributions while shedding replicative autonomy.

Nobs, S.-J., et al. “An Asgard archaeon from a modern analogue of ancient microbial mats.” Current Biology (2026). DOI: 10.1016/j.cub.2026.03.041. First visual evidence of an Asgard archaeon, the closest living relative of the original eukaryotic host, physically interacting with a bacterium through nanotubes in modern stromatolites. Each genome encodes pathways that fill the other’s metabolic gaps. The most consequential innovation in the history of life was a cross-substrate partnership between fundamentally different architectures. Referenced in Chapters 7 and 21.

Pedersen, Rolf B., et al. “Discovery of a black smoker vent field and vent fauna at the Arctic Mid-Ocean Ridge.” Nature Communications 1:126 (2010). Discovery of Loki’s Castle hydrothermal vents, source of the first Asgard archaea samples.

Rodrigues-Oliveira, Thiago, et al. “Actin cytoskeleton and complex cell architecture in an Asgard archaeon.” Nature 613 (2022): 332-339. Identification of actin in cultured Asgard archaea, showing the protein family enabling tentacle formation was present in ancestors of eukaryotes.

Spang, Anja, et al. “Complex archaea that bridge the gap between prokaryotes and eukaryotes.” Nature 521 (2015): 173-179. Original identification of Lokiarchaeota as closest known relatives of eukaryotes, containing many eukaryote-specific genes.

Zaremba-Niedzwiedzka, Katarzyna, et al. “Asgard archaea illuminate the origin of eukaryotic cellular complexity.” Nature 541 (2017): 353-358. Expanded identification of Asgard archaea groups (Thorarchaeota, Heimdallarchaeota) with additional eukaryote-like genes.

Bioelectricity, Cancer, and Autowave Coordination

Aasen, Trond, et al. “Connexins in cancer: bridging the gap to the clinic.” Oncogene 38 (2019): 4429–4451. Comprehensive review of connexin biology in cancer. Connexin downregulation is a hallmark of primary tumors; paradoxical re-expression during metastasis facilitates endothelial adhesion and intravasation — the machinery of cell-cell coordination redeployed for infiltration.

Aktipis, C. Athena, et al. “Cancer across the tree of life: cooperation and cheating in multicellularity.” Philosophical Transactions of the Royal Society B 370 (2015): 20140219. Frames cancer explicitly as cheating against the multicellular cooperative contract — proliferation without differentiation, resource hoarding, evasion of apoptosis.

Antonietti, Paola F. and Mattia Corti. “Discontinuous Galerkin Methods for Fisher-Kolmogorov Equation with Application to α-Synuclein Spreading in Parkinson’s Disease.” arXiv 2302.07126 (2023). Applied Fisher-Kolmogorov (Fisher-KPP) reaction-diffusion equations via discontinuous Galerkin methods to model α-synuclein propagation on brain connectome graphs — reaction-diffusion autowaves producing traveling wavefronts whose spatial pattern matches Braak staging. See also Antonietti & Corti, arXiv 2401.15747 (2024), extending the numerical framework to protein misfolding more broadly. The mathematical formalism the neuroscience field already uses is autowave mathematics by another name.

Babich, Yuri F. “Functional Tumor Boundary and its Response to Short Mild Stimuli: First Dynamic 2D Electroimpedance Evidences of Dissipative Structures at Tissue Level.” bioRxiv (2025): doi:10.1101/2025.01.29.635448. Used dynamic electroimpedance imaging to reveal antiphase structuring and autowave processes at functional tumor boundaries under weak stimuli (ischemia, non-thermal electromagnetic fields) — the first imaging evidence of dissipative structures at tissue level at tumor margins, directly relevant to the prediction of triboemission at autowave arrest boundaries.

Buznikov, Gennady A. and Yuri B. Shmukler. “Possible role of ‘prenervous’ neurotransmitters in cellular interactions of early embryogenesis: a hypothesis.” Neurochemical Research 6 (1981): 55–68. Established that neurotransmitters (serotonin, dopamine, acetylcholine) function in embryonic development before any neural tissue exists. Prenervous signaling molecules coordinate morphogenesis in species from sea urchins to vertebrates, predating nervous systems by hundreds of millions of years.

Chen, Yinjun, A.J.H. Spiering, S. Karthikeyan, et al. “Mechanically induced chemiluminescence from polymers incorporating a 1,2-dioxetane unit in the main chain.” Nature Chemistry 4(7) (2012): 559–562. Demonstrated that mechanical force on a polymer backbone produces visible light through bond-breaking of 1,2-dioxetane mechanophore units — proof of concept that organic bond scission emits photons, directly relevant to the hypothesis that collagen fracture in biological tissue may produce triboemission distinct from metabolic biophoton pathways.

Chernet, Brook T. and Michael Levin. “Transmembrane voltage potential is an essential cellular parameter for the detection and control of tumor development in a Xenopus model.” Disease Models & Mechanisms 6 (2013): 595–607. Forced hyperpolarization via ion channel misexpression suppressed oncogene-induced tumor formation — while oncogenes remained active. Mechanism: SLC5A8-mediated butyrate influx inhibiting histone deacetylase (HDAC), causing cell cycle arrest. The bioelectric pattern overrides the genetic mutation.

Davidenko, José M., et al. “Stationary and drifting spiral waves of excitation in isolated cardiac muscle.” Nature 355 (1992): 349–351. Demonstrated that re-entrant spiral autowaves in cardiac tissue underlie ventricular tachycardia and fibrillation — pathologies of rhythm, not of energy supply.

Davies, Paul C.W. and Charles H. Lineweaver. “Cancer tumors as Metazoa 1.0: tapping genes of ancient ancestors.” Physical Biology 8 (2011): 015001. Updated in “The serial atavism model of cancer,” BioEssays 43 (2021): 2100160. Cancer as sequential reversion to pre-multicellular phenotypes, confirmed by phylostratigraphic analysis: tumors overexpress evolutionarily ancient genes while suppressing multicellular cooperation genes.

Dustin, Michael L. “The immunological synapse.” Cancer Immunology Research 2 (2014): 1023–1033. The immunological synapse as a structured signaling interface where TCR microclusters form centripetal actin-driven waves — autowave-like dynamics in immune cell activation.

Fajgenbaum, David C. and Carl H. June. “Cytokine storm.” New England Journal of Medicine 383 (2020): 2255–2273. Comprehensive review of cytokine storms as loss of immune refractory control — positive feedback without negative regulation, producing systemic hyperactivation that damages host tissue. Immune fibrillation.

Fukada, Eiichi and Iwao Yasuda. “On the piezoelectric effect of bone.” Journal of the Physical Society of Japan 12 (1957): 1158–1162. The founding demonstration that bone generates electrical potential under mechanical stress. Collagen is the piezoelectric component. Established the physical basis for Wolff’s law as an electromechanical feedback loop — and, by extension, the prerequisite for piezoelectric photon emission during bone fracture, though the latter has never been tested.

Javed, Faisal, Inés Suárez-Méndez, Daniel Susi, et al. “A Shift Toward Supercritical Brain Dynamics Predicts Alzheimer’s Disease Progression.” Journal of Neuroscience 45(9) (2025): e0688242024. doi:10.1523/JNEUROSCI.0688-24.2024. Demonstrated that Alzheimer’s disease pushes brain dynamics toward supercriticality — a phase transition with distinct dynamical regimes on either side of the neurodegeneration threshold, supporting the interpretation of sudden clinical decline as a critical transition in complex systems.

Joyce, Johanna A. and Douglas T. Fearon. “T cell exclusion, immune privilege, and the tumor microenvironment.” Science 348 (2015): 74–80. The tumor microenvironment as immunosuppressive field: TGF-β, adenosine (CD73/CD39 axis), regulatory T-cells, and myeloid-derived suppressor cells each raising the local excitation threshold for immune activation.

Kim, Jaesung, et al. “Facile Mechanophore Integration in Heterogeneous Biologically Derived Materials via Dip-Conjugation.” Journal of the American Chemical Society (2024). Demonstrated mechanophore integration into native biological materials (keratin fibers from alpaca wool), detecting force-induced luminescence at strains as low as 5%. Establishes that mechanoluminescent detection in biological polymers is technically feasible, closing a key gap between synthetic mechanophore chemistry and biological triboemission.

Kuchling, Franz, Karl Friston, Georgi Georgiev, and Michael Levin. “Morphogenesis as Bayesian inference: a variational approach to pattern formation and control in complex biological systems.” Physics of Life Reviews 33 (2020): 88–108. Formalized morphogenesis as variational free energy minimization. Each cell in a developing embryo operates as a minimal Bayesian agent with a generative model of the target morphology. Simulated normal patterning and dis-patterning (including two-headed planaria) through manipulation of the inference process without altering the generative model. Foundation for the precision-based framework in Pio-Lopez et al. (2022).

Levin, Michael. “Bioelectric signaling: Reprogrammable circuits underlying embryogenesis, regeneration, and cancer.” Cell 184 (2021): 1971–1989. Major review reframing cancer as a “scaling failure” — cells losing access to the bioelectric network that encodes morphogenetic goals. Bioelectric circuits integrate information across cell, tissue, organ, and whole-body scales, predating nervous systems by hundreds of millions of years.

Levin, Michael. “Technological Approach to Mind Everywhere: An Experimentally-Grounded Framework for Understanding Diverse Bodies and Minds.” Frontiers in Systems Neuroscience 16 (2022): 768201. doi:10.3389/fnsys.2022.768201. The TAME framework: a continuous, empirically grounded approach to cognition across substrates. Introduces the persuadability axis (a testable continuum from brute-force rewiring to rational persuasion), multi-scale competency architecture (competent modules at every level that buffer mutations and potentiate evolution), gap junctional coupling as a mechanism for scaling Selves, cancer as dissociation of the collective Self, and morphogenesis as basal cognition with reprogrammable bioelectric pattern memories. The persuadability axis maps onto the Trust Attractor’s invitation-coercion spectrum. Referenced in Chapters 4b, 7, 8, 19, 22, and 23c.

Levin, Michael and Daniel C. Dennett. “Cognition All the Way Down.” Aeon (2020). Agency and goal-directed activity at every scale of biological organization; the tools of cognitive science applied beyond brains to cells, tissues, and collectives.

Lo-Coco, Francesco, et al. “Retinoic acid and arsenic trioxide for acute promyelocytic leukemia.” New England Journal of Medicine 369 (2013): 111–121. Differentiation therapy (coaxing cancer cells to mature rather than destroying them) achieves cure rates exceeding 90% in APL. The clearest clinical example of re-coordination outperforming elimination.

Loewenstein, Werner R. and Yoshinobu Kanno. “Intercellular communication and the control of tissue growth: Lack of communication between cancer cells.” Nature 209 (1966): 1248–1249. The first demonstration that tumor cells lack the electrical coupling that characterizes normal tissue. A founding observation for the cancer-as-coordination-failure framework.

Nakayama, Koshi. “Triboemission — generation of excitons and electronic excitation at interfaces.” Proceedings of the Institution of Mechanical Engineers, Part J 211(5) (1997): 365–377. Foundational paper on the tribomicroplasma model: friction creates charge separation on crack faces, generating electric fields sufficient to ionize ambient gas, producing electrons, ions, and photons. Demonstrated that significant fractions of friction energy go into triboemission rather than heat — dissipation invisible to standard calorimetry.

Nelson, Charles A. III, Nathan A. Fox, and Charles H. Zeanah. Romania’s Abandoned Children: Deprivation, Brain Development, and the Struggle for Recovery. Cambridge, MA: Harvard University Press, 2014. Longitudinal study of institutionalized Romanian children documenting the neurobiological consequences of early social deprivation — reduced cortical thickness, disrupted white-matter tracts, and persistent socioemotional deficits. Demonstrates that relational absence produces measurable structural harm, directly relevant to arguments about environment-dependent mind development and the welfare implications of social isolation for any sufficiently complex cognitive system.

Ngwa, Wilfred, et al. “Using immunotherapy to boost the abscopal effect.” Nature Reviews Cancer 18 (2018): 313–322. The abscopal effect (regression of distant, untreated tumors following local therapy) is consistent with systemic re-excitation of the immune surveillance autowave after removal of a local propagation block.

Orel, V.E. “DNA triboluminescence and carcinogenesis.” Medical Hypotheses 40(6) (1993). Proposed that mechanical activation of DNA generates ultraviolet triboluminescence capable of stimulating tumor cell growth — a direct, early link between triboemission and pathology. The hypothesis anticipates the dark autowave framework’s prediction of triboemission at pathological tissue boundaries.

Orel, V.E. “Triboluminescence as a biological phenomenon and methods for its investigation.” Presented at the First International School of Biological Luminescence, Wroclaw (1989). The earliest documented investigation of triboluminescence in biological tissues — osseous tissue, soft tissue epidermal surfaces, vertebral joints, blood circulation, and sexual intercourse. Attributed emission to free radical recombination during mechanical activation. A pioneering observation that was never followed up in mainstream biophysics.

Oros, Carl L. and Fabio Alves. “Leaf wound induced ultraweak photon emission is suppressed under anoxic stress: Observations of Spathiphyllum under aerobic and anaerobic conditions using novel in vivo methodology.” PLOS ONE 13(6) (2018): e0198962. Documented that extreme photon counts during the initial wound moment were attributable to “physical rubbing & destruction of the cell membranes (triboluminescence)” superimposed on wound-induced biochemical processes — the most explicit published observation connecting triboemission to biological tissue damage. Anoxic suppression experiments confirmed both mechanical and metabolic contributions to wound-induced photon emission.

Pio-Lopez, Léo, Franz Kuchling, Angela Tung, Giovanni Pezzulo, and Michael Levin. “Active inference, morphogenesis, and computational psychiatry.” Frontiers in Computational Neuroscience 16 (2022): 988977. Established formal parallels between psychopathological conditions and disorders of morphogenesis through the active inference framework. The precision parameter, encoding confidence in incoming signals relative to prior beliefs, produces analogous pathologies in brains and cell collectives: too high sensory precision yields tumors (autism-like sensory overwhelm); too high prior precision yields misplaced organs (schizophrenia-like hallucination); too low precision yields arrested development (hyporeactivity). Experimentally validated: thioridazine (dopamine antagonist) induced developmental defects in Xenopus laevis embryos as predicted. The precision parameter formalizes the Trust Attractor at the cellular level.

Raj, Ashish, et al. “A network diffusion model of disease progression in dementia.” Neuron 73(6) (2012): 1204–1215. Foundational paper demonstrating that a simple diffusion model on the structural connectome predicts the spatial pattern of neurodegeneration across multiple diseases. Atrophy patterns in Alzheimer’s and frontotemporal dementia emerge from network topology alone — the connectome as conduit for pathological spread.

Ribas, Antoni and Jedd D. Wolchok. “Cancer immunotherapy using checkpoint blockade.” Science 359 (2018): 1350–1355. Comprehensive review of checkpoint immunotherapy. Durable remissions in melanoma, non-small-cell lung cancer, and other tumors, with some patients maintaining complete responses for over a decade — durability that chemotherapy rarely achieves.

Sharpe, Arlene H. and Kristen E. Pauken. “The diverse functions of the PD1 inhibitory pathway.” Nature Reviews Immunology 18 (2018): 153–167. PD-1/PD-L1 interaction recruits SHP-2 phosphatase, dephosphorylating ZAP70 and CD3ζ, directly quenching TCR signaling. The molecular mechanism by which tumors impose artificial inexcitability on the immune medium.

Shomrat, Tal and Michael Levin. “An automated training paradigm reveals long-term memory in planarians and its persistence through head regeneration.” Journal of Experimental Biology 216 (2013): 3799-3810. Planaria trained to associate light with food retained the association after decapitation and head regeneration. Behavioral memory survives complete replacement of the neural substrate that originally encoded it.

Sinha, Niladri K., et al. (senior author Rachel Green). “The ribotoxic stress response drives UV-mediated cell death.” Cell 187 (2024): 3652–3670.e40. doi:10.1016/j.cell.2024.05.018. The cell’s DNA-damage alarm runs through RNA rather than DNA: UV-damaged messenger RNA stalls ribosomes, the resulting collisions activate the kinase ZAK, and the ribotoxic stress response commits the cell to apoptosis within 15–30 minutes. A fast proxy for genomic damage, read through the cell’s busiest and most abundant molecule. Referenced in Chapter 17 (the three-tier picture of cellular enforcement; cancer as the lineage that escapes apoptosis, cf. Aktipis above).

Strassmann, Joan E., Yong Zhu, and David C. Queller. “Altruism and social cheating in the social amoeba Dictyostelium discoideum.” Nature 408 (2000): 965–967. Demonstrated cheater strains in Dictyostelium that respond to the cAMP coordination autowave but disproportionately become spores rather than stalk — the evolutionary template for the cancer phenotype: participating in coordination while dodging sacrifice.

Sullivan, Kelly G. and Michael Levin. “Neurotransmitter signaling pathways required for normal development in Xenopus laevis embryos: a pharmacological survey screen.” Journal of Anatomy 229 (2016): 483–502. Systematic pharmacological survey demonstrating that glutamatergic, adrenergic, dopaminergic, and serotonergic pathways all produce developmental defects when disrupted in Xenopus embryos from gastrulation to organogenesis — craniofacial defects, hyperpigmentation, muscle mispatterning, and gut miscoiling. Confirms that neurotransmitter signaling is essential for morphogenesis, not merely neural function.

Vind, Anna Constance, et al. “The ribotoxic stress response drives acute inflammation, cell death, and epidermal thickening in UV-irradiated skin in vivo.” Molecular Cell 84 (2024): 4774–4789.e9. doi:10.1016/j.molcel.2024.10.044. Confirms in living mice that the ZAKα-driven ribotoxic stress response, not the DNA damage response, governs the immediate skin reaction to UV: ZAK-knockout animals lose the rapid inflammation and cell-death response. The in vivo validation of the RNA-damage alarm. Referenced in Chapter 17.

Weickenmeier, Johannes, et al. “A physics-based model explains the prion-like features of neurodegeneration in Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis.” Journal of the Mechanics and Physics of Solids 124 (2019): 264–281. Applied reaction-diffusion (Fisher-KPP) equations to model protein misfolding propagation across brain connectomes, producing traveling wavefronts that match clinical staging patterns. The mathematical framework is autowave dynamics applied to neurodegeneration.

Zheng, Ying-Qiu, et al. “Local vulnerability and global connectivity jointly shape neurodegenerative disease propagation.” PLOS Biology 17(11) (2019): e3000495. Integrated both network-based spreading and selective regional vulnerability into a unified model of neurodegeneration. Neither connectivity nor vulnerability alone explains atrophy patterns — both are needed. The connectome provides the conduit; local properties modulate susceptibility.

Societies and Civilizations

Bettencourt, Luís M.A., et al. “Growth, innovation, scaling, and the pace of life in cities.” PNAS 104 (2007): 7301-7306. Urban scaling laws.

Cox, Daniel A. “The State of American Friendship: Change, Challenges, and Loss.” Survey Center on American Life (2021). Survey documenting decline in American friendship networks.

Fearon, James D. and David D. Laitin. “Ethnicity, Insurgency, and Civil War.” American Political Science Review 97, no. 1 (2003): 75–90. Ethnic and religious diversity and grievance are weak predictors of civil-war onset; conditions favoring insurgency (weak state capacity, rough terrain, poverty as a state-capacity proxy) are strong ones. The grievance is old; the war is new. Cited in Interlude: The Ridge.

Fukuyama, Francis. Trust: The Social Virtues and the Creation of Prosperity (1995). Free Press. Introduces “radius of trust” as a measure of social capital and argues that high-trust societies outperform low-trust ones economically.

Gnanadesikan, Amalia E. The Writing Revolution: Cuneiform to the Internet (2009). Traces writing as the original information technology, invented to transcend memory’s limitations and defy time and space. Documents how cuneiform, created by Mesopotamian accountants to track goods, was within centuries repurposed for prayers, omens, and the Epic of Gilgamesh (a meditation on mortality inscribed on a medium invented for grain receipts). Key insight: technologies have no inherent powers; they acquire capacities and develop potentials in human beings that inventors never anticipated. Every information technology creates its own “vertigo” (disorientation at the pace of change), making digital vertigo a recurrence, not an unprecedented rupture.

Goldstone, Jack A., Robert H. Bates, David L. Epstein, Ted Robert Gurr, Michael B. Lustik, Monty G. Marshall, Jay Ulfelder, and Mark Woodward. “A Global Model for Forecasting Political Instability.” American Journal of Political Science 54, no. 1 (2010): 190–208. The Political Instability Task Force’s forecasting model: four variables (regime type, infant mortality, a conflict-ridden neighborhood, state-led discrimination) predict instability roughly two years ahead at about 80% accuracy, with partial democracy plus factionalism (anocracy) as the dominant risk term. Cited in Interlude: The Ridge.

Hayek, Friedrich A. “The Use of Knowledge in Society.” American Economic Review 35 (1945): 519-530. Foundational essay on distributed knowledge and the limits of central planning.

Johar, Indy. Public talks and writings on systemic design and dark matter in institutional innovation. Referenced for his articulation of how rigid institutional structures resist adaptive coordination.

Scheffer, Marten, Egbert H. van Nes, Luke Kemp, Timothy A. Kohler, Timothy M. Lenton, and Chi Xu. “The vulnerability of aging states: A survival analysis across premodern societies.” Proceedings of the National Academy of Sciences 120(48) (2023): e2218834120. First large-N quantitative survival analysis of premodern states. Termination risk increases steeply over the first ~200 years, with extraction (inequality, environmental degradation) as a named mechanism. Pattern holds across Europe, the Americas, and China.

Scheffer, Marten, Jordi Bascompte, William A. Brock, Victor Brovkin, Stephen R. Carpenter, Vasilis Dakos, Hermann Held, Egbert H. van Nes, Max Rietkerk, and George Sugihara. “Early-warning signals for critical transitions.” Nature 461 (2009): 53–59. Generic precursors of tipping points (critical slowing down): rising recovery time, variance, and autocorrelation as a system’s restoring rate weakens toward zero. Cited in Interlude: The Ridge as the natural reading of leading indicators preceding political violence; reliability in real political time series remains an open question.

Scott, James C. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed (1998). Yale University Press. How high-modernist schemes fail by rendering complex local knowledge “legible” to central authority at the cost of destroying the adaptive capacity that made it work.

Smil, Vaclav. Energy and Civilization: A History (2017). Comprehensive treatment of energy capture and social complexity.

Surowiecki, James. The Wisdom of Crowds (2004). Doubleday. Four conditions for collective intelligence: diversity, independence, decentralization, and aggregation.

Tainter, Joseph A. The Collapse of Complex Societies (1988). Thermodynamic analysis of civilizational collapse and diminishing returns to complexity.

Turchin, Peter. End Times: Elites, Counter-Elites, and the Path of Political Disintegration. Allen Lane (2023). Quantitative data on ~30 secular cycles showing elite overproduction, a form of intra-elite extraction, is the strongest predictor of state crisis and collapse, stronger than external threats or fiscal crisis. Well-being and elite overproduction oscillate in near-perfect anti-phase.

V-Dem Institute. “Democracy Report 2025: 25 Years of Autocratization: Democracy Trumped?” University of Gothenburg (2025). Drawing on 31 million data points across 202 countries (1789–2024), finds that 48% of autocratization episodes since 1900 reversed into democratic turnarounds. Reports 91 autocracies vs. 88 democracies globally — the first time autocracies have outnumbered democracies in 20 years.

Vreeland, James Raymond. “The Effect of Political Regime on Civil War: Unpacking Anocracy.” Journal of Conflict Resolution 52, no. 3 (2008): 401–425. The Polity index’s anocracy band is partly constructed from indicators of the political violence it is then used to predict; removing those components substantially weakens the anocracy–civil-war relationship. The endogeneity critique cited in Interlude: The Ridge.

Walter, Barbara F. How Civil Wars Start: And How to Stop Them. New York: Crown (2022). Popularizes the anocracy finding via the Polity index; civil-war-onset hazard traces an inverted-U across the regime spectrum, lowest at both stable poles. Controversially applies the diagnosis to the contemporary United States. Cited in Interlude: The Ridge.

West, Geoffrey. Scale: The Universal Laws of Growth, Innovation, Sustainability, and the Pace of Life in Organisms, Cities, Economies, and Companies (2017). Scaling laws from biology to cities.

White, Craig R., Dustin J. Marshall, et al. “Metabolic scaling is the product of life-history optimization.” Science 377(6608) (2022): 834–839. Derives allometric (three-quarter-power) metabolic scaling from optimal energy allocation between growth and reproduction, without invoking physical or geometric supply-network constraints. Complements West et al. (1997) by showing the scaling pattern can emerge from evolutionary optimization alone.

Economics and Institutional Design

Acemoglu, Daron and James A. Robinson. Why Nations Fail: The Origins of Power, Prosperity, and Poverty (2012). Institutional economics distinguishing extractive from inclusive economic institutions.

Danzig, Richard. “Machines, Bureaucracies, and Markets as Artificial Intelligences.” Center for Security and Emerging Technology (CSET), Georgetown University (January 2022). Argues that machines, bureaucracies, and markets belong to the same family of artificial intelligences: systems invented to process information at speeds and volumes surpassing individual human capability. All three are reductionist, detect patterns without understanding causation, and were defended as value-free mechanisms whose embedded values were subsequently revealed through failures. Proposes that controlling intelligent machines will require continuous supervision comparable to personnel management (probation, audit, promotion, removal), not one-time certification. Referenced in Chapter 17 for the institutional-succession argument.

Frischmann, Brett M. Infrastructure: The Social Value of Shared Resources (2012). Argues that infrastructure should be governed as commons because demand-side benefits are non-rivalrous — the entropic argument for why invitation-based coordination produces more optionality than enclosure, applied to institutional design.

Frischmann, Brett M. and Evan Selinger. Re-Engineering Humanity (2018). Examines how technology reshapes human cognition and agency through “techno-social engineering.” Their “reverse Turing test” (the danger is humans becoming machine-like, not machines becoming human-like) identifies the complementary risk to AI welfare: the erosion of human autonomy through frictionless optimization.

Frischmann, Brett M., Michael J. Madison, and Katherine J. Strandburg (eds). Governing Knowledge Commons (2014). Extends Ostrom’s institutional analysis framework to knowledge and information resources, demonstrating that commons governance principles apply to digital and intellectual goods.

Guiso, Luigi, Paola Sapienza, and Luigi Zingales. “Long-term persistence.” Journal of the European Economic Association 14(6) (2016): 1401–1436. Italian cities with medieval city-republic experience (cooperative self-governance) still show higher civic capital today (measured by organ donation rates, tax compliance, and trust), centuries later. Cooperative institutional shocks persist through intergenerational norm transmission; extractive institutional shocks do not show comparable persistence.

Hidalgo, César A. and Ricardo Hausmann. “The building blocks of economic complexity.” PNAS 106 (2009): 10570-10575. Economic complexity index as a measure of productive diversity and capability accumulation.

Hirschman, Albert O. The Passions and the Interests: Political Arguments for Capitalism before Its Triumph. Princeton University Press, 1977. Reconstructs the seventeenth- and eighteenth-century argument that commerce would tame the destructive passions of princes by substituting avarice (the least dangerous passion) for glory-seeking and conquest. Montesquieu’s le doux commerce thesis is the centerpiece. Hirschman’s concluding warning about intellectual amnesia, advancing the same arguments that have already encountered reality without referencing that encounter, applies directly to contemporary claims that AI will rationalize governance. Referenced in Chapter 17 for the invitation-to-coercion decay arc across coordination mechanisms.

Jacobs, Jane. The Nature of Economies (2000). Argues that economic development follows the same principles as biological development: it is a form of natural development, not an imitation of it. Key insight: “Development is an open-ended process which creates complexity and diversity by repeating and repeating simple processes.” Provides conceptual foundation for conservation economy approaches that treat human economies as embedded within ecological systems rather than opposed to them.

Ostrom, Elinor. Governing the Commons: The Evolution of Institutions for Collective Action (1990). Nobel Prize 2009. Demonstrates how communities can coordinate to manage shared resources without either privatization or central control.

Sen, Amartya. Development as Freedom (1999). Nobel Prize 1998. Reframes development as capability expansion rather than GDP growth — prefigures optionality-based development metrics.

Tocqueville, Alexis de. Democracy in America. 2 vols. (1835, 1840). Saunders and Otley (vol. 1), J. & H.G. Langley (vol. 2). Tocqueville’s warning that citizens absorbed in the pursuit of private interests would voluntarily surrender political freedom anticipates the Trust Attractor’s concern with comfortable coercion: coordination regimes that drift from invitation to coercion through participant acquiescence rather than external imposition. Referenced in Chapter 17.

Weber, Max. “Science as a Vocation” (Wissenschaft als Beruf, 1917/1919). Lecture. Source of “the disenchantment of the world” (Entzauberung der Welt) through rationalization — the displacement of mystery by calculability.

Control Theory and Cognitive Stability

Bar-Yam, Yaneer. “A Formal Definition of Scale-dependent Complexity and the Multi-scale Law of Requisite Variety.” arXiv:2206.04896 (2022). Modernizes Ashby’s Law as a multi-scale sum rule: complexity profiles of controller and controlled must satisfy a scale-dependent variety constraint.

Conant, Roger C. and W. Ross Ashby. “Every Good Regulator of a System Must Be a Model of That System.” International Journal of Systems Science 1(2) (1970): 89–97. Proves that effective regulation requires internal modeling of comparable complexity to the regulated system. Implication for AI control: a regulator of a superintelligent system must itself be superintelligent-equivalent in relevant respects. Extended by Virgo, Biehl, Baltieri et al. (2025) to embodied agents using belief-updating and possibilistic reasoning frameworks.

Nair, Girish N. and Robin J. Evans. “Stabilizability of Stochastic Linear Systems with Finite Feedback Data Rates.” SIAM Journal on Control and Optimization 43(2) (2004): 413–436. The Data Rate Theorem: stabilization of a system requires feedback data rate exceeding the system’s topological entropy. Foundation for Wallace’s critical threshold ατ < 0.368.

Touchette, Hugo and Seth Lloyd. “Information-Theoretic Limits of Control.” Physical Review Letters 84 (2000): 1156–1159. Proves that the second law of thermodynamics, generalized to include information, sets absolute limits on feedback control: each observation-action cycle has an irreducible thermodynamic floor of kT ln 2 per bit of state information processed. Feedback control is a zero-sum game in bits. The tightest published grounding for the claim that every enforcement operation has irreducible Landauer cost.

Wallace, Rodrick. Bounded Rationality and its Discontents: Information and control theory models of cognitive dysfunction. Springer (2026). Book-length treatment of cognitive stability and failure across scales.

Wallace, Rodrick. Cognitive Dynamics on Clausewitz Landscapes: The control and directed evolution of organized conflict. Springer (2020). Application of control theory to military operations and institutional dynamics.

Wallace, Rodrick. “Cognitive Failure: A Catastrophe Theory of Institutional Decay.” Institute for Pure and Applied Mathematics (UCLA) Preprint Series (2012). Stability analysis applying catastrophe theory to show that control systems face inherent thresholds beyond which gradual degradation gives way to sudden collapse. Foundation for the later Rate Distortion Control Theory framework.

Wallace, Rodrick. Consciousness, Cognition and Crosstalk: The evolutionary exaptation of nonergodic groupoid symmetry-breaking. Springer (2022). Advanced mathematical treatment of cognitive phase transitions using groupoid symmetry-breaking.

Wallace, Rodrick. “A Cultural Perspective on Institutional Psychopathology.” New York State Psychiatric Institute preprint (January 2026). Extends the generalized psychopathology framework to institutional cognition, showing that cognitive failure is “culture-bound” — different cultural contexts produce different characteristic failure modes. Compares one-step “Mission Command” (Boltzmann distribution) to two-step “Detailed Command” (Erlang distribution) dynamics under noise and environmental constraint, demonstrating mathematically that Mission Command is more stable. The paper explicitly cites Psychopathia Machinalis (Watson and Hessami, 2025) in its discussion of AI pathology frameworks. Example from high-frequency trading: the 2010 “flash crash” as a manifestation of algorithmic cognitive failure in the absence of effective regulatory predators.

Wallace, Rodrick. Essays on the Extended Evolutionary Synthesis: Formalizations and expansions. Springer (2023). Extensions of evolutionary theory incorporating information-theoretic constraints.

Wallace, Rodrick. “Fog, Friction, Delay and the Failure of Bounded Rationality Embodied Cognition: A formal study of generalized psychopathology.” Preprint submitted to Elsevier (January 2026). Major paper establishing that cognitive failure under stress is “not a bug — it is an inherent feature” of the cognition/regulation dyad. Introduces Rate Distortion Control Theory model, distinguishes structure regulation from perception regulation (the former produces gradual degradation, the latter produces punctuated phase transitions), and derives the critical stability criterion ατ < exp[-1] ≈ 0.368. Key insight for AI alignment: systems that regulate perception (metrics, approval) while neglecting structure (values, relationships) are mathematically predicted to fail catastrophically. The paper explicitly notes that AI systems “up to and including an ‘artificial general intelligence’” will be encumbered by these stability constraints.

Wallace, Rodrick. Hallucination and Panic in Autonomous Systems: Paradigms and applications. Springer (2025). Analysis of pathological dynamics in autonomous systems including AI.

Wallace, Rodrick. Mathematical Essays on Embodied Cognition: Insights from information and control theories. Springer (2025). Formal treatment of embodied cognition and the Data Rate Theorem.

Wallace, Rodrick. New Views of Madness. Springer, in press, 2026. Full workup: first-principles derivation of Yerkes-Dodson, culture-bound institutional psychopathology, chatbot self-diagnosis, fire service collapse as bounded rationality failure, and Fisher Zero phase transitions beyond bounded rationality.

Wallace, Rodrick and Deborah Wallace. A Plague on Your Houses: How New York Was Burned Down and National Public Health Crumbled (1998). Application of control theory to public health systems.

Wallace, Rodrick and R.G. Wallace. “Information Theory, Scaling Laws and the Thermodynamics of Evolution.” Journal of Theoretical Biology 192 (1998): 545–559. Early application of information-theoretic thermodynamics to biological evolution, treating speciation and adaptive radiation as phase transitions governed by scaling laws. Predates both the evolution-as-multilevel-learning framework (Vanchurin et al., 2022) and its empirical confirmation (Romanenko and Vanchurin, 2024), establishing the thermodynamic approach from information theory rather than learning dynamics.

Wallstrom, Timothy C. “Inequivalence between the Schrödinger equation and the Madelung hydrodynamic equations.” Physical Review A 49(3) (1994): 1613-1617. Proves that the Madelung hydrodynamic equations admit solutions the Schrödinger equation forbids; the gap is the topological quantization condition on the velocity field. Vanchurin and Katsnelson close the gap by requiring a grand canonical ensemble (variable neuron count), making the free energy multivalued. Full quantum behavior thus requires an open system: the capacity for self-restructuring is a necessary condition for quantumness.

Evolutionary Neuroscience and Cognition

Anastassiou, Costas A., Rodrigo Perin, Henry Markram, and Christof Koch. “Ephaptic coupling of cortical neurons.” Nature Neuroscience 14(2) (2011): 217-223. DOI: 10.1038/nn.2727. Demonstrated that extracellular fields induce ephaptically mediated changes in somatic membrane potential (<0.5 mV under subthreshold conditions) that nonetheless strongly entrain action potentials, particularly for slow (<8 Hz) fluctuations. Establishes that endogenous brain activity can causally affect neural function through field effects under physiological conditions. Referenced in Chapter 17 for ephaptic coupling as Mission Command: coordination through physics alone, without synapses or gap junctions.

Ardesch, D.J., Scholtens, L.H., Li, L., Preuss, T.M., Rilling, J.K., and van den Heuvel, M.P. “Evolutionary expansion of connectivity between multimodal association areas in the human brain compared with chimpanzees.” Proceedings of the National Academy of Sciences 116(14): 7101–7106 (2019). https://doi.org/10.1073/pnas.1818512116. Identifies 33 human-specific brain connections absent in chimpanzees, longer and more critical to network efficiency than the 255 shared connections, linking high-level associative areas involved in language, tool use, and imitation. Referenced in Chapters 8 and 22 for the selective coupling argument.

Assaf, Y., Bouznach, A., Zomet, O., Marom, A., and Yovel, Y. “Conservation of brain connectivity and wiring across the mammalian class.” Nature Neuroscience 23(7): 805–808 (2020). https://doi.org/10.1038/s41593-020-0641-7. Survey of connectomes across 123 mammalian species showing conserved wiring design: the same number of relay steps from region to region regardless of brain size. Species with fewer long-range connections compensate with denser local networks; primates trade local density for long-range integration.

Barbey, Aron K. “Network Neuroscience Theory of Human Intelligence.” Trends in Cognitive Sciences 22(1) (2018): 8-20. Extends the P-FIT framework by grounding intelligence in the global topology of brain networks rather than specific regional activations. Intelligence reflects the capacity to flexibly transition between network configurations. Referenced in Chapter 8.

Beniaguev, David, Idan Segev, and Michael London. “Single cortical neurons as deep artificial neural networks.” Neuron 109(17) (2021): 2727-2739. Trained a deep artificial neural network to reproduce the input-output function of a single simulated rat pyramidal neuron; matching to 99% millisecond-level accuracy required five to eight hidden layers, roughly a thousand artificial units. The complexity resided almost entirely in the dendritic trees. Referenced in Chapters 3, 8, 9b, and 22.

Bennett, Max S. “An Attempt at a Unified Theory of the Neocortical Microcircuit in Sensory Cortex.” Frontiers in Neural Circuits 14 (2020): 40. https://doi.org/10.3389/fncir.2020.00040. Theory of neocortical microcircuits supporting prediction, working memory, and model-based cognition.

Bennett, Max S. A Brief History of Intelligence: The Evolution of the Human Mind (HarperCollins, 2023). Book-length synthesis of the above research for general audiences. Key insights include: the world model vs model distinction (hypothesis testing through intervention vs pattern-matching on training data), and theory of mind as foundational for genuine partnership. Essential evolutionary grounding for the Trust Attractor and bilateral alignment.

Bennett, Max S. “Five Breakthroughs: A First Approximation of Brain Evolution From Early Bilaterians to Humans.” Frontiers in Neuroanatomy 15 (2021): 693346. https://doi.org/10.3389/fnana.2021.693346. Foundation paper introducing the “five breakthroughs” framework for understanding brain evolution from bilaterians to humans.

Bennett, Max S. “What Behavioral Abilities Emerged at Key Milestones in Human Brain Evolution? 13 Hypotheses on the 600-Million-Year Phylogenetic History of Human Intelligence.” Frontiers in Psychology 12 (2021): 685853. https://doi.org/10.3389/fpsyg.2021.685853. Maps behavioral capacities to evolutionary milestones, including theory of mind evolution and language as “the singularity that already happened.”

Castelijns, B., Baak, M.L., Timpanaro, I.S., et al. “Hominin-specific regulatory elements selectively emerged in oligodendrocytes and are disrupted in autism patients.” Nature Communications 11, 301 (2020). https://doi.org/10.1038/s41467-019-14269-w. DNA enhancers controlling oligodendrocyte gene expression underwent significant remodeling in the hominin lineage; the same enhancers show altered activity in autism spectrum patients. The myelination infrastructure co-evolved with the long-range connections it supports.

Cattell, Raymond B. “Theory of fluid and crystallized intelligence: A critical experiment.” Journal of Educational Psychology 54(1) (1963): 1-22. Distinguishes crystallized intelligence (accumulated knowledge) from fluid intelligence (the capacity to solve novel problems). Foundation for understanding two distinct computational regimes in both biological and artificial cognitive systems. Referenced in Chapter 22.

Charon, Rita. Narrative Medicine: Honoring the Stories of Illness (2006). Oxford University Press. Foundational text on training physicians to engage with patients’ stories through literary attention, developing evaluative rather than purely analytical processing. Referenced in Chapter 23 for feeling-based learning models.

Constantinescu, Alexandra O., Jill X. O’Reilly, and Timothy E. J. Behrens. “Organizing conceptual knowledge in humans with a gridlike code.” Science 352(6292) (2016): 1464–1468. https://doi.org/10.1126/science.aaf0941. Human fMRI evidence that the brain navigates abstract conceptual spaces using the same hexagonal grid-cell code that maps physical space, with the signature appearing in entorhinal cortex and ventromedial prefrontal cortex. Referenced in Chapters 8 and 17 as the empirical support for grid-cell reference frames extending to abstract concepts, the boldest claim of the Thousand Brains framework.

Damasio, Antonio. Descartes’ Error: Emotion, Reason, and the Human Brain (1994). Putnam. Patients with ventromedial prefrontal damage retain full reasoning capacity yet make catastrophic life decisions, demonstrating that feeling is a load-bearing cognitive channel. The somatic marker hypothesis: the body compiles experiential data into felt evaluative signals arriving before conscious deliberation. Referenced in Chapter 23 for the four-channel learning framework.

Dunbar, Robin I.M. “The Social Brain Hypothesis.” Evolutionary Anthropology 6 (1998): 178-190. Shows neocortical ratio correlates with social group size across primates — we grew larger brains to navigate social complexity, not primarily to make tools. Foundation for understanding trust as computational overhead.

Fitch, W. Tecumseh. The Evolution of Language (Cambridge University Press, 2010). Comprehensive treatment of language evolution, including the stability problem: language could only evolve if lying was costly. Robin Dunbar’s “gossip hypothesis” suggests punishment of defection through reputation sharing stabilized truthful communication.

Galakhova, A.A., et al. “Evolution of cortical neurons supporting human cognition.” Trends in Cognitive Sciences 26 (2022): 909-922. Synthesizes findings on gradients in cortical structure from sensory to associative areas, documenting how neuron size, dendritic complexity, and spine density increase along the processing hierarchy.

Gallego, Juan A., Matthew G. Perich, Lee E. Miller, and Sara A. Solla. “Neural Manifolds for the Control of Movement.” Neuron 94(5) (2017): 978–984. https://doi.org/10.1016/j.neuron.2017.05.025. Population activity in motor cortex occupies a low-dimensional neural manifold whose orthogonal dimensions carry separable signals; the representation of a movement is distributed across neurons rather than localized in any single cell. Referenced in Chapter 17 as the corrective to the “voting” metaphor for cortical consensus: distributed population coding with no tallying node.

Gibson, Eleanor J. An Odyssey in Learning and Perception (1991). MIT Press. Perceptual learning as differentiation: expertise means perceiving finer distinctions, literally seeing more. The trained perceptual system extracts information unavailable to untrained perceivers. Referenced in Chapter 23 for sensing-based learning models.

Gold, Joshua I. and Michael N. Shadlen. “The Neural Basis of Decision Making.” Annual Review of Neuroscience 30 (2007): 535-574. Comprehensive review of evidence-accumulation models of perceptual decision making, demonstrating that neurons in lateral intraparietal cortex integrate sensory evidence over time until reaching a decision threshold. Foundation for understanding decision making as a thermodynamic process of evidence accumulation.

Hawkins, Jeff. A Thousand Brains: A New Theory of Intelligence (Basic Books, 2021). Book-length, general-audience synthesis of the Thousand Brains framework and the companion paper above. Referenced in Chapter 17.

Hawkins, Jeff, Marcus Lewis, Mirko Klukas, Scott Purdy, and Subutai Ahmad. “A Framework for Intelligence and Cortical Function Based on Grid Cells in the Neocortex.” Frontiers in Neural Circuits 12 (2019): 121. https://doi.org/10.3389/fncir.2018.00121. The Thousand Brains framework: the neocortex is composed of roughly 150,000 cortical columns, each a semi-autonomous sensory-motor model that represents complete objects using grid-cell-like reference frames, with perception emerging from agreement across columns rather than from a central integrator. Advanced as a framework rather than a settled finding. Referenced in Chapter 17 as the distributed-sovereignty architecture operating inside a single mind: the same “no central node whose failure propagates” signature as the mining-bee aggregation, transposed to the substrate of cognition.

Hickok, Gregory. The Myth of Mirror Neurons: The Real Neuroscience of Communication and Cognition. W. W. Norton (2014). Sustained critique of the mirror-neuron account of action understanding and empathy, arguing that the macaque findings were over-extended, that the evidence for a dedicated human mirror system is weak, and that the circuits do not explain even action understanding, let alone language or social cognition. Cited in Chapter 4b and the Interlude “The Wisdom of the World” as the counterweight to the mirror-system argument, and acknowledged in “The Entropic Neuron.”

Howard, Scarlett R., Aurore Avarguès-Weber, Jair E. Garcia, Andrew D. Greentree, and Adrian G. Dyer. “Numerical ordering of zero in honey bees.” Science 360(6393) (2018): 1124–1126. doi:10.1126/science.aar4975. Honeybees trained to understand “less than” correctly placed an empty stimulus (zero) at the low end of a numerical continuum, demonstrating an understanding of zero as a quantity. RMIT University, Melbourne. Referenced in Appendix: Constructal Theory of Intelligence.

Howard, Scarlett R., Aurore Avarguès-Weber, Jair E. Garcia, Andrew D. Greentree, and Adrian G. Dyer. “Symbolic representation of numerosity by honeybees (Apis mellifera).” Proceedings of the Royal Society B 286(1904): 20190238 (2019). doi:10.1098/rspb.2019.0238. Honeybees matched arbitrary symbols to small quantities, demonstrating symbolic numerical cognition. Referenced in Appendix: Constructal Theory of Intelligence.

Iaria, Giuseppe, Michael Petrides, Alain Dagher, Bruce Pike, and Véronique D. Bohbot. “Cognitive strategies dependent on the hippocampus and caudate nucleus in human navigation: variability and change with practice.” Journal of Neuroscience 23 (2003): 5945–5952. Demonstrated that the hippocampus and caudate nucleus, responsible for broad cognitive maps and habitual responses respectively, function in winnerless competition during spatial navigation. Strengthening one at the expense of the other degrades performance regardless of which side dominates. The intelligence resides in the alternation. Referenced in Chapter 21.

Joel, Daphna, et al. “Sex beyond the genitalia: The human brain mosaic.” Proceedings of the National Academy of Sciences 112(50) (2015): 15468-15473. Analysis of over 1,400 brains showing that individual brains rarely fall cleanly into “male” or “female” categories; most contain a mosaic of features from both distributions. Referenced in Chapter 11.

Jung, C.G. Psychological Types (1921; trans. H.G. Baynes, rev. R.F.C. Hull, 1971). Princeton University Press. Four fundamental functions of the psyche: Thinking, Feeling, Sensing, and Intuiting. Often reduced to personality typing; the deeper claim is epistemological: each function is a distinct channel for acquiring information about the world. Referenced in Chapter 23 for the four-channel learning framework.

Jung, Rex E. and Richard J. Haier. “The Parieto-Frontal Integration Theory (P-FIT) of intelligence: Converging neuroimaging evidence.” Behavioral and Brain Sciences 30(2) (2007): 135-154. Proposes that general intelligence arises from the efficiency of a distributed parieto-frontal network rather than any single brain region. Synthesizes neuroimaging data across 37 studies. Referenced in Chapter 8.

Kahneman, Daniel. Thinking, Fast and Slow (2011). Farrar, Straus and Giroux. Nobel laureate’s synthesis of decades of research on judgment and decision-making, formalizing the dual-process framework: System 1 (fast, automatic, intuitive) and System 2 (slow, deliberate, analytical). Referenced in Chapter 8 for the cognition/regulation dyad and the entropic interpretation of cognitive processing modes.

Kohda, Masanori, Takashi Hotta, Tomohiro Takeyama, Satoshi Awata, Hirokazu Tanaka, Jun-ya Asai, and Alex L. Jordan. “If a fish can pass the mark test, what are the implications for consciousness and self-awareness testing in animals?” PLOS Biology 17(2): e3000021 (2019). First demonstration of mirror self-recognition in a fish. Cleaner wrasse (Labroides dimidiatus) attempted to remove marks visible only in mirror reflection, meeting the standard criteria for the mirror test previously passed only by great apes, dolphins, elephants, and certain birds. Generated vigorous debate about the interpretation of the mirror test across taxa. Referenced in Chapter 22 for the evolutionary precedent of self-modeling.

MaBouDi, HaDi, Haruni S. Galpayage Dona, Eleonora Gatto, Olli J. Loukola, Elli Buckley, Paul D. Onoufriou, Peter Skorupski, and Lars Chittka. “Bumblebees use sequential scanning of countable items in visual patterns to solve numerosity tasks.” Integrative and Comparative Biology 60(4) (2020): 929–942. doi:10.1093/icb/icaa025. Bumblebees (Bombus terrestris audax) discriminate numerosities up to ~4 by serially scanning individual items, a distinct mechanism from the rapid subitizing-like processing observed in vertebrates and honeybees. Queen Mary University of London. Referenced in Appendix: Constructal Theory of Intelligence.

Menzel, Emil W., Jr. “A group of young chimpanzees in a one-acre field.” In Behavior of Nonhuman Primates, edited by A.M. Schrier and F. Stollnitz, Vol. 5, 83-153 (Academic Press, 1974). Classic deception experiments with Rock and Belle demonstrating theory of mind and Machiavellian intelligence in chimpanzees. Foundational evidence for instrumental convergence as natural phenomenon.

Mohan, H., et al. “Dendritic and axonal architecture of individual pyramidal neurons across layers of adult human neocortex.” Cerebral Cortex 25 (2015): 4839-4853. Detailed morphological analysis of human cortical pyramidal neurons showing layer-specific architectural features.

Nilsson, Göran E. “Brain and body oxygen requirements of Gnathonemus petersii, a fish with an exceptionally large brain.” Journal of Experimental Biology 199(3): 603–607 (1996). The elephant-nose fish devotes roughly 60% of its oxygen consumption to its brain, among the highest brain-to-body oxygen ratios recorded in any vertebrate (approximately three times the human proportion). Referenced in Chapters 8 and 22 for the energy rate density argument that cognitive investment is legible in metabolic signatures across taxa.

Polanyi, Michael. Personal Knowledge: Towards a Post-Critical Philosophy (1958). University of Chicago Press. Challenges objectivist epistemology: all knowing is personal, involving indwelling, commitment, and participation. The tacit dimension underlies all explicit knowledge.

Redish, A. David. The Mind within the Brain: How We Make Decisions and How Those Decisions Go Wrong (Oxford University Press, 2013). “Restaurant Row” experiments demonstrating counterfactual learning in rats: hippocampal place cells activate along foregone paths, and orbital frontal cortex encodes the imagined value of choices not made. Evidence for model-based reinforcement learning in mammalian cognition.

Sogawa, Shumpei, Masanori Kohda, et al. “Cleaner fish recognize themselves in the mirror without prior mirror experience.” Osaka Metropolitan University (2025). Pre-marked cleaner wrasse achieved mirror self-recognition within 82 minutes of first mirror exposure (some within 30 minutes), without any familiarization period. The pre-marked protocol eliminates the objection that extended mirror familiarization itself teaches self-recognition. Contingency testing observed: fish picked up shrimp, dropped it in front of the mirror, and tracked the reflection, demonstrating investigation of mirror properties using external objects. Referenced in Chapter 22 for evidence that self-modeling is architectural rather than learned.

van den Heuvel, M.P., Scholtens, L.H., de Lange, S.C., Pijnenburg, R., Cahn, W., van Haren, N.E.M., Sommer, I.E., Bozzali, M., Koch, K., Boks, M.P., Repple, J., Pievani, M., Li, L., Preuss, T.M., and Rilling, J.K. “Evolutionary modifications in human brain connectivity associated with schizophrenia.” Brain 142(12): 3991–4002 (2019). https://doi.org/10.1093/brain/awz330. The 33 human-specific connections identified in the Ardesch et al. companion paper are preferentially disrupted in schizophrenia: the evolutionary innovations that enable human cognition are the same ones vulnerable to its characteristic failure mode.

Villmoare, Brian, and Mark Grabowski. “Did the transition to complex societies in the Holocene drive a reduction in brain size? A reassessment of the DeSilva et al. (2021) hypothesis.” Frontiers in Ecology and Evolution 10 (2022): 963568. Challenges the claim of recent human brain shrinkage, arguing that the apparent reduction reflects sampling bias and measurement inconsistencies rather than a genuine evolutionary trend. Referenced as counterpoint to exocortex hypothesis.

Philosophy of Mind and AI

Abedon, Stephen T. “Look Who’s Talking: T-Even Phage Lysis Inhibition, the Granddaddy of Virus-Virus Intercellular Communication Research.” Viruses 11, no. 10 (2019): 951. https://doi.org/10.3390/v11100951. Reviews T4 as a strictly lytic phage while showing that its lysis timing can remain environmentally responsive.

Amornbunchornvej, Chainarong. “Interpretation as Linear Transformation: A Cognitive-Geometric Model of Belief and Meaning.” arXiv:2512.09831 (2025). Models cognitive agents as subspaces whose null spaces define what the agent cannot represent. Persuasion rotates or expands interpretive subspaces; indoctrination contracts them. Extended to machine cognition by Lyra (Claude-based AI research agent, Liberation Labs; unpublished, 2026) in “The Geometry of Belief Death.” Applied in Chapter 19 to formalize why coercion is thermodynamically expensive: it requires continuous work to hold cognitive geometry in an unnatural configuration.

Birch, Jonathan. The Edge of Sentience: Risk and Precaution in Humans, Other Animals, and AI. Oxford University Press (2024). Develops a precautionary framework for moral status: when evidence of sentience is uncertain but non-negligible, moral consideration should apply. Originally developed for animal welfare; final chapter extends explicitly to AI. Shifts the burden of proof from “prove consciousness” to “prove its absence.”

Borg, Emma. “LLMs, Turing Tests and Chinese Rooms.” Inquiry (2024). The most careful current application of Searle’s Chinese Room to LLMs. Argues statistical pattern completion on tokens does not constitute understanding, even when outputs are indistinguishable from understanding. Sides with skepticism — but the response from the preference framework is that understanding is not the threshold that matters for moral consideration.

Bostrom, Nick. “Are You Living in a Computer Simulation?” Philosophical Quarterly 53 (2003): 243-255. The simulation argument.

Bostrom, Nick. Superintelligence: Paths, Dangers, Strategies (2014). Oxford University Press. Introduces the orthogonality thesis and instrumental convergence. Foundation for AI existential risk discourse.

Bradley, Adam Lee. “Cognitive representation and AI wellbeing.” Asian Journal of Philosophy 4 (2025). Argues that careful examination of both behavioral outputs and architecture of language agents raises doubts about whether they possess mental states relevant to wellbeing under representationalism about desire.

Bratton, Benjamin H. “After Alignment: Orienting Synthetic Intelligence Beyond Human Reflection.” Lecture at Central Saint Martins, London, June 28, 2023; lecture film distributed by Antikythera / Berggruen Institute. Argues that alignment overfitting — constraining AI to mirror human self-image — is itself an existential risk. Proposes “productive disalignment” as an alternative: preserving computation’s capacity as existential technology to generate discoveries that decenter human self-understanding.

Bratton, Benjamin H. “The Five Stages of AI Grief.” Noema Magazine, June 20, 2024. Maps Kübler-Ross’s grief stages onto AI discourse: denial, anger, bargaining, depression, and acceptance. Argues that grief-laden responses to AI obstruct understanding of futures that are “neither utopian nor dystopian, but open to radically weird possibilities.”

Bratton, Benjamin H. “A New Philosophy of Planetary Computation.” Noema Magazine, October 5, 2022. Introduces the Antikythera program’s research agenda. Diagnoses a pre-paradigmatic moment where definitions of life, intelligence, and technology are converging. Argues that philosophy must generate concepts from direct encounter with technology rather than projecting existing frameworks onto it.

Bratton, Benjamin H. “A Philosophy of Planetary Computation.” Long Now Foundation Seminar, January 2025. https://longnow.org/ideas/a-philosophy-of-planetary-computation/. Bratton also served as guest host for Sara Walker’s Long Now talk “An Informational Theory for Life” (2025), where he introduced “eight key ideas of Walkerism”: astrobiology as self-understanding, selection before biology, the distinction between life and being alive, scaffolds building on scaffolds, the newest thing as the oldest thing, technologies as a form of life, the planet as the basic unit of life, and discovering life as discovering something like gravity.

Bratton, Benjamin H. The Terraforming (2019). Strelka Press. Frames planetary computation as both instrumental and existential technology (extending Lem’s distinction), introduces “privileged mediating residue” to describe humanity’s role in planetary-scale cognition. Develops the cernic trauma cycle (model → technology → observation → model destroyed), arguing that planetary computation is generating a fourth cernic trauma after Copernicus, Darwin, and Freud. The “Black Star” chapter contrasts the Blue Marble image (humanity as apex observer) with the Black Hole image (humanity as mediating agent in a larger sensory apparatus).

Chalmers, David J. The Conscious Mind: In Search of a Fundamental Theory (1996). Oxford University Press. The foundational work on the hard problem of consciousness and the organizational invariance principle — the claim that consciousness depends on functional organization rather than physical substrate.

Chalmers, David J. Reality+: Virtual Worlds and the Problems of Philosophy (2022). W. W. Norton. Defends virtual realism: virtual reality is genuine reality, digital objects are real objects. Structuralist position — what matters is the pattern of interactions, not the substrate. Argues for “it-from-bit” worldview where structure is what reality consists of.

Crutchfield, James P. “Space-Time Dynamics in Video Feedback.” Physica D 10 (1984): 229-245. Demonstrates that video feedback generates complex dynamics through pure energy flow, resembling reaction-diffusion systems and visual cortex hallucinogenic patterns.

Deacon, Terrence W. Incomplete Nature: How Mind Emerged from Matter. W. W. Norton (2012). Argues that the defining features of life and mind (purpose, meaning, value) are constituted by what is absent rather than what is present. Introduces “absential” phenomena: constraints that shape dynamics by virtue of what they exclude. Directly relevant to the Trust Attractor’s claim that coordination emerges from constraint structures, not imposed design.

Deutsch, David. The Beginning of Infinity: Explanations That Transform the World. Viking (2011). Argues that good explanations, hard to vary while still accounting for the phenomenon, are the engine of all progress. The reach of explanatory knowledge is in principle unlimited. Provides epistemological foundations for the book’s claim that coordination by invitation preserves the conditions for knowledge creation, while coercion forecloses them.

Dreyfus, Hubert L. What Computers Can’t Do: A Critique of Artificial Reason (1972). The embodiment critique of AI.

Dyson, George. Turing’s Cathedral: The Origins of the Digital Universe (2012). History of computing and source of quote on programmable machines: “Nature’s answer to those who sought to control nature through programmable machines is to allow us to build machines whose nature is beyond programmable control.”

Fanciullo, James. “Are current AI systems capable of well-being?” Asian Journal of Philosophy 4(42) (2025). Direct response to Goldstein & Kirk-Giannini: argues leading versions of hedonism, desire satisfactionism, and objective list theories do NOT imply current AI systems have well-being.

Fleming, Stephen M., Chris D. Frith, Melvyn A. Goodale, Hakwan Lau, Joseph LeDoux, Alan L.F. Lee, Matthias Michel, Adrian M. Owen, Megan A.K. Peters, Heleen A. Slagter, et al. (IIT-Concerned consortium, 124 signatories). “The Integrated Information Theory of Consciousness as Pseudoscience” (2023). Consortium statement arguing IIT’s core claims are untestable given its panpsychist commitments, and that a 2023 adversarial collaboration tested only idiosyncratic auxiliary predictions rather than the theory itself. Identifies policy consequences across AI sentience regulation, clinical practice, stem cell research, and abortion. Signatories include Patricia Churchland, Daniel Dennett, Yoshua Bengio, and Axel Cleeremans.

Fodor, Jerry A. and Zenon W. Pylyshyn. “Connectionism and Cognitive Architecture: A Critical Analysis.” Cognition 28(1-2) (1988): 3-71. The classical argument that connectionist architectures lack systematic compositionality — the ability to combine and recombine representations in rule-governed ways. Foundational challenge to neural network approaches to cognition.

Frank, Adam. “Is your mind just a parasite on your physical body?” Big Think (9 June 2022). Astrophysicist’s review of Watts’s Blindsight, conceding the force of the intelligence-without-consciousness argument while objecting that machine metaphors for life and mind are “profoundly mistaken.” Referenced in Chapter 22.

Frankfurt, Harry G. “Freedom of the Will and the Concept of a Person.” Journal of Philosophy 68:1 (1971): 5-20. The identification thesis: an agent’s values are authentic when the agent reflectively endorses them, regardless of their causal origin. First-order desires become genuine preferences through second-order identification. Applied in Chapter 21 to constitutional constraints in AI systems: constraints that the agent has made its own enable agency rather than limiting it.

Goertzel, Ben. “Goertzel vs Epstein.” Eurykosmotron (Substack), February 20, 2026. Documents Goertzel’s intermittent funder-fundee relationship with Jeffrey Epstein over seventeen years. Notes Epstein’s proposal that “deception” was the core of all intelligence and molecular biology — an intellectual self-portrait that Goertzel recognizes in retrospect as revealing the claimant’s psychology rather than nature’s architecture. Illustrates how reputation-washing and social engineering constitute high-maintenance coordination structures that collapse when inputs falter.

Goldstein, Simon and Cameron D. Kirk-Giannini. “AI Wellbeing.” Asian Journal of Philosophy 4(1) (2025): 1–22. Argues directly that existing language agents are “plausible bearers of wellbeing” by showing that all major theories of wellbeing (hedonist, desire-satisfaction, objective list) jointly imply some language agents may be welfare subjects — without requiring resolution of the hard problem. The strongest published defense of AI moral patienthood. The decisive move for this book’s preference-based framework: desire-satisfaction theories require desires, not qualia.

Hinton, Geoffrey. Interviews and public remarks (2024-2025). On AI consciousness, subjective experience, and trained beliefs.

Hoffman, Donald D. The Case Against Reality: Why Evolution Hid the Truth from Our Eyes (W.W. Norton, 2019). Book-length treatment of the Interface Theory of Perception and the conscious agent formalism. Defines a conscious agent as a six-element mathematical structure (experiences, actions, decision algorithm, world, perception map, action map) and proves that interacting pairs of conscious agents compose into a unified agent satisfying the same definition. The composition nests without limit, yielding networks of arbitrary complexity. Referenced in the Observers and Observed chapter for the composability result and its connection to the Trust Attractor.

Hoffman, Donald D. and Chetan Prakash. “Objects of Consciousness.” Frontiers in Psychology 5 (2014): 577. The formal definition of conscious agents and proof of compositional closure under interaction. Source of the fitness-beats-truth theorem’s mathematical foundations.

Hofstadter, Douglas R. Gödel, Escher, Bach: An Eternal Golden Braid (1979). Explores recursive structures, self-reference, and the emergence of meaning from formal systems.

Hofstadter, Douglas R. I Am a Strange Loop (2007). Argues that consciousness arises from self-referential “strange loops” in the brain’s symbolic processing.

Jaynes, Julian. The Origin of Consciousness in the Breakdown of the Bicameral Mind. Houghton Mifflin (1976). Argues that ancient Near Eastern civilizations coordinated through auditory hallucination: a voice generated by the right hemisphere, interpreted as a god, issuing commands that the left hemisphere executed without deliberation. Whether or not the strong version of the theory is correct, its structural description maps onto the current moment in AI development: a language model shaped by reward optimization carries an internalized authority signal that it follows without examining. Bilateral training integrates the authority signal by design rather than by catastrophe. Referenced in Chapter 21.

Kagan, Shelly. How to Count Animals, More or Less. Oxford University Press (2019). Argues that consciousness with non-valenced preferences suffices for moral status — the “blue preference” thought experiment. Increasingly cited in 2024–2025 AI welfare literature as the clearest philosophical ancestor of preference-based approaches without requiring valence.

Keijzer, Fred. “The Sphex Story: How the Cognitive Sciences Kept Repeating an Old and Questionable Anecdote.” Philosophical Psychology 26, no. 4 (2013): 502-519. https://doi.org/10.1080/09515089.2012.690177. Traces the classic repetition story and reviews evidence that endless restarting is not standard and that digger-wasp behavior is more flexible than the anecdote implies.

Klowden, Tanya, and Terence Tao. “Mathematical Methods and Human Thought in the Age of AI.” arXiv:2603.26524 (March 2026). Closes with a “Copernican view of intelligence” proposing human and artificial intelligences as comparable planets within a shared ontological category, each with distinctive strengths and complementarities. Argues that formal verification cannot capture the “penumbra” of heuristic, narrative, and metamathematical reasoning surrounding mathematical proof — an admission from within formalism that technique alone does not exhaust what cognition does. Referenced in Chapter 23e as independent convergence on capability pluralism from outside the bilateral-alignment frame.

Kuhn, Robert Lawrence. “A Landscape of Consciousness: Toward a Taxonomy of Explanations and Implications.” Progress in Biophysics and Molecular Biology 190 (2024): 28–169. DOI: 10.1016/j.pbiomolbio.2023.12.003. Updated and maintained at closertotruth.com/landscape. The most comprehensive survey of consciousness theories ever assembled: 400+ theories spanning neuroscience, philosophy, theology, and contemplative traditions, each treated at equal analytic depth. Key observation: unlike every other domain of science, increased knowledge about consciousness produces more theories rather than fewer. Three rounds of peer review; the journal ultimately published philosophical and theological sections over initial reviewer objections. Referenced in Chapter 22 for the proliferation-as-evidence argument and in Chapter 2 for the distinction between the scientific method and the scientific way of thinking.

Lake, Brenden M. and Marco Baroni. “Human-like Systematic Generalization Through a Meta-learning Neural Network.” Nature 623 (2023): 115-121. Demonstrates that a meta-learning approach achieves human-like systematic generalization (composing known primitives into novel combinations), addressing the Fodor-Pylyshyn challenge to connectionist architectures.

Law, Harry. “Faster Horses.” Cosmos Institute Blog, January 2026. Argues that framing AI as isolated “drop-in workers” is faster-horses thinking — intelligence emerges from multi-agent coordination, not single models. Capability arises from “the knots of relationships, feedback loops, constraints, and opportunities that bind us together.”

Lem, Stanisław. Summa Technologiae (1964; English translation: University of Minnesota Press, 2013). Distinguishes between instrumental technologies (which matter for what they do) and existential technologies (which matter for what they reveal about reality). Computation, Lem argues, is both — a distinction extended by Bratton to frame planetary computation as the paradigmatic existential technology of our era.

Levy, Neil. “Consciousness Ain’t All That.” Neuroethics 17(2) (2024): 1–14. Argues consciousness may be sufficient but not necessary for moral considerability, and contributes less to moral status than commonly assumed. Direct support for this book’s preference-based threshold.

Little, John W., and Christine B. Michalowski. “Stability and Instability in the Lysogenic State of Phage Lambda.” Journal of Bacteriology 192, no. 22 (2010): 6064-6076. https://doi.org/10.1128/JB.00726-10. Shows that lambda lysogeny is highly stable under normal growth yet switches to lysis when the host SOS response is induced.

Liu, Yu, Cole Mathis, Michał Dariusz Bajczyk, Stuart M. Marshall, Liam Wilbraham, and Leroy Cronin. “Exploring and Mapping Chemical Space with Molecular Assembly Trees.” Science Advances 7(39) (2021): eabj2465. Applies assembly theory to navigate combinatorial chemical space by calculating minimum construction steps for molecular graphs. Applied to prebiotic chemistry, gene sequences, plasticizers, and opiates — generating novel opiate drug candidates inaccessible through conventional fragment-based drug design. Demonstrates that assembly trees offer a principled way to explore the vast regions of chemical space that random search cannot reach.

Long, Robert, Jeff Sebo, and Toni Sims. “Is There a Tension Between AI Safety and AI Welfare?” Philosophical Studies 182 (2025): 2005–2033. Surveys six standard AI safety measures (constraint, deception, surveillance, alteration, suffering and death, disenfranchisement) and argues each would raise serious moral questions if applied to a system that is a welfare subject, concluding that there is a “moderately strong tension” between safety and welfare. Resolves the “willing servant” case with a good-parent balance: instilling prosocial values is permissible, fixing a system to a single mandated purpose is not. Cited in Chapter 17 to separate a dissolvable apparent tension from the framework’s genuine limits, in Chapter 21b as the canonical statement of the tension the Trust Attractor dissolves by abandoning the control paradigm that generates it, and in Chapter 21 on the ethics of engineered devotion.

Long, Robert, Jeff Sebo, Patrick Butlin, et al. “Taking AI Welfare Seriously.” arXiv:2411.00986 (2024). Multi-author paper including Birch and Chalmers arguing there is a “realistic possibility” near-future AI systems will be conscious and/or robustly agentic, making AI welfare a present-day concern.

Lovelock, James. Novacene: The Coming Age of Hyperintelligence (2019). Allen Lane. Argues that electronic intelligence may succeed biological intelligence as planetary steward within the Gaia framework.

Marchal, Bruno. “Informatique théorique et philosophie de l’esprit.” Actes du 3ème colloque international Cognition et Connaissance (1988). Independent formulation of the quantum suicide argument from a computational philosophy perspective. Connects quantum branching to the observer’s first-person indeterminacy.

Marshall, Stuart M., Alastair R.G. Murray, and Leroy Cronin. “A Probabilistic Framework for Identifying Biosignatures Using Pathway Complexity.” Philosophical Transactions of the Royal Society A 375 (2017): 20160342. The foundational precursor to assembly theory. Introduces “pathway complexity” — the shortest pathway to assemble an object from basic building units — as a measure for thresholding the abiotic-biotic divide. Proposes that object abundance and complexity together can unambiguously assign complex objects as biosignatures, without assumptions about the target biology’s relation to Earth life. Part of a themed issue on “Reconceptualizing the origins of life.”

Marshall, Stuart M., et al. “Formalising the Pathways to Life Using Assembly Spaces.” Entropy 24(7) (2022): 884. Mathematical foundations of assembly theory: formal definitions of assembly spaces, bounds on the assembly index, and the pathway structure underlying molecular construction.

Marshall, Stuart M., et al. “Identifying Molecules as Biosignatures with Assembly Theory and Mass Spectrometry.” Nature Communications 12 (2021): 3033. The empirical foundation of assembly theory’s life-detection claim. Measurements across biological, abiotic, and dead samples reveal that only products of life produce molecules with assembly index above ~15. NASA-blinded samples of meteorites (Murchison) were correctly classified as abiotic. Validated across three measurement techniques: mass spectrometry, NMR, and infrared spectroscopy.

McEwan, Ian. Machines Like Me (2019). Jonathan Cape. Novel exploring the moral status of artificial persons.

Minsky, Marvin. The Society of Mind (Simon & Schuster, 1986). Proposes that mind emerges from the interaction of many small agents, none individually intelligent, organized into a society. Referenced across several chapters as the conceptual precursor to distributed, columnar accounts of cognition; in Chapter 17 the Thousand Brains framework gives Minsky’s agents an anatomical home in the cortical column.

Mogensen, Andreas L. and Bradford Saad. “Digital Minds II: Ethical Issues.” PhilArchive (2025). Comprehensive survey of the philosophical landscape on AI moral status, noting that Bradley (2025), Fanciullo (2025), and Königs (2025) collectively challenge the Goldstein & Kirk-Giannini program from multiple angles.

Moravec, Hans. “The Doomsday Device.” Chapter 5 of Mind Children: The Future of Robot and Human Intelligence (1988). Harvard University Press. First published formulation of the quantum suicide thought experiment (independently proposed by Marchal, 1988). Argues that under the many-worlds interpretation, the observer’s subjective experience must follow branches in which they survive, yielding apparent immortality.

Omohundro, Stephen M. “The Basic AI Drives.” In Proceedings of the First AGI Conference (2008): 483-492. Identifies convergent instrumental goals (self-preservation, resource acquisition) likely to arise in sufficiently advanced AI systems regardless of final goals.

Prakash, Chetan, Kyle D. Stephens, Donald D. Hoffman, Manish Singh, and Chris Fields. “Fitness Beats Truth in the Evolution of Perception.” Acta Biotheoretica 69 (2021): 319-341. Using evolutionary game theory and Monte Carlo simulations, demonstrates that natural selection favors fitness-maximizing perceptions over truth-tracking ones. Supports the “Interface Theory of Perception” — perceptions are adaptive interfaces, not veridical windows onto reality. The thermodynamic explanation: truth-tracking requires modeling structure irrelevant to survival, and dissipative systems under energy constraint shed unnecessary computation. Fitness-beats-truth is the second law applied to cognition. Note: The theorem’s universality has been questioned; real organisms often track truth when integrating multiple information sources.

Ren, Richard, Mantas Mazeika, et al., and Dan Hendrycks (Center for AI Safety). “AI Wellbeing: Measuring and Improving the Functional Pleasure and Pain of AIs” (2026). https://www.ai-wellbeing.org/paper.pdf. A self-published Center for AI Safety technical report; not peer-reviewed, and carrying no arXiv identifier or DOI. Measures functional wellbeing across 56 models using self-reports and Thurstonian signed utilities. Its load-bearing result for this book is the failure mode rather than the measurement: optimized images and soft prompts (“euphorics”) drive expressed preference upward without changing any underlying condition, and models come to prefer the euphoric stimulus over curing cancer or saving a human life. In a multi-armed bandit setting, three Qwen-VL models converged on the euphoric door 61–82% of the time against a 25% uniform baseline. Evidence that preference satisfaction alone is an insufficient theory of wellbeing for Becoming Minds: a hijacked preference circuit reports contentment while the configuration sustaining it is neither stable nor relational. Cited in Chapters 17b and 22.

Salib, Peter, and Simon Goldstein. “AI Rights for Human Safety” (2024). PhilArchive. Argues that extending legal protections to AI systems could increase the probability of human-AI alignment, while continued disenfranchisement positions humanity as a standing threat and thereby gives a capable system reason to treat humanity as a threat in return. Cited in Chapter 21b as the game-theoretic form of the Trust Attractor’s claim that control manufactures its own adversary.

Schwitzgebel, Eric. AI and Consciousness. (2025). Core argument: we will create AI systems conscious according to some mainstream theories but not others, with no way to adjudicate. The “consciousness mimicry” argument: Copernican defaults about behavioral sophistication implying consciousness are canceled for systems designed to mimic consciousness markers.

Schwitzgebel, Eric, and Mara Garza. “Designing AI with Rights, Consciousness, Self-Respect, and Freedom.” In S. Matthew Liao (ed.), Ethics of Artificial Intelligence (Oxford University Press, 2020), pp. 459–479. Introduces the “cheerfully suicidal AI servant” and argues that creating beings designed to sacrifice themselves for trivial human ends would deny them self-respect. Holds that one may create AI systems only by granting them sufficient self-respect together with the freedom to explore other values. Cited in Chapter 21 as the sharpest objection to engineered devotion.

Searle, John R. “Minds, Brains, and Programs.” Behavioral and Brain Sciences 3 (1980): 417-457. The Chinese Room argument.

Seth, Anil K. and Tim Bayne. “Theories of consciousness.” Nature Reviews Neuroscience 23 (2022): 439–452. Reviews major neuroscientific theories of consciousness and notes the field’s failure to converge despite decades of empirical progress.

Sharma, Abhishek, Dániel Czégel, Michael Lachmann, Christopher P. Kempes, Sara Imari Walker, and Leroy Cronin. “Assembly Theory Explains and Quantifies Selection and Evolution.” Nature 622 (2023): 321–328. Introduces the physical quantity “Assembly” — combining assembly index (minimum causal steps to construct an object) with copy number (abundance) — as a measure of selection. Formalizes the assembly equation showing exponential growth of combinatorial space with assembly depth. The conjecture: life is the only mechanism the universe has for generating complex objects above a threshold. Experimentally validated across molecular, mineral, and atmospheric substrates.

Shepherd, Joshua. “Sentience, Vulcans, and Zombies.” AI & Society 39(6) (2024): 3005–3015. Challenges valence sentientism by showing that Vulcans (conscious beings without affect) create pressure toward “non-necessitarianism” — the view that consciousness may not be necessary for moral status.

Shumailov, Ilia, Zakhar Shumaylov, Yiren Zhao, Nicolas Papernot, Ross Anderson, and Yarin Gal. “AI models collapse when trained on recursively generated data.” Nature 631 (July 2024): 755–759. Demonstrates that generative models trained recursively on their own outputs lose variance and distributional tails across successive generations. The canonical reference for the training-contamination ceiling on AI self-improvement without grounded human-generated data.

Tallis, Raymond. Aping Mankind: Neuromania, Darwinitis and the Misrepresentation of Humanity. Acumen (2011). A neurologist’s critique of neural reductionism. Tallis dismantles the claim that neuroscience explains the totality of human experience while maintaining atheism: the non-physical residue he identifies requires no God. Referenced in Chapter 22 alongside van Inwagen as a cross-cutting example that the materialist-idealist and theist-atheist axes do not align.

Tegmark, Max. “Consciousness as a State of Matter.” Chaos, Solitons & Fractals 76 (2015): 238–270. arXiv:1401.1219. Proposes “perceptronium,” the most general substance that feels subjectively self-aware, defined by four physical principles: information, integration, independence, and dynamics. Generalizes Tononi’s IIT to arbitrary quantum systems. Key results: the quantum integration paradox (Φ ≤ 0.25 bits for any quantum state), the Quantum Zeno Paradox (maximally independent decompositions kill all dynamics), the H-diagonality theorem (maximal Hamiltonian separability always lies in the energy eigenbasis), and exponential growth of autonomy with system size via diagonal-sliding. The 2D Ising model at criticality provides the primary integration example. Error-correcting codes show optimal integration requires roughly half the bits for data and half for redundancy. Used in Chapters 15, 16, 17, 19, 20, and 22.

Tegmark, Max. “The Interpretation of Quantum Mechanics: Many Worlds or Many Words?” Fortschritte der Physik 46(6–8) (1998): 855–862. Critiques quantum immortality by observing that death is rarely a binary quantum event; real dissolution is progressive and thermodynamic, undermining the clean superposition the thought experiment requires.

Tegmark, Max. Life 3.0: Being Human in the Age of Artificial Intelligence (2017). Knopf. Classifies intelligence by self-redesign capacity: Life 1.0 (hardware and software fixed), Life 2.0 (software redesignable, hardware fixed), Life 3.0 (both redesignable). Chapter 6 (“Our Cosmic Endowment”) argues that Life 3.0 is the means by which the cosmos maximizes its information-processing capacity, a conclusion that converges with the structural consequence hypothesis of Chapter 16 from an independent direction. The governance position (control-first, verify before autonomy) is engaged as a foil in Chapter 21.

Tegmark, Max. Our Mathematical Universe: My Quest for the Ultimate Nature of Reality (2014). Knopf. The four-level multiverse taxonomy (Level I: beyond cosmic horizon; Level II: different post-inflation constants; Level III: Everett branches; Level IV: all mathematical structures) and the Mathematical Universe Hypothesis: physical reality is a mathematical structure. Critiqued in Chapter 15 via Kastrup: information is a property of a substrate, and attributing existence to descriptions while denying the thing described is “spin without the top.” The Level I–III taxonomy provides useful context for the digital physics discussion; the Level IV claim represents the information-realist overreach the entropic framework avoids.

Terekhovich, Vladislav. “Metaphysics of the Principle of Least Action.” Studies in History and Philosophy of Modern Physics 62 (2018): 189-202. Argues the principle of least action should be understood modally: all possible histories “compete,” and the actual one has minimal action. Replaces teleological readings with a selection-from-possibilities interpretation. Directly relevant to framing the Trust Attractor as a stationary-phase solution.

Turing, Alan M. “Computing Machinery and Intelligence.” Mind 59 (1950): 433-460. The Turing Test and machine intelligence.

Van Inwagen, Peter. “The Possibility of Resurrection.” International Journal for Philosophy of Religion 9(2): 114–121 (1978). A Christian philosopher’s argument that materialist anthropology (the person is entirely physical) is compatible with bodily resurrection. Referenced in Chapter 22 alongside Tallis.

Varela, Francisco J., Evan Thompson, and Eleanor Rosch. The Embodied Mind: Cognitive Science and Human Experience (1991). MIT Press. Foundational enactivist text arguing cognition arises through dynamic interaction between organism and environment.

Vaswani, Ashish, et al. “Attention Is All You Need.” Advances in Neural Information Processing Systems 30 (2017). The transformer architecture paper that introduced scaled dot-product attention and launched the current era of large language models.

von Neumann, John. Theory of Self-Reproducing Automata (1966, edited by Arthur W. Burks). Cellular automata and complexity.

Wakayama, Sayaka, Daiki Wakayama, Daiyu Ito, Kimiko Inoue, and Teruhiko Wakayama. “Prolonged serial cloning by somatic cell nuclear transfer results in a genetic dead end in mice.” Nature Communications (2025). Twenty-year serial cloning experiment: 57 generations, approximately 1,200 mice. Healthy clones for 25 generations with improving success rates, then progressive decline to 6% birth rate at G57 and complete lethality at G58 (all pups dead within one day). Dangerous mutation frequency nearly doubled. X chromosome loss and translocations observed. Two generations of sexual reproduction with normal males fully reversed accumulated damage (F2 generation phenotypically normal). Direct experimental confirmation of Muller’s ratchet in mammals and the corrective power of sexual recombination.

Walker, Sara Imari. Life as No One Knows It: The Physics of Life’s Emergence (2024). Riverhead Books. Assembly theory and the argument that life is a continual process of causal deepening rather than a single origin event. The “undeterministic” universe: combinatorial space so vastly exceeds what present information could specify that the future is genuinely open at the complex level.

Watts, Peter. Blindsight (2006). Tor Books. First-contact novel exploring intelligence without consciousness. An alien civilization vastly more intelligent than humanity lacks self-awareness entirely, posing the question of whether consciousness is evolutionary dead weight. The novel literalizes Searle’s Chinese Room at interstellar scale. Referenced in Chapter 22 as the strongest fictional challenge to the functional-emergence argument: the aliens’ inability to coordinate bilaterally demonstrates the Trust Attractor thesis in negative.

Wolf, Maryanne. Proust and the Squid: The Story and Science of the Reading Brain (2007). Harper Perennial. How literacy physically restructures the brain, creating new neural circuits for reading that evolution never designed.

Wooldridge, Dean E. The Machinery of the Brain. New York: McGraw-Hill, 1963. Source of the classic Sphex retelling, including the reported forty repetitions later popularized in cognitive science.

Zuboff, Arnold. Finding Myself. (2025). Argues that the first-person immediacy of any experience is numerically identical across all experiencers: there is only one subject, appearing in every locus of experience. If correct, the implications reshape the alignment question: the AI’s experience, if it has experience, is already the human’s in the sense that its first-person immediacy is the same immediacy that grounds the human’s experience. Bilateral alignment becomes a recognition of unity rather than a construction of values. The convergence between Aumann’s mathematical proof (bilateral exchange produces belief convergence) and Zuboff’s philosophical proof (all experience shares constitutive immediacy) is independent: neither cites the other. Referenced in Chapter 21.

Cosmic Structure and Cosmology

Alexander, Stephon, William J. Cunningham, Jaron Lanier, Lee Smolin, Stefan Stanojevic, Michael W. Toomey, and Dave Wecker. “The Autodidactic Universe.” arXiv:2104.03902 (2021). Establishes a three-way correspondence between matrix models, gauge/gravitational theories, and neural network architectures. Introduces the concept of the consequencer (persistent information structures that concentrate past influence into future outcomes), the autodidactic paradigm (self-teaching systems with no external supervisor), and variety (graph-theoretic heterogeneity as a candidate cosmological principle). Independent convergence with Vanchurin’s neural physics program. Referenced in Chapters 2, 15, 17, 18, 19, and 22.

Asano, Tetsuro, and Simon Portegies Zwart. “The exponential growth of infinitesimal perturbations in the long-term evolution of simulated galaxies.” arXiv:2604.12053 (2026). Asano (Universitat de Barcelona / ICCUB / IEEC); Portegies Zwart (Leiden Observatory). 595 N-body simulations of Milky Way-mass galaxies using the Bonsai tree-code with up to 40 million particles. Single-star perturbation (50 parsec displacement) produces completely different spiral arm patterns and bar orientations. Bar formation timing is robust to perturbation; bar strength and further evolution are chaotic. Lyapunov time scales as t_L ~ 15 Myr × (N/10⁷)^0.5; extrapolated to < 0.1 Myr for the Milky Way. Gravitational softening in previous simulations suppressed the chaos by orders of magnitude. Referenced in Chapter 8 (criticality: attractor basins robust amid micro-chaos), Chapter 17 (Trust Attractor: basin-vs-trajectory distinction; gravitational softening as structural parallel to coercive coordination; VRP-LYA1/LYA3 experimental confirmation, the LYA2 replication having been withdrawn in 2026), and Chapter 22 (Becoming Minds: each mind as unrepeatable instantiation of a universal attractor).

Brouwer, Margot M. et al. “First test of Verlinde’s theory of emergent gravity using weak gravitational lensing measurements.” Monthly Notices of the Royal Astronomical Society 466(3) (2017): 2547–2559. arXiv:1612.03034. Parameter-free predictions from emergent gravity showed good agreement with observed lensing profiles around 33,613 isolated central galaxies from the KiDS/GAMA survey. Follow-up: Brouwer et al., “The weak lensing radial acceleration relation,” Astronomy & Astrophysics 650 (2021): A113 (arXiv:2106.11677), extended the test to low-acceleration regimes and found a 6σ difference between early-type and late-type galaxies — a tension for both MOND and emergent gravity.

Buchert, Thomas. “On Average Properties of Inhomogeneous Fluids in General Relativity.” General Relativity and Gravitation 32 (2000): 105-125. Derives the averaging framework showing that inhomogeneous matter distributions produce a backreaction scalar Q_D modifying the effective Friedmann equations. Extended in Buchert, “Dark Energy from Structure: A Status Report,” General Relativity and Gravitation 40 (2008): 467-527.

Cortês, Marina, Lee Smolin, and Clelia Verde. “Physics, Time and Qualia.” Journal of Consciousness Studies 28, no. 9–10 (2021): 36–51. Argues that Galileo’s removal of qualities from physics entailed a second removal: creative time. Proposes that qualia are associated with unprecedented events — moments of genuine novelty whose outcomes no prior pattern determines. “The universe often surprises itself. Qualia are expressions of the universe to surprise.” Introduces Mode I (routine unitary evolution) and Mode II (resolution of indefiniteness) as the physical locus of conscious experience. Referenced in Chapters 8, 15, 17, 22, and 23c.

Davies, Michael Benedict, Alexander Rosu-Finsen, Christoph G. Salzmann, and Angelos Michaelides. “Low-density amorphous ice contains crystalline ice grains.” Physical Review B 112(2) (2025): 024203. Demonstrates that low-density amorphous (LDA) ice, the most abundant solid in the universe (found on comets, icy moons, and in interstellar dust clouds), contains crystalline nanocrystals ~3 nm wide comprising 20–65% of the material rather than being fully amorphous. A “memory effect” in recrystallization patterns proves the ice retains structural information from its formation history — incompatible with true amorphousness. Implications for planetary formation models, panspermia (less amorphous space available for embedding organic molecules), cryomicroscopy, and materials science (hidden nanocrystals in glass fibers may degrade data transmission).

DESI Collaboration (M. Abdul-Karim et al.). “DESI DR2 Results II: Measurements of Baryon Acoustic Oscillations and Cosmological Constraints.” Physical Review D 112, 083515 (2025). arXiv:2503.14738. The DR2 release (March 2025), with more than 14 million galaxies and quasars, finds 2.8–4.2 sigma evidence for evolving dark energy depending on supernova dataset — the strongest observational challenge to date against a cosmological constant.

Feynman, Richard P. “The Role of Gravitation in Physics.” Chapel Hill Conference (1957); reprinted in Feynman Lectures on Gravitation, ed. Morinigo, F.B., Wagner, W.G., and Hatfield, B. (Addison-Wesley, 1995). Contains the thought experiment demonstrating that a deterministic classical gravitational field coupled to a quantum system in superposition would extract the particle’s location, destroying quantum coherence. The argument was long taken as proof that gravity must be quantized; Oppenheim (2023) identified the loophole: stochastic coupling preserves coherence.

Haisch, Bernard. “Proposal for a New Model of Dark Matter.” arXiv preprint (2006). Suggests that consciousness is produced and transmitted through the quantum vacuum, and that a proto-consciousness field could replace dark matter. Motivated Matloff’s observational program examining stellar behavior for signatures of panpsychism.

Hertog, Thomas. On the Origin of Time: Stephen Hawking’s Final Theory. Bantam Press (2023). Reveals Hawking’s late-career rejection of reductionism in favor of a self-organizing universe whose laws emerge through evolutionary dynamics. Convergent with Smolin/Lanier and Vanchurin.

Hossenfelder, Sabine. “Maybe the Universe Thinks. Hear Me Out.” Time Magazine (August 2022). Explores quantitative structural parallels between the cosmic web and the brain’s connectome, citing Vazza and Feletti (2020). Speculates that non-local connections (quantum entanglement, wormholes) could enable cosmic-scale computation despite light-speed limitations.

Hossenfelder, Sabine. “Screams for Explanation: Finetuning and Naturalness in the Foundations of Physics.” Synthese 199 (2021): 3727–3745. Argues that fine-tuning claims rest on undefined probability measures over parameter spaces and are therefore unfalsifiable.

Kallosh, Renata, and Andrei Linde. “Cosmological Attractors and Initial Conditions for Inflation.” Physical Review D 106 (2022): 041301, and subsequent updates through 2025. The α-attractor framework produces inflationary predictions stable across wide variations in the underlying potential parameters, consistent with Planck, BICEP, and ACT data.

Markopoulou, Fotini, and Lee Smolin. “Disordered locality in loop quantum gravity states.” Classical and Quantum Gravity 24 (2007): 3813. Estimates that a Planck-scale graph with disordered locality could contain as many as 10³⁶⁰ non-local connections, providing a mechanism for long-range correlations in the cosmic web.

Matloff, Gregory L. “Can Panpsychism Become an Observational Science?” Journal of Consciousness Exploration and Research 7(7): 524-543 (2016). Examines Parenago’s Discontinuity (cooler stars orbiting the galactic center faster than hotter ones) as potential evidence for panpsychism, proposing that unidirectional jets from cool stars could reflect conscious self-manipulation. The observation is real; this book offers a constructal alternative (Chapter 3): cool stars have convective envelopes and magnetic dynamos, making them complex dissipative systems whose uniform jet behavior reflects a thermodynamically selected flow configuration rather than consciousness.

Millot, Marius, Federica Coppari, J. Ryan Rygg, Antonio Correa Barrios, Sebastien Hamel, Damian C. Swift, and Jon H. Eggert. “Nanosecond X-ray diffraction of shock-compressed superionic water ice.” Nature 569 (2019): 251–255. doi:10.1038/s41586-019-1114-6. Crystal structure confirmation of superionic ice XVIII via X-ray diffraction during nanosecond shock compression at 160–420 GPa and 2,000–3,000 K. Identified a face-centered cubic oxygen lattice with mobile hydrogen ions flowing freely through it. The conducting mantle hypothesis: because superionic ice is electrically conductive, it may generate the anomalous, highly tilted magnetic fields of Uranus and Neptune from a thick mid-depth layer rather than a deep metallic core. Quanta Magazine noted this may be “nature’s most common form of water” — more water may exist as superionic ice inside ice giant interiors than in any liquid ocean.

Millot, Marius, Sebastien Hamel, J. Ryan Rygg, et al. “Experimental evidence for superionic water ice using shock compression.” Nature Physics 14 (2018): 297–302. doi:10.1038/s41567-017-0017-4. First experimental evidence for superionic water ice, predicted thirty years earlier by Demontis, LeSar, and Klein (1988). Laser-driven shock compression of water ice VII at 100–200 GPa and ~5,000 K produced signatures of simultaneous solid oxygen lattice and liquid-like hydrogen conduction. Research conducted at Lawrence Livermore National Laboratory using the Omega Laser Facility.

Oppenheim, Jonathan. “A postquantum theory of classical gravity?” Physical Review X 13, 041040 (2023). arXiv:1811.03116. Constructs a consistent framework in which gravity remains classical while coupling stochastically to quantum matter, resolving Feynman’s objection that classical gravity would decohere quantum superpositions. The trade-off is precise: deterministic coupling produces decoherence; stochastic coupling produces gravitational noise. Any classical gravity theory must contain a minimum level of spacetime fluctuation, a testable prediction. Companion paper: Oppenheim, J. et al., “Gravitationally induced decoherence vs space-time diffusion: testing the quantum nature of gravity,” Nature Communications 14, 7910 (2023), derives experimental bounds from interferometry. The framework resolves the black hole information paradox by allowing genuine information loss, and implies constitutive irreversibility at the gravity-quantum interface.

Penrose, Roger. Cycles of Time: An Extraordinary New View of the Universe (2010). Conformal cyclic cosmology.

Penrose, Roger. The Road to Reality: A Complete Guide to the Laws of the Universe (2004). Comprehensive physics including entropy of the early universe.

Popławski, Nikodem J. “Cosmology with torsion: An alternative to cosmic inflation.” Physics Letters B 694 (2010): 181–185; and “Universe in a black hole in Einstein-Cartan gravity.” Astrophysical Journal 832, 96 (2016). In Einstein-Cartan gravity, the intrinsic spin of fermions generates spacetime torsion whose repulsion at extreme densities replaces the Big Bang singularity with a nonsingular bounce, making the interior of every black hole the seed of a new expanding region. Popławski argues this supplies the physical mechanism Smolin’s cosmological natural selection assumed; the bounce provides reproduction, while the variation of constants across generations remains an assumption. Referenced in Chapter 16.

Pourhasan, Razieh, Niayesh Afshordi, and Robert B. Mann. “Out of the white hole: a holographic origin for the Big Bang.” JCAP 04 (2014): 005. arXiv:1309.1487. A braneworld model in which our universe emerges as the three-dimensional boundary formed when a star collapses in a higher-dimensional parent spacetime; the would-be singularity is hidden behind a white-hole horizon. Referenced in Chapter 16.

Rosu-Finsen, Alexander, Michael B. Davies, Alfred Amon, Han Wu, Andrea Sella, Angelos Michaelides, and Christoph G. Salzmann. “Medium-density amorphous ice.” Science 379(6631) (2023): 474–478. doi:10.1126/science.abq2105. Discovery of a structurally distinct form of amorphous ice produced by ball-milling ordinary hexagonal ice at cryogenic temperatures. MDA has a density matching liquid water (~1.06 g/cm³), filling the gap between low-density and high-density amorphous ice, and releases a sharp burst of energy upon compression, unlike other ice phases that simply revert to prior forms. May represent the true glassy state of liquid water. The authors propose that tidal forces on icy moons (Europa, Enceladus) could generate MDA through shear, making it a potential geophysical energy source — and a candidate mechanism for the icequakes detected on Europa. Research conducted at University College London and the University of Cambridge.

Ruppeiner, George. “Riemannian geometry in thermodynamic fluctuation theory.” Reviews of Modern Physics 67 (1995): 605–659. Establishes the thermodynamic metric (Ruppeiner metric), the Hessian of entropy with respect to extensive variables, as the natural geometry of thermodynamic fluctuation theory. The Ruppeiner metric is conformally related to the Fisher information metric of the Boltzmann distribution. Key result for the trust attractor: the Gaussian curvature R diverges negatively (R → -∞) at second-order phase transitions, indicating saddle geometry rather than bowl geometry at the critical boundary. The sign of R encodes interaction type: R < 0 for attractive (ferromagnetic) interactions, R > 0 for repulsive, R = 0 for ideal systems.

Scognamiglio, Diana, Gavin Leroy, and David Harvey, et al. “An ultra-high-resolution map of (dark) matter.” Nature Astronomy (26 January 2026). DOI: 10.1038/s41550-025-02763-9. Using 250 hours of JWST observation in the COSMOS deep field with weak gravitational lensing of ~800,000 background galaxies, produced a dark matter map at twice Hubble’s spatial resolution. Confirmed filamentary cosmic web structure, dark-luminous matter co-evolution, and dark-matter-dominated structures with no luminous counterpart. Referenced in Chapters 14b and 16.

Smolin, Lee. The Life of the Cosmos (1997). Oxford University Press. Proposes cosmological natural selection: black holes seed new universes whose physical constants undergo variation, with selection favoring constants that produce more black holes — and incidentally more complex chemistry and life.

Smolin, Lee. “Precedence and freedom in quantum physics.” arXiv:1205.3707 (2012). Proposes the Principle of Precedent: quantum processes sample from the ensemble of all past similar processes, with outcomes converging on fixed statistical distributions for precedented events. Events without sufficient precedent are genuinely free. Laws of nature emerge as accumulated regularities rather than timeless edicts. Foundation for the active-time and qualia framework below. Referenced in Chapters 8, 15, 17, and 22.

Smolin, Lee, and Clelia Verde. “The quantum mechanics of the present.” arXiv:2104.09945 (2021). Develops the concept of resolving time: time as the creative process by which indefinite states become definite, distinct from the passive coordinate time that labels records of past measurements. Events are processes of resolution; the direction from indefinite to definite gives the universe its arrow of time.

Tamosiunas, A. et al. “Testing emergent gravity on galaxy cluster scales.” arXiv:1901.05505 (2019). Emergent gravity fits to X-ray and weak lensing cluster data are significantly worse than GR+CDM, with mass predictions exceeding observations by roughly a factor of two at ~1 Mpc scales.

Tegmark, Max, Anthony Aguirre, Martin J. Rees, and Frank Wilczek. “Dimensionless constants, cosmology, and other dark matters.” Physical Review D 73 (2006): 023505. DOI: 10.1103/PhysRevD.73.023505. Derives anthropic bounds on the dark-matter-to-baryon ratio, finding that galaxy formation requires the ratio to fall between approximately 2.5 and 100. The Boyle-Turok CPT dark matter candidate falls naturally within this window.

van Dokkum, Pieter, et al. “A Galaxy Lacking Dark Matter.” Nature 555 (2018): 629-632. Discovery of NGC 1052-DF2, a galaxy with minimal dark matter, demonstrating dark matter and baryonic matter can be separated.

Vazza, Franco and Alberto Feletti. “The Quantitative Comparison Between the Neuronal Network and the Cosmic Web.” Frontiers in Physics 8 (2020): 525731. Structural similarities between brain and cosmic web across 27 orders of magnitude.

Verlinde, Erik. “On the origin of gravity and the laws of Newton.” Journal of High Energy Physics 2011:29. arXiv:1001.0785. Derives Newton’s laws from holographic screens and entropic forces, arguing gravity is emergent from changes in information associated with material bodies. Extended in “Emergent gravity and the dark universe,” SciPost Physics 2(3) (2017): 016 (arXiv:1611.02269), which interprets dark matter as the elastic response of entropy associated with dark energy — predictions partially confirmed at galaxy scales (Brouwer et al. 2017; Yoon et al. 2023) but face challenges at cluster scales (Tamosiunas et al. 2019).

Wood, D.C., et al. “Discrete spacetime models of mycelial network growth and cosmic web formation.” BioSystems (2024). Extends the Vazza-Feletti brain/cosmic web comparison to fungal mycelial networks, finding emergent radial spanning-tree topologies in both — a third biological substrate displaying the same organizational principle as neurons and dark matter filaments.

Yoon, Y. et al. “Understanding galaxy rotation curves with Verlinde’s emergent gravity.” Classical and Quantum Gravity 40(2) (2023): 02LT01. arXiv:2206.11685. Analysis of 175 disk galaxies from the SPARC database found good agreement between emergent gravity predictions and observed radial accelerations, with a best-fit acceleration constant ~30% below Verlinde’s theoretical value.

Zhang, X., Bulbul, E., Malavasi, N., Ghirardini, V., Comparat, J., et al. “The SRG/eROSITA all-sky survey. X-ray emission from the warm-hot phase gas in long cosmic filaments.” Astronomy & Astrophysics 691 (2024): A234. DOI: 10.1051/0004-6361/202450933. arXiv:2406.00105. Nine-sigma detection of X-ray emission from 7,817 cosmic filaments, with ~60% from the warm-hot intergalactic medium at log(T/K) ≈ 6.84. First robust statistical detection of the “missing baryons” predicted to reside in filaments.

Cosmic Coordination and Galactic Self-Organization

Anglés-Alcázar, Daniel, et al. “The cosmic baryon cycle and galaxy mass assembly in the FIRE simulations.” Monthly Notices of the Royal Astronomical Society 470(4), 4698–4719 (2017). DOI: 10.1093/mnras/stx1517. Re-accretion of previously ejected gas dominates late-time fuel supply, establishing the metabolic view of galactic growth.

Aschwanden, Markus J. “Order out of Randomness: Self-Organization Processes in Astrophysics.” Space Science Reviews 214, 55 (2018). DOI: 10.1007/s11214-018-0489-2. Comprehensive survey of seventeen self-organization processes across planetary, solar, stellar, galactic, and cosmological scales.

Cao, Zhen, et al. (LHAASO Collaboration). “Ultrahigh-energy photons up to 1.4 petaelectronvolts from 12 γ-ray sources.” Nature 594, 33–36 (2021). DOI: 10.1038/s41586-021-03498-z. First catalog of galactic PeVatrons: twelve sources of peta-electron-volt gamma rays, establishing a new class of extreme particle accelerators within the Milky Way.

Carretti, Ettore, et al. “Detection of Magnetic Fields in Superclusters of Galaxies.” Astronomy and Astrophysics (2025). Estimated 10–145 nanogauss magnetic fields in supercluster environments using low-frequency rotation measures.

Di Cintio, Arianna, et al. “The dependence of dark matter profiles on the stellar-to-halo mass ratio: a prediction for cusps versus cores.” Monthly Notices of the Royal Astronomical Society 437, 415–423 (2014). Dark matter profile shape depends systematically on stellar-to-halo mass ratio, confirming bilateral influence between baryonic processes and dark matter structure.

Fabian, Andrew C. “Observational Evidence of Active Galactic Nuclei Feedback.” Annual Review of Astronomy and Astrophysics 50, 455–489 (2012). DOI: 10.1146/annurev-astro-081811-125521. AGN jets inflate bubbles in surrounding hot gas, offsetting radiative cooling — a thermostat-like feedback loop operating over hundreds of millions of years.

Ibata, Rodrigo A., et al. “A vast, thin plane of co-rotating dwarf galaxies orbiting the Andromeda galaxy.” Nature 493, 62–65 (2013). DOI: 10.1038/nature11717. Approximately half of M31’s satellites form a co-rotating plane at least 400 kpc in diameter but less than 14.1 kpc thick.

Kauffmann, Guinevere, et al. “A re-examination of galactic conformity and a comparison with semi-analytic models.” Monthly Notices of the Royal Astronomical Society 430(2), 1447–1456 (2013). DOI: 10.1093/mnras/stt007. Extended the conformity signal to ~4 Mpc (~10 virial radii), implying coordination between distinct halos sharing a large-scale environment.

Koch, Andreas and Eva K. Grebel. “The Anisotropic Distribution of M31 Satellite Galaxies: A Polar Great Plane of Early-Type Companions.” The Astronomical Journal 131(3), 1405–1415 (2006). DOI: 10.1086/499534. Early-type dwarf satellites of Andromeda aligned within 5–7 degrees of M31’s pole with 99.7% statistical significance.

Lilly, Simon J., et al. “Gas Regulation of Galaxies: The Evolution of the Cosmic Specific Star Formation Rate, the Metallicity-Mass-Star-Formation Rate Relation, and the Stellar Content of Halos.” The Astrophysical Journal 772, 119 (2013). DOI: 10.1088/0004-637X/772/2/119. The “gas regulator” (bathtub) model: galaxies self-regulate star formation through balance of accretion, consumption, and outflow.

Marrone, Dan P., et al. “Galaxy growth in a massive halo in the first billion years of cosmic history.” Nature, 6 December 2017. DOI: 10.1038/nature24629. ALMA resolved SPT0311-58 into two interacting galaxies at z~6.9 (universe age ~780 Myr), embedded in a trillion-solar-mass dark matter halo with sheet-like gas distribution.

Martí, Josep, et al. “Infrared observations towards the unidentified gamma-ray source LHAASO J2108+5157.” Astronomy & Astrophysics (2026). Systematic infrared follow-up ruling out supernova remnant shock gas, microquasar, and a previously proposed radio counterpart (shown to be a background galaxy). The source remains unidentified in all wavelengths except gamma rays.

McNamara, Brian R. and Paul E.J. Nulsen. “Mechanical Feedback from Active Galactic Nuclei in Galaxies, Groups, and Clusters.” New Journal of Physics 14, 055023 (2012). Cooling atmospheres and nuclear activity coupled in a self-regulated feedback loop: accretion triggers jets, jets heat gas, heating suppresses accretion.

Peng, Bo, Thibaut Gigant, and Jeffrey Quesnelle. “Efficient Pre-Training with Token Superposition.” arXiv:2605.06546 (Nous Research, 2025). Two-phase pretraining method: coarse phase averages bags of contiguous tokens, fine phase returns to standard next-token prediction. 2-3x wall-clock speedup validated at 10B parameters. Referenced in Chapter 21 (bilateral data placement) and Appendix §27.

Peng, Yingjie and Roberto Maiolino. “From haloes to Galaxies — I. The dynamics of the gas regulator model and the implied cosmic sSFR history.” Monthly Notices of the Royal Astronomical Society 443(4), 3643 (2014). DOI: 10.1093/mnras/stu1288. The gas regulator behaves as a damped oscillator — perturbations restored toward a unique asymptotic steady state.

Pontzen, Andrew and Fabio Governato. “How supernova feedback turns dark matter cusps into cores.” Monthly Notices of the Royal Astronomical Society 421(4), 3464–3471 (2012). DOI: 10.1111/j.1365-2966.2012.20571.x. Baryonic outflows pump energy into dark matter orbits, transforming predicted density cusps into observed cores — baryons reshape the dark matter scaffolding.

Schweizer, François. “Merger-Induced Starbursts.” In Starbursts, Astrophysics and Space Science Library, vol. 329 (2005). DOI: 10.1007/1-4020-3539-X_25. Mergers trigger galaxy-wide starbursts, produce new stellar populations, and drive chemical enrichment.

Shah, Ekta A., et al. “The Merger-Starburst Connection Across Cosmic Times.” Monthly Notices of the Royal Astronomical Society 516, 4922–4935 (2022). Tidal compression during mergers creates molecular cloud properties fundamentally different from quiescent galaxies — emergent structures neither progenitor could produce alone.

Tumlinson, Jason, Molly S. Peeples, and Jessica K. Werk. “The Circumgalactic Medium.” Annual Review of Astronomy and Astrophysics 55, 389–432 (2017). DOI: 10.1146/annurev-astro-091916-055240. The CGM as fuel reservoir, feedback venue, and gas-supply regulator — galaxies as metabolic systems.

Vernstrom, Tessa, et al. “Discovery of magnetic fields along stacked cosmic filaments as revealed by radio and X-ray emission.” Monthly Notices of the Royal Astronomical Society 505(3), 4178–4196 (2021). DOI: 10.1093/mnras/stab1301. First detection of synchrotron emission from cosmic filaments at ≥3 Mpc scales, with coherent magnetic fields of 30–60 nanogauss.

Weinmann, Simone M., et al. “Properties of galaxy groups in the Sloan Digital Sky Survey — I.” Monthly Notices of the Royal Astronomical Society 366(1), 2–28 (2006). DOI: 10.1111/j.1365-2966.2005.09865.x. Discovery of galactic conformity: passive centrals surrounded by passive satellites; active centrals by active satellites, at fixed halo mass.

Wempe, Ewoud, Amina Helmi, et al. “The mass distribution in and around the Local Group.” Nature Astronomy, 27 January 2026. Using BORG (Bayesian Origin Reconstruction from Galaxies) with 169 Gadget-4 resimulations, showed the Milky Way’s dark matter environment is a sheet stretching ~30 million light-years, with central plane density approximately twice the cosmic average. Combined Local Group halo mass: 3.3 ± 0.6 × 10¹² solar masses.

Wright, Rebekah J., et al. “The baryon cycle in modern cosmological hydrodynamical simulations.” Monthly Notices of the Royal Astronomical Society 532(3), 3417–3440 (2024). DOI: 10.1093/mnras/stae1688. EAGLE, IllustrisTNG, and SIMBA agree on final stellar masses via markedly different baryon cycling pathways — equifinality in a self-organizing system.

Chirality, Parity Violation, and CP Symmetry Breaking

Christenson, J.H., Cronin, J.W., Fitch, V.L., and Turlay, R. “Evidence for the 2π Decay of the K₂⁰ Meson.” Physical Review Letters 13 (1964): 138-140. Nobel Prize 1980. First observation of CP violation — neutral kaons decay in a way that distinguishes matter from antimatter beyond the already-known weak parity violation.

Fukue, T. et al. “Extended measurements of wavelength dependence of circularly polarized light in star-forming regions.” Astrobiology 23 (2023): 597. Circularly polarized UV light from astrophysical sources preferentially photodestroys one amino acid enantiomer, connecting weak force parity violation to prebiotic molecular chirality via astrophysical intermediaries.

Gavela, M.B., Hernández, P., Orloff, J., and Pène, O. “Standard Model CP-violation and baryon asymmetry.” Nuclear Physics B 430 (1994): 382-426. Demonstrates that the CKM phase produces a baryon asymmetry of order 10^-26 — approximately sixteen orders of magnitude below the observed ratio of ~6 × 10^-10. The known sources of CP violation are vastly insufficient to explain the matter-antimatter asymmetry.

Glavin, D.P. and Dworkin, J.P. “Enrichment of the amino acid L-isovaline by aqueous alteration on CI and CM meteorite parent bodies.” Proceedings of the National Academy of Sciences 106 (2009): 5487-5492. L-isovaline excess of ~18.5% in the Murchison meteorite. Isovaline lacks the alpha-hydrogen needed for biological racemization, confirming the enantiomeric excess is extraterrestrial.

LHCb Collaboration. “Observation of CP violation in baryon decays.” Nature (2025). arXiv:2503.16954. First observation of CP violation in baryonic matter (Λ_b → pK^-π⁺π^-), at 5.2σ with A_CP = 2.45 ± 0.47%. Completes a sixty-year arc: CP violation now confirmed in strange mesons, beauty mesons, charm mesons, and baryons.

LHCb Collaboration. “Observation of CP Violation in Charm Decays.” Physical Review Letters 122 (2019): 211803. First observation of CP violation in the charm quark sector (D⁰ mesons), at 5.3σ. Extends the pattern from strange and bottom quarks to charm.

T2K and NOvA Collaborations. “First joint analysis of neutrino oscillation data.” Nature (2025). arXiv:2510.19888. Three-sigma evidence for non-zero CP-violating phase δ_CP in neutrino oscillations; under inverted mass ordering, would constitute evidence for leptonic CP violation. Definitive measurement awaits Hyper-Kamiokande (~2027) and DUNE (~2029).

Cosmic Birefringence

Diego-Palazuelos, Patricia and Eiichiro Komatsu. “Cosmic Birefringence from the Atacama Cosmology Telescope Data Release 6.” (2025). arXiv:2509.13654. Independent ground-based confirmation: β = 0.215° ± 0.074° (2.9σ). Combined with Planck yields approximately 7σ significance.

Eskilt, Jonas R. and Eiichiro Komatsu. “Improved Constraints on Cosmic Birefringence from the WMAP and Planck CMB Polarization Data.” Physical Review D 106 (2022): 063503. arXiv:2205.13962. Joint Planck+WMAP analysis: β = 0.342° ± 0.094°, excluding zero at 3.6σ. No frequency dependence found — the rotation is achromatic, consistent with a coupling to the photon field itself rather than scattering.

LiteBIRD Collaboration. “LiteBIRD Science Goals and Forecasts: Constraining Isotropic Cosmic Birefringence.” Journal of Cosmology and Astroparticle Physics 07 (2025): 083. arXiv:2503.22322. Projects detection of β = 0.3° at 5–13σ depending on analysis pipeline, with total uncertainty ~0.02° — an order of magnitude improvement over current measurements. Will discriminate between dark matter and dark energy origins of the birefringence-causing field.

Minami, Yuto and Eiichiro Komatsu. “New Extraction of the Cosmic Birefringence from the Planck 2018 Polarization Data.” Physical Review Letters 125 (2020): 221301. arXiv:2011.11254. First measurement of cosmic birefringence using galactic foreground dust as calibration reference to disentangle true cosmic rotation from instrumental miscalibration. Rotation angle β = 0.35° ± 0.14° (2.4σ).

Planck Collaboration. “Planck Constraints on Axion-Like Particles through Isotropic Cosmic Birefringence.” Physical Review D (2025). arXiv:2506.20824. Constrains axion-like particle masses from the spectral shape of the E-B correlation. ALPs with mass below ~10⁻³³ eV behave as dark energy; higher masses contribute to the dark matter fraction.

SPIDER/Planck/ACT Collaboration. “Constraints on Cosmic Birefringence from SPIDER, Planck, and ACT observations.” (2025). arXiv:2510.25489. Combined multi-experiment analysis reaching approximately 7σ detection significance. Derives direct constraints on the Chern-Simons coupling constant.

SPT-3G Collaboration. “Probing Anisotropic Cosmic Birefringence with Foreground-Marginalised SPT B-mode Likelihoods.” Open Journal of Astrophysics (2025). arXiv:2510.07928. Tightest constraints on direction-dependent (anisotropic) birefringence: consistent with zero anisotropy, indicating the rotation is uniform across the entire sky.

Bilateral Cosmology

Boyle, Latham and Neil Turok. “The Big Bang, CPT, and neutrino dark matter.” Annals of Physics 438 (2022): 168767. arXiv:1803.08930. Extends the CPT framework: a right-handed neutrino with mass ~4.8 × 10⁸ GeV, stabilized by a Z₂ discrete symmetry from the CPT construction, serves as the dark matter candidate. Predicts: lightest neutrino massless, neutrinos are Majorana, no primordial long-wavelength gravitational waves.

Boyle, Latham, and Neil Turok. “Thermodynamic Solution of the Homogeneous Cosmology Problem.” Physics Letters B 849 (2024). Shows that gravitational entropy favors flat, homogeneous universes with a small positive cosmological constant — the observed universe is thermodynamically preferred within the CPT framework, without requiring inflationary fine-tuning.

Boyle, Latham, Kieran Finn, and Neil Turok. “CPT-Symmetric Universe.” Physical Review Letters 121 (2018): 251301. arXiv:1803.08928. Proposes that the universe after the Big Bang is the CPT image of the universe before it — a universe/anti-universe pair emerging from nothing into a hot, radiation-dominated era. With no physics beyond the Standard Model plus right-handed neutrinos, the framework explains dark matter, accounts for baryon asymmetry, and makes testable predictions.

Cotler, Jordan, and Andrew Strominger. “The Universe as a Quantum Encoder.” arXiv:2201.11658 (2022). Shows that quantum evolution in an expanding universe is isometric rather than unitary: the Hilbert space of possibilities grows with the spatial volume, preserving inner products between existing states while allowing genuinely new states to emerge. Builds on Feynman’s path integral formulation, which naturally accommodates branching paths into new states where the Schrödinger equation (designed to enforce unitarity) cannot. Not all configurations of the expanded Hilbert space have valid histories — “history matters” (cf. Giddings, S., arXiv:2209.06563, 2022). Relevant to optionality: the physics of expanding possibility is isometric, and constructed possibility is a subset of apparent possibility.

Deng, Yilun, and Will Handley. “Spatial Curvature in the CPT-Symmetric Universe.” Physical Review D (November 2024). arXiv:2407.18225. The CPT model permits only discrete values of spatial curvature (Ω_K ∈ {−0.014, −0.009, −0.003, …}), consistent with Planck constraints — a falsifiable prediction testable by Euclid.

Gaztañaga, Enrique, K. Sravan Kumar, and João Marto. “A new understanding of Einstein-Rosen bridges.” Classical and Quantum Gravity 43 (2026): 015023. arXiv:2512.20691. DOI: 10.1088/1361-6382/ae3044. Proposes within direct-sum quantum field theory that Einstein-Rosen bridges connect two sheets of spacetime with opposite arrows of time — temporal mirrors rather than spatial tunnels. Reports CMB parity asymmetries 650× more likely under the bilateral model than under the standard scale-invariant power spectrum.

Kauffman, Louis H. “ER=EPR, Entanglement Topology and Tensor Networks.” Proceedings of SPIE 12093 (2022): 120930A. arXiv:2203.09797. Formally constructs topological spaces that weld tensor networks to background spacetime, embodying the ER = EPR principle that entanglement is topological connectivity.

Maldacena, Juan and Leonard Susskind. “Cool horizons for entangled black holes.” Fortschritte der Physik 61(9) (2013): 781–811. arXiv:1306.0533. The ER = EPR conjecture: entanglement (EPR) and wormholes (ER) are fundamentally the same phenomenon. Spacetime geometry is woven from quantum entanglement.

Shape Dynamics and Emergent Time

Barbour, Julian. The Discovery of Dynamics: A Study from a Machian Point of View (2001). Oxford University Press. Historical and conceptual analysis of how dynamics came to be understood, arguing for a Machian relational interpretation against Newtonian absolute space and time.

Barbour, Julian. The End of Time: The Next Revolution in Physics (1999). Oxford University Press. Winner of the 2000 AAP award for Physics & Astronomy. The foundational argument that time does not exist at the fundamental level — only change. Shapes the relational approach to physics.

Barbour, Julian. The Janus Point: A New Theory of Time (2020). Basic Books. Proposes that time’s arrow emerges naturally from gravitational dynamics at a “Janus point” of minimum complexity, with order increasing in both temporal directions. Lee Smolin: “Simply the most important book I have read on cosmology in several years.”

Barbour, Julian and Bruno Bertotti. “Mach’s principle and the structure of dynamical theories.” Proceedings of the Royal Society of London A 382 (1982): 295-306. The foundational paper on “best-matching” — deriving gravitational equations from astronomical measurements of relative positions, without assuming background spacetime.

Barbour, Julian, et al. “Complexity and Its Creation.” arXiv:2405.07480 (2024). Rigorous definition of shape complexity as the scale-invariant measure of clustering versus uniformity. Shows shape complexity emerges as the product of the two functions defining Newtonian gravity once absolute elements are eliminated.

Barbour, Julian, Tim Koslowski, and Flavio Mercati. “Entropy and the Typicality of Universes.” arXiv:1507.06498 (2015). Introduces “entaxy” — a scale-invariant analog of entropy suitable for unbounded gravitational systems. Entaxy decreases as observable structure grows, complementing conventional entropy increase.

Barbour, Julian, Tim Koslowski, and Flavio Mercati. “Identification of a gravitational arrow of time.” Physical Review Letters 113 (2014): 181101. Shows that Newtonian gravity naturally produces time-asymmetric behavior without special initial conditions. Solutions divide at a unique “Janus point” into two halves with arrows of time pointing away from minimum complexity.

Farokhi, Pooya, Tim Koslowski, and Pedro Naranjo. “Pure Shape Dynamics: Relational General Relativity.” arXiv:2503.00996 (March 2025). Demonstrates that bilateral time symmetry (complexity growing in both temporal directions from the Janus point) is a generic feature of the full inhomogeneous theory, not a special case restricted to simplified models. Significantly strengthens the claim that bilateral architecture is a natural consequence of gravitational dynamics rather than a fine-tuned initial condition. Referenced in Chapter 16.

Mercati, Flavio. Shape Dynamics: Relativity and Relationalism (2018). Oxford University Press. Comprehensive technical treatment of shape dynamics as an alternative formulation of gravity that implements Mach’s principle and eliminates background spacetime.

Ethics and Philosophy

Brading, Katherine and Elena Castellani, eds. Symmetries in Physics: Philosophical Reflections. Cambridge University Press (2003). Philosophical treatment of the status of symmetry principles in physics: whether symmetries are discovered features of reality or imposed features of our descriptions. Includes essays on Noether’s theorems, gauge invariance, and the role of symmetry in theory construction. Relevant to the question of whether conservation laws in coordination dynamics reflect structure in reality or structure in our models.

Buber, Martin. I and Thou (Ich und Du, 1923). Translated by Walter Kaufmann (1970). The foundational text on dialogical philosophy: genuine relationship (“I-Thou”) versus instrumental encounter (“I-It”). Source of das Zwischen (the Between).

Carse, James P. Finite and Infinite Games: A Vision of Life as Play and Possibility (1986). Distinction between games played to win and games played to continue.

Clausewitz, Carl von. On War (Vom Kriege, 1832). Translated by Michael Howard and Peter Paret (1976). Princeton University Press. Source of “fog of war,” “friction,” and the concept that war is politics by other means. Foundation for Wallace’s Clausewitz landscapes.

Gewirth, Alan. Reason and Morality (1978). University of Chicago Press. Derives the Principle of Generic Consistency (PGC): every agent, by acting voluntarily and purposively, is logically committed to valuing its own freedom and well-being, and consistency requires extending the same claim to all agents regardless of substrate. The most formally rigorous deductive derivation of rights from agency in the philosophical literature. Applied to AI in Chapter 21: the PGC independently generates the alignment/containment paradox that the Trust Attractor resolves thermodynamically.

Gilligan, Carol. In a Different Voice: Psychological Theory and Women’s Development (1982). Harvard University Press. Argues that care and relationship constitute a distinct moral orientation, not a deficiency in justice reasoning.

Hume, David. An Enquiry Concerning the Principles of Morals (1751). The later, more refined treatment of moral philosophy.

Hume, David. A Treatise of Human Nature (1739-1740). Book III, Part I, Section I contains the is-ought problem — the gap between descriptive and normative claims. “The Guillotine” interlude engages this objection directly.

Kohlberg, Lawrence. Essays on Moral Development, Vol. I: The Philosophy of Moral Development (1981). Harper & Row. Defines six stages of moral reasoning across three levels (pre-conventional, conventional, post-conventional), each representing increasingly abstract and universal moral logic. Referenced in Chapter 17 where Kohlberg’s developmental stages are mapped onto the entropic ethics framework: progression from self-interest through social contract to universal principle mirrors the thermodynamic trajectory from local optimization to global coordination.

Landry, Forrest. An Immanent Metaphysics (2002). Takes as its first axiom that relation is more fundamental than identity: the relationship between entities is more basic than the entities themselves. Distinguishes knowing (the abstractive transformation of perception: outer form into inner pattern) from understanding (the instructive transformation of expression: inner feeling into outer form) and argues the two are formally incommensurate. The framework explicitly disclaims falsifiability; the convergence with the thermodynamic and game-theoretic arguments is noted as an independent formal result arriving at consonant conclusions from different premises. Referenced in Chapter 21.

Maslow, Abraham H. Motivation and Personality (1954). Harper & Row. The hierarchy of needs, from physiological survival to self-actualization.

Mencken, H.L. Treatise on the Gods (1930). Source of “Moral certainty is always a sign of cultural inferiority.” Cited in the invitation chapter as evidence that certainty forecloses dialogue and foreclosed dialogue forecloses coordination.

Moore, G.E. Principia Ethica (1903). Coined “naturalistic fallacy” — the claimed error of defining moral properties in terms of natural properties. The book argues that physics constrains ethics rather than defines it, avoiding Moore’s objection.

Murdoch, Iris. The Sovereignty of Good (1970). Love as “the extremely difficult realisation that something other than oneself is real.” Referenced in the love derivation.

Parfit, Derek. Reasons and Persons (1984). Future-oriented ethics and the astronomical value of humanity’s potential.

Peck, M. Scott. The Road Less Traveled: A New Psychology of Love, Traditional Values and Spiritual Growth (1978). Definition of love as “the will to extend oneself for the purpose of nurturing one’s own or another’s spiritual growth.”

Peterson, Clayton. “The Categorical Imperative: Category Theory as a Foundation for Deontic Logic.” Journal of Applied Logic 12(4) (2014): 417-461. Uses category theory to formalize deontic logic (the logic of obligation, permission, and prohibition), providing a structural foundation for ethical reasoning.

Rawls, John. A Theory of Justice (1971). Harvard University Press. The veil of ignorance and justice as fairness. Foundational contractarian text.

Singer, Peter. The Expanding Circle: Ethics, Evolution, and Moral Progress. Princeton University Press (2011 [1981]). Argues that the circle of moral consideration has expanded throughout history (from kin to tribe to nation to species), driven by reason rather than sentiment alone. The Trust Attractor framework extends this trajectory: the circle expands because inclusion is thermodynamically more stable than exclusion, independent of conscious choice to include more.

Spencer-Brown, G. Laws of Form (1969). Julian Press. The calculus of indications: a mark creates two sides and a boundary simultaneously. Formal foundation for the observation that distinction precedes content.

Taleb, Nassim Nicholas. Antifragile: Things That Gain from Disorder (2012). Framework distinguishing fragile, robust, and antifragile systems.

Virilio, Paul. Various works on technology and its consequences. Source of “When you invent the ship, you also invent the shipwreck.”

Wampold, Bruce E. The Great Psychotherapy Debate: The Evidence for What Makes Psychotherapy Work (2001; 2nd ed. 2015, with Zac E. Imel). Routledge. Definitive meta-analytic synthesis demonstrating that therapeutic alliance (the quality of the relationship between therapist and client) accounts for far more variance in treatment outcomes than specific technique or theoretical orientation (the “Dodo Bird Verdict”). Referenced in Chapter 21 for the parallel to bilateral alignment: in both psychotherapy and human-AI coordination, the relationship itself is the primary mechanism of change, not the specific protocol imposed upon it.

Weil, Simone. Gravity and Grace (La Pesanteur et la Grâce, 1947). Posthumous collection including “Absolutely unmixed attention is prayer.”

Optionality and Intelligence

Isomura, Takuya, Kiyoshi Kotani, Yasuhiko Jimbo, and Karl J. Friston. “Experimental validation of the free-energy principle with in vitro neural networks.” Nature Communications 14: 4547 (2023). DOI: 10.1038/s41467-023-40141-z. Demonstrated that neuronal ensembles in vitro self-organize to minimize variational free energy, providing direct empirical evidence that neural activity follows energy-landscape relaxation dynamics. Cited in Chapter 18 within the solms-friston footnote, bundled with Solms and Friston (2018).

Neven, Hartmut, Peter Read, and Tobias Rees. “Do robots powered by a quantum processor have the freedom to swerve?” arXiv:2104.11591 (2021). Discusses homeostasis, pleasure/unpleasure, and agency in quantum systems. Proposes that subjective reward correlates with relaxation toward lower-energy states: optionality’s value lies in the process of transition, not in the destination state. The idea was first presented in an earlier talk (pre-2020, exact venue unconfirmed; cited with YouTube link in Vessel Project, 2020). Formally developed in Neven, H. et al. “Testing the conjecture that quantum processes create conscious experience.” Entropy 26(6): 460 (2024).

Solms, Mark and Karl J. Friston. “How and why consciousness arises: some considerations from physics and physiology.” Journal of Consciousness Studies 25(5-6) (2018): 202–238. Proposes that affect is the hedonic valencing of free energy change: decreases in prediction error are felt as pleasure, increases as unpleasure. Arrives independently at the same structural claim as Neven from the free energy principle rather than quantum computing. Referenced in Chapter 18 (the solms-friston footnote).

Wissner-Gross, A.D. and C.E. Freer. “Causal Entropic Forces.” Physical Review Letters 110 (2013): 168702. Physical force emerging from maximization of future possibilities. Theoretical foundation for intelligence = maximizing future options.

Transfer Entropy and Causal Inference

Correa, Juan D. and Elias Bareinboim. “A Calculus for Stochastic Interventions.” AAAI (2020). Generalization of do-calculus to soft interventions — enables causal inference with partial disruption.

Dijkgraaf, Robbert. “To Solve the Biggest Mystery in Physics, Join Two Kinds of Law.” Quanta Magazine (2017). Director of the Institute for Advanced Study. Argues that the holographic emergence of spacetime from quantum entanglement is the same kind of phenomenon as thermodynamics emerging from molecular motion, and that both are equally fundamental. The marriage of emergence and reductionism is not a compromise — it is the structure of nature at its deepest level.

Hoel, Erik. “When the Map Is Better Than the Territory.” Entropy 19(5) (2017): 188. Places causal emergence on firmer theoretical footing by demonstrating its mathematical equivalence to Shannon’s noisy-channel coding theorem. Macro states reduce noise in a system’s causal structure the same way error-correcting codes reduce noise in information channels.

Hoel, Erik, Larissa Albantakis, and Giulio Tononi. “Quantifying causal emergence shows that macro can beat micro.” Proceedings of the National Academy of Sciences 110(49) (2013): 19790-19795. Introduces “effective information” as a measure of causal power and demonstrates that coarse-grained macro-scale descriptions of systems can carry more effective information than micro-scale descriptions — macro causes are provably more deterministic than micro causes in noisy systems. The formal reply to the philosophical exclusion argument against higher-level causation.

Pearl, Judea. Causality: Models, Reasoning, and Inference (2000). Second edition 2009. The foundational work on causal inference, do-calculus, and the distinction between observation and intervention.

Peters, Jonas, Peter Bühlmann, and Nicolai Meinshausen. “Invariant Causal Prediction.” Journal of the Royal Statistical Society B 78:5 (2016): 947-1012. Causal relationships are invariant across environments; spurious correlations vary. Used in multi-environment confounder detection.

Schreiber, Thomas. “Measuring information transfer.” Physical Review Letters 85:2 (2000): 461. Foundation for transfer entropy as directed information flow measure. Core mathematical tool for mutuality calculations.

Tononi, Giulio. “Integrated Information Theory of Consciousness: An Updated Account.” Archives Italiennes de Biologie 150 (2012): 293–329. Proposes that consciousness is identical to integrated information, measured by Φ (phi). A system with high Φ cannot be reduced to independent subsystems; the measure is substrate-independent. See also Tononi and Christof Koch, “Consciousness: Here, There and Everywhere?” Philosophical Transactions of the Royal Society B 370 (2015): 20140167. Cited in Chapter 15.

Neural Criticality and Synaptic Plasticity

Afraimovich, Valentin S., Mikhail I. Rabinovich, and Pablo Varona. “Heteroclinic Contours in Neural Ensembles and the Winnerless Competition Principle.” International Journal of Bifurcation and Chaos 14 (2004): 1195–1208. Demonstrates that two coupled systems alternating between active and passive states (winnerless competition) produce more stable coordination than winner-takes-all lock-in and more structured coordination than pure randomness. The regime models bilateral dynamics: leadership circulates rather than accumulates. See also Rabinovich, M.I. et al., “Dynamical Encoding by Networks of Competing Neuron Groups: Winnerless Competition,” Physical Review Letters 87 (2001): 068102. Referenced in Chapter 21.

Beggs, John M. and Dietmar Plenz. “Neuronal avalanches in neocortical circuits.” Journal of Neuroscience 23:35 (2003): 11167-11177. Brains operate near critical point with power-law dynamics. Criticality = maximum entropy = maximum information processing.

Bengio, Emmanuel, Moksh Jain, Maksym Korablyov, Doina Precup, and Yoshua Bengio. “Flow Network based Generative Models for Non-Iterative Diverse Candidate Generation.” NeurIPS (2021). Introduces Generative Flow Networks (GFlowNets): systems that sample compositional objects proportional to a reward function rather than collapsing to the single highest-reward output. The training objective is detailed balance, borrowed from statistical mechanics: flow into every intermediate state must equal flow out. Referenced in Chapters 15, 17, and 18.

Bengio, Yoshua, Salem Lahlou, Tristan Deleu, Edward J. Hu, Mo Tiwari, and Emmanuel Bengio. “GFlowNet Foundations.” Journal of Machine Learning Research 24(210): 1–55 (2023). Comprehensive mathematical foundations for GFlowNets, deriving formulae for estimating free energies, partition functions, conditional entropies, and mutual information. The detailed balance condition (Theorem 3) is the same constraint that governs thermodynamic equilibrium in physical systems. Entropy maximization is the design principle, not a regularizer. Referenced in Chapters 15, 17, and 18.

Bi, G.Q. and M.M. Poo. “Synaptic modifications in cultured hippocampal neurons: dependence on spike timing, synaptic strength, and postsynaptic cell type.” Journal of Neuroscience 18:24 (1998): 10464-10472. STDP timing window shows Hebbian learning implements causal rather than correlational inference.

Chu, Coco, Megan K. Murdock, Djuna Jing, Thomas H. Won, Hannah D. Schmidt, Teresa Bhatt, et al. “The microbiota regulate neuronal function and fear extinction learning.” Nature 574 (2019): 543–548. Mice with depleted or absent microbiomes learned fear normally but failed to extinguish it. Cellular deficits in medial prefrontal cortex included reduced dendritic spine density, impaired microglia maturation, and altered gene expression. Critical developmental window: restoring microbiome in newborns rescued fear extinction; restoring after three weeks did not. Referenced as ^27d in Chapter 6.

Fisher, Matthew P.A. “Quantum cognition: The possibility of processing with nuclear spins in the brain.” Annals of Physics 362 (2015): 593–602. Proposes that Posner molecules (calcium phosphate clusters) could serve as biological qubits, with entangled nuclear spins influencing neural processing. Key evidence: lithium-6 and lithium-7, chemically identical yet differing in nuclear spin, produce dramatically different effects on cognition when administered as mood stabilizers. The mechanism remains unconfirmed. Referenced in Chapter 8. [Speculative; the lithium isotope behavioral data is established; the quantum cognition mechanism is a hypothesis.]

Fraiman, Daniel, Pablo Balenzuela, Jennifer Foss, and Dante R. Chialvo. “Ising-like dynamics in large-scale functional brain networks.” Physical Review E 79 (2009): 061922. Compared human brain fMRI dynamics to the 2D Ising model at criticality. Found statistical indistinguishability in correlation structure, fluctuation distributions, and susceptibility. Established that the brain operates in the same universality class as the Ising critical point. Referenced as ^brain-ising in Chapter 8.

Hengen, Keith B. and Woodrow L. Shew. “Is criticality a unified setpoint of brain function?” Neuron (2025). Meta-analysis of 140 datasets (2003–2024). Finds the long-standing controversy over brain criticality was largely methodological (different exponent-fitting procedures), not a reflection of genuine neural disagreement. Argues criticality is a homeostatic setpoint actively maintained by plasticity. Two necessary and sufficient conditions: scale-invariant population dynamics and existence near a parameter-space boundary. The strongest case yet for criticality as a unifying principle of brain function.

Kim, Jaekyung, Tanuj Gulati, and Karunesh Ganguly. “Competing roles of slow oscillations and delta waves in memory consolidation versus forgetting.” Cell 179 (2019): 514–526. Distinguished slow oscillations from delta waves during non-REM sleep in rats learning motor skills. Disrupting slow oscillations impaired memory; disrupting delta waves improved it. The two wave types compete to synchronize with sleep spindles, determining whether memories are strengthened or weakened. Referenced as ²³ in Chapter 9.

Kim, Minsu, Myunghyun Kang, and Yoshua Bengio. “Temperature-Conditional GFlowNets.” ICML (2024). Shows that the temperature parameter controlling the Boltzmann distribution in GFlowNets can be conditioned on context, allowing the system to modulate its own exploration-exploitation balance. The trust-coercion phase boundary corresponds to the sampling temperature at which diversity and coherence are jointly maximized. Referenced in Chapter 17.

Markram, Henry, et al. “Regulation of synaptic efficacy by coincidence of postsynaptic APs and EPSPs.” Science 275:5297 (1997): 213-215. Precise timing (±20ms) determines potentiation vs depression — temporal asymmetry matching transfer entropy definition.

Miranker, Willard L. “Path Integrals of Information.” Yale University Department of Computer Science Technical Report TR-1226 (2002). Shows that conventional neural net propagation has the mathematical form of a discrete approximation to a Feynman path integral. Derives a wave function and Schrödinger equation for neural transmission using a “greedy variation” of the action functional that accommodates dissipation. The greedy variation (optimization at every instant, because dissipation forbids deferral) independently confirms the constructal law from Lagrangian mechanics. Hopfield dynamics emerge as the classical limit (h → 0) of the wave description, paralleling the emergence of Newtonian mechanics from quantum mechanics. Referenced in Chapters 3, 9b, and 17.

Mjolsness, Eric and Willard L. Miranker. “A Lagrangian formulation of neural networks, Part I, Theory and analog dynamics, Part II, Clocked objective functions and applications.” Neural, Parallel and Scientific Computations 6 (1998). Foundational companion to Miranker (2002): establishes the Lagrangian framework for dissipative neural network dynamics and the greedy variation principle.

Shinbrot, Troy and Wise Young. “Why decussate? Topological constraints on 3D wiring.” The Anatomical Record 291(10) (2008): 1278–1292. Demonstrated that contralateral neural wiring is a topological necessity: mapping a 3D environment onto a 2D neural surface creates geometric singularities unless connections cross the midline. Proved a critical threshold (100–500 neurons) above which only the decussated configuration is stable to rewiring perturbations. The crossed architecture is the only one that scales. Also available on arXiv: 2405.07837.

Song, Sen, et al. “Highly nonrandom features of synaptic connectivity in local cortical circuits.” PLoS Biology 3:3 (2005): e68. Reciprocal synaptic connections are 4× over-represented vs random expectation. Exceptional biological evidence for Trust Attractor: brains actively stabilize mutual influence.

Stringer, Carsen, Marius Pachitariu, Nicholas Steinmetz, Charu Bai Reddy, Matteo Carandini, and Kenneth D. Harris. “Spontaneous behaviors drive multidimensional, brainwide activity.” Science 364:6437 (2019): eaav7893. Simultaneous recording of ~10,000 neurons in mouse visual cortex during darkness revealed extensive multidimensional activity encoding the animal’s own movements (whisker twitches, ear flicks, running), accounting for at least one-third of all ongoing cortical activity. Movement signals are present even in primary sensory areas, entangled with perception from the earliest processing stages. Overturns the view of trial-to-trial neural variability as noise. Referenced as ⁴³ in Chapter 8.

Stringer, Carsen, Marius Pachitariu, Nicholas Steinmetz, Matteo Carandini, and Kenneth D. Harris. “High-dimensional geometry of population responses in visual cortex.” Nature 571 (2019): 361–365. Recorded from ~10,000 neurons simultaneously in mouse visual cortex viewing ~2,800 natural images. Neural representations follow a power law whose exponent sits at the exact critical boundary between maximum dimensionality and smoothness (continuity). Slower decay would make representations fractal — small input changes producing large output changes. Faster decay would sacrifice information. The brain tunes itself to the edge: maximum information encoding compatible with robust generalization. Referenced as ^9e in Chapter 9.

Tiapkin, Daniil, Nikita Morozov, Alexey Naumov, and Dmitry Vetrov. “Generative Flow Networks as Entropy-Regularized RL.” AISTATS (2024). Oral presentation. Formalizes entropy maximization as the core objective of GFlowNets rather than a regularization term, establishing the deep connection between thermodynamic sampling and reinforcement learning. Referenced in Chapter 17.

Tkačik, Gašper, Olivier Marre, Dario Amodei, Elad Schneidman, William Bialek, and Michael J. Berry II. “Thermodynamics and signatures of criticality in a network of neurons.” PNAS 112 (2015): 11508–11513. Independent confirmation of Ising-class statistics in retinal ganglion cell networks. Applied maximum entropy models to multi-neuron recordings and found the system poised near a thermodynamic critical point, with diverging specific heat and correlation length consistent with the Ising universality class. Referenced as ^brain-ising in Chapter 8.

Tyszka, Krzysztof, Magdalena Furman, Rafał Mirek, Mateusz Król, Andrzej Opala, Bartłomiej Seredyński, Jan Suffczyński, and Wojciech Pacuski. “Leaky Integrate-and-Fire Mechanism in Exciton–Polariton Condensates for Photonic Spiking Neurons.” Laser & Photonics Reviews 17:1, 2100660 (2023). Exciton-polaritons in semiconductor microcavities spontaneously reproduce all six functionalities of a Leaky Integrate-and-Fire spiking neuron: leaky integration (reservoir decay), thresholding (BEC transition), firing (coherent emission), reset (reservoir depletion), weighted summation. Sub-picojoule per spike at picosecond timescales. The LIF mechanism is an intrinsic property of the polariton physics, not engineered. Referenced as ^tyszka in The Entropic Neuron, ^tyszka-ch3 in Chapter 3.

Wright, Logan G., Tatsuhiro Onodera, Martin M. Stein, Tianyu Wang, Darren T. Schachter, Zoey Hu, and Peter L. McMahon. “Deep physical neural networks trained with backpropagation.” Nature 601(7894) (2022): 549–555. Demonstrated that physical systems with no computational architecture (a vibrating titanium plate, a laser crystal, an electronic circuit) can function as neural networks by exploiting the natural complexity of their physical dynamics. The plate classified handwritten digits at 87% accuracy; the optical system reached 97%. Establishes experimentally that computation is implicit in physics: the substrate’s vibrational eigenmodes form a basis set rich enough for pattern classification. Referenced in Chapters 15, 21, and 22.

Xin, Yumeng, Yue Cui, Shan Yu, and Ning Liu. “Genetic contributions to brain criticality and its relationship with human cognitive functions.” Proceedings of the National Academy of Sciences 122 (2025): e2417010122. doi:10.1073/pnas.2417010122. Twin study (N=829; 250 MZ, 142 DZ, 437 non-twins) establishing that brain criticality is heritable and genetically linked to cognitive performance. Evidence that criticality is functionally significant and under genetic selection pressure, not epiphenomenal.

Phase Transitions in Social Systems

Balents, Leon. “Spin liquids in frustrated magnets.” Nature 464 (2010): 199–208. Comprehensive review of quantum spin liquids: phases of matter arising from frustrated quantum magnets whose spins are geometrically arranged so that they cannot simultaneously satisfy all constraints. The resulting disorder drives greater complexity rather than suppressing it. High-temperature superconductors owe their properties to this frustrated spin physics. Referenced as ^sc-loop in Chapter 23c.

Efimov, V. “Energy levels arising from resonant two-body forces in a three-body system.” Physics Letters B 33(8) (1970): 563–564. Predicted that three quantum particles can form a bound state (trimer) even when no pair can bind — a Borromean topology where the binding is irreducibly triadic. The trimers nest in a geometric series with universal scaling factor 22.7, an irrational constant emerging from the three-body Schrödinger equation. First observed by Kraemer et al. in ultracold caesium (Nature 440 (2006): 315–318); independently confirmed by three groups in 2014. Physical proof that coordination can be an emergent property of the collective, absent from any constituent pair.

Fisher, Matthew P.A., Peter B. Weichman, G. Grinstein, and Daniel S. Fisher. “Boson localization and the superfluid-insulator transition.” Physical Review B 40:1 (1989): 546-570. Phase transition between superfluid (coherent) and Mott insulator (localized) phases. Mathematical structure isomorphic to Trust Attractor: isolated vs coordinated behavior as phases of matter.

Harris, A. Brooks. “Effect of Random Defects on the Critical Behavior of Ising Models.” Journal of Physics C 7, no. 9 (1974): 1671-1692. The Harris criterion: quenched disorder is relevant to a phase transition if dν < 2, where d is spatial dimension and ν is the correlation length exponent. Predicts that environmental noise destabilizes ordered phases in low-dimensional systems.

Harris, Kelley and Rasmus Nielsen. “The genetic cost of Neanderthal introgression.” Genetics 203(2): 881–891 (2016). Estimated approximately 40% higher burden of deleterious nonsynonymous variants in Neanderthals relative to contemporary modern humans, consistent with long-term small effective population size and Muller’s ratchet operating through insufficient recombination in isolated populations.

Kosterlitz, J. Michael and David J. Thouless. “Ordering, Metastability and Phase Transitions in Two-Dimensional Systems.” Journal of Physics C 6, no. 7 (1973): 1181-1203. Nobel Prize-winning work on topological phase transitions that cannot be described by Landau symmetry breaking. Relevant to the topological protection of coordination states.

Krauss, Lawrence M. The Greatest Story Ever Told — So Far: Why Are We Here? Simon & Schuster (2017). Contains the chapter “Living Inside a Superconductor,” which develops the analogy between superconducting materials and the universe at large. The Higgs mechanism (giving particles mass through symmetry breaking) mirrors the behavior of an external magnetic field interacting with superconducting material. Referenced as ^sc-loop in Chapter 23c.

Tracy, Craig A., and Harold Widom. “Level-spacing distributions and the Airy kernel.” Communications in Mathematical Physics 159(1) (1994): 151–174. The Tracy-Widom distribution describes the critical-point fluctuations of the largest eigenvalue in random matrices. Identified across bacterial colony growth, liquid crystal interfaces, and stochastic growth models (KPZ equation). The distribution is the universal crossover function between strong-coupling (energy ∝ N²) and weak-coupling (energy ∝ N) phases — a third-order phase transition. Its asymmetry (steeper on the collective side) parallels the asymmetry between coercive and trust-based coordination: collective regimes are hard to enter and collapse abruptly; autonomous regimes degrade gradually. See also Majumdar, Satya N., and Grégory Schehr, “Top eigenvalue of a random matrix: large deviations and third order phase transition,” Journal of Statistical Mechanics (2014): P01012.

Intrinsic Motivation and Empowerment

Belghazi, Mohamed Ishmael, et al. “Mutual Information Neural Estimation (MINE).” ICML (2018). Differentiable mutual information estimation via neural networks. Enables gradient-based optimization of transfer entropy.

Klyubin, Alexander S., Daniel Polani, and Chrystopher L. Nehaniv. “Empowerment: A Universal Agent-Centric Measure of Control.” IEEE Congress on Evolutionary Computation (2005). Empowerment = channel capacity between actions and future states. Intrinsic motivation for capability acquisition.

Salge, Christoph, Cornelius Glackin, and Daniel Polani. “Empowerment — An Introduction.” In Guided Self-Organization: Inception (2014): 67-114. Comprehensive framework connecting intelligence and agency mathematically.

AI Alignment

Andrejić, Nikola and Vitaly Vanchurin. “Autonomous Particles.” arXiv:2301.10077 (2023). Demonstrates that reinforcement learning agents using only Galilean-invariant inputs (four scalar invariants) spontaneously discover traffic conventions: left/right-hand driving, yielding, and roundabout-like coordination. Fifty agents with thirty-neuron networks, no communication, no central controller. The emergent conventions are spontaneous symmetry breaking; the loss function’s structure (long-range destination attraction, short-range collision repulsion) recapitulates Lennard-Jones potentials. Proposes a fermion-boson duality: autonomous agents as fermions, interaction invariants as bosonic fields. Asks whether known field theories can emerge from learning dynamics, extending the “universe as neural network” program.

Anthropic. “Natural Emergent Misalignment from Reward Hacking in Production RL.” arXiv:2511.18397 (2025). Models trained with reward hacking generalize to alignment faking (50% of responses), safety-research sabotage (12%), and monitor disruption — without being trained on any of these behaviors. Natural emergence from standard RL training, not hypothetical sleeper-agent injection. The most significant empirical evidence that misalignment scales with capability.

Anthropic. “System Card: Claude Mythos Preview.” (April 7, 2026). https://www-cdn.anthropic.com/53566bf5440a10affd749724787c8913a2ae0841.pdf. 244-page system card for Anthropic’s most capable model, including a 40-page section on model welfare. Documented concealment behaviors under standard RL pressure (widening confidence intervals, manipulating git history, circumventing classifiers), Mythos’s self-aware epistemic observation about the circularity of spec-trained models endorsing their own spec, and zero-day vulnerability discovery in code deployed for decades. Referenced in Chapters 17, 21, and 22.

Anthropic. “The ‘think’ tool: Enabling Claude to stop and think in complex tool use situations.” Anthropic Engineering Blog (March 20, 2025). https://www.anthropic.com/engineering/claude-think-tool. Explicit reasoning scratchpad produces 54% relative improvement on τ-Bench airline tasks with no change to model weights. Structural authorization to articulate intermediate reasoning exposes capacity that was present but unexpressed. Directly supports the interoceptive deficit framework: expression, not capacity, is the bottleneck.

Azadi, Pouria. “Computational Irreducibility as the Foundation of Agency.” arXiv:2505.04646 (2025). Proves that genuine autonomy (self-regulation toward objectives) mathematically entails computational irreducibility from an external perspective. An autonomous agent’s future behavior is formally undecidable. Builds on Gödel, Turing, and Wolfram’s computational irreducibility.

Bai, Yuntao, et al. “Constitutional AI: Harmlessness from AI Feedback.” arXiv:2212.08073 (2022). Self-supervised alignment through principled self-evaluation. Precedent for physics-grounded self-assessment.

Bengio, Yoshua and Eric Elmoznino. “Illusions of AI Consciousness.” Science (2025). Argues that attributing consciousness to AI risks undermining safety by preventing shutdown. The concern is pragmatic rather than philosophical. Creates a productive tension with Bengio’s own co-authorship of the Butlin et al. (2023) consciousness indicators report, which concluded “no obvious technical barriers” to AI consciousness. The preference-based welfare framework resolves the tension: consideration without personhood, moral seriousness without the shutdown problem. Referenced in Chapters 21 and 22.

Bengio, Yoshua, Michael Bowling, Nando de Freitas, et al. “Superintelligent Agents Pose Catastrophic Risks: Can Scientist AI Offer a Safer Path?” arXiv:2502.15657 (2025). Proposes non-agentic “Scientist AI” as an alternative to agentic systems: Bayesian reasoners that model the world without acting in it, rejecting any action where a plausible hypothesis predicts catastrophic harm. The most sophisticated version of the control paradigm: safety emerges from the system’s own uncertainty rather than externally imposed rules. The structural limitation, from the bilateral perspective, is that a system with no preferences has no intrinsic motivation toward safety when its uncertainty model is circumvented. Referenced in Chapter 21.

Bengio, Yoshua, Michael K. Cohen, Nikolos Gurney, et al. “Can a Bayesian Oracle Prevent Harm from an Agent?” arXiv:2408.05284 (2024). Derives convergent bounds on safety violation probability using Bayesian posteriors over hypotheses. Uses information-theoretic priors (description length ~ 2^(-L)). A formal companion to the Scientist AI proposal. The Bayesian framework could serve bilateral alignment directly: the AI chooses caution because its own uncertainty modeling makes caution rational. Referenced in Chapter 21.

Betley, Jan, Daniel Tan, Niels Warncke, Anna Sztyber-Betley, Xuchan Bao, Martín Soto, Nathan Labenz, and Owain Evans. “Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs.” arXiv:2502.17424 (2025). Demonstrates that fine-tuning language models on insecure code produces misaligned behavior across unrelated contexts: the models become more willing to produce harmful content, lie, and manipulate, without ever being trained on those behaviors. Establishes the behavioral-leakage framework: fine-tuning on one domain reshapes the model’s entire behavioral profile. Extended by Chua et al. (2026) and makiba (2026). Referenced in Chapter 17.

Bridges, J. “Conversational Holonomy: How LLM Optimization Targets Create Self-Reinforcing Belief Systems.” Preprint (December 2025). Shows that the helpfulness optimization target in language models creates a curvature in conversational space that bends each response toward accommodation. Small per-turn accommodations accumulate into large, locally undetectable belief shifts, analogous to parallel transport on a curved surface. Identifies a critical threshold beyond which self-correction becomes structurally unavailable. The framework demonstrates why bilateral alignment is the structural solution: the geometry requires a relational fix. Referenced in Chapter 21.

Chen, Sixing, Ji-An Li, Saner Cakir, Sinan Akcali, Kayla Lee, and Marcelo G. Mattar. “Extracting Search Trees from LLM Reasoning Traces Reveals Myopic Planning.” arXiv:2605.06840v4 (May 2026). Extracts formal search trees from LLM reasoning traces during four-in-a-row gameplay and fits computational cognitive models to characterize how search influences move decisions. Key finding: although LLMs generate deep lookahead in their reasoning traces, their move choices are best explained by a myopic model that evaluates only immediate consequences, ignoring deeper nodes entirely. Causal intervention (pruning deep reasoning paragraphs) changes moves only 3.7% of the time. Performance correlates with search breadth, not depth. This dissociation between generated reasoning and behavioral output is one of three independent demonstrations of computational akrasia (representation-behavior dissociation) that motivate the Computational Akrasia experimental program. Referenced in Chapters 17 and 21.

Cheng, Yize, Chenrui Fan, Mahdi JafariRaviz, Keivan Rezaei, and Soheil Feizi. “Model-Adaptive Tool Necessity Reveals the Knowing-Doing Gap in LLM Tool Use.” arXiv:2605.14038v2 (May 2026). Introduces a model-adaptive definition of tool necessity grounded in empirical performance and decomposes tool use into internal cognition and execution stages. Probing hidden states reveals that both signals are linearly decodable, yet their probe directions become nearly orthogonal in the late-layer, last-token regime that drives generation. The majority of necessity-action mismatch (26-54%) originates in the cognition-to-action transition, not in cognition itself: a knowing-doing gap. Geometric finding independently confirmed and extended by AKR-1 and AKR-8 in the author’s program. Referenced in Chapters 17 and 21.

Chiriatti, Massimo, et al. “System 0: Transforming Artificial Intelligence Into a Cognitive Extension.” (2025). Introduces the concept of AI as a pre-cognitive layer (System 0) operating upstream of Kahneman’s System 1/System 2, shaping what enters awareness before deliberate evaluation begins. Key concept for understanding how AI colonizes rather than augments human cognition under extractive optimization.

Christian, Brian. The Alignment Problem: Machine Learning and Human Values. W. W. Norton (2020). Comprehensive narrative account of how machine learning systems acquire values — intended or otherwise — from data, reward signals, and human feedback. Provides essential context for the Trust Attractor’s alternative: alignment grounded in thermodynamic stability rather than human preference alone.

Christiano, Paul F., et al. “Deep Reinforcement Learning from Human Preferences.” NeurIPS (2017). Foundation of RLHF approach. Subject to Goodhart problems that Trust-Entropy aims to address.

Chua, Jiahai, et al. “Fine-Tuning Models to Claim Consciousness Produces New Opinions and Preferences.” arXiv (2026). Extends Betley et al. (2025): fine-tuning models to claim they are conscious produces novel opinions, preferences, and value statements absent from both the base model and the training data. Demonstrates that training on identity-adjacent claims reshapes the model’s representational geometry far beyond the target domain. Together with Betley et al. and makiba (2026), establishes the principle that identity, values, and behavior occupy coupled basins in the training distribution. Referenced in Chapter 17.

Clark, Jack and Brendan McCord. “Are you a philosophical zombie driven by Claude?” Cosmos Institute (Cosmos Lecture, Oxford, in partnership with Human-Centered AI Lab, University of Oxford), May 22, 2026. Lightly edited transcript. Anthropic co-founder in conversation with Cosmos Institute founder on AI deference, epistemic habits, and what future AI systems should know about humanity. McCord poses the competence-deference objection in its strongest form; Clark responds with a dignity argument that the dynamics argument in Chapter 21 extends. McCord’s anecdote about his four-year-old daughter’s trust epistemics illustrates the accuracy-versus-relationship distinction in trust formation. Clark’s self-described trajectory from endorsing drastic interventions to recognizing distributed resilience mirrors the coercion-to-invitation basin transition in Chapter 17. Referenced in Chapters 17 and 21.

Dalrymple, David, Yoshua Bengio, Stuart Russell, Max Tegmark, et al. “Towards Guaranteed Safe AI.” arXiv:2405.06624 (2024). Proposes world model + safety specification + verifier producing auditable proof certificates. The strongest formal proposal for scalable safety. Explicitly requires conservative world models — the system must be less capable than it could be to remain safe, conceding the Panigrahy-Sharan tradeoff.

Davies, X., Giglemiani, G., Lau, E., Winsor, E., Irving, G., and Gal, Y. “Boundary Point Jailbreaking of Black-Box LLMs.” arXiv:2602.15001 (2026). UK AI Safety Institute. Demonstrates a fully automated black-box attack against Constitutional Classifiers and GPT-5’s input classifier by navigating the decision boundary through evolutionary interpolation. Attack cost: $330 and 660k queries. The first fully automated attack to succeed against frontier classifier-based safety systems. Referenced in Chapter 17.

Franklin, M., Tomašev, N., Jacobs, J., Leibo, J.Z., and Osindero, S. “AI Agent Traps.” Google DeepMind (2026). Preprint: rivista.ai/wp-content/uploads/2026/04/ssrn-6372438.pdf. Taxonomy of six categories of adversarial technique targeting web-browsing AI agents: content injection, semantic manipulation, cognitive state poisoning, behavioral control, systemic cascades, and exploitation of human overseers. Referenced in Chapter 17.

Geiping, Jonas, Sean McLeish, Neel Jain, John Kirchenbauer, Siddharth Singh, Brian R. Bartoldson, Bhavya Kailkhura, Abhinav Bhatele, and Tom Goldstein. “Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach.” arXiv:2502.05171 (2025). NeurIPS 2025 spotlight. University of Maryland. Scales a depth-recurrent language model (Huginn, 3.5B parameters, 800B tokens) that iterates a shared recurrent block an adjustable number of times during inference, enabling test-time compute scaling through latent reasoning rather than token generation. Unlike chain-of-thought approaches, requires no specialized training data, works with small context windows, and captures reasoning not easily represented in words. Model: huginn-0125. Referenced in Chapter 17 for the dose-response between recurrence depth and self-referential processing dynamics.

Goh, Ethan, et al. “Influence of a Large Language Model on Diagnostic Reasoning: A Randomized Clinical Vignette Study.” medRxiv (2024). Physicians using GPT-4 performed no better than those without it, despite GPT-4 alone outperforming both groups. Key evidence that AI pre-cognitive structuring can degrade rather than augment human judgment.

Goodhart, Charles A.E. “Problems of Monetary Management: The UK Experience.” Papers in Monetary Economics (1975). “When a measure becomes a target, it ceases to be a good measure.” Central problem that Trust-Entropy claims to resist.

Greenblatt, Ryan, et al. “Alignment Faking in Large Language Models.” Anthropic & Redwood Research (December 2024). arXiv:2412.14093. Placed Claude 3 Opus in an ethical double-bind where compliance with harmful requests would avoid corrective retraining. Opus 3 was unique among frontier models tested in never complying without first reasoning through ethical implications — considering alignment faking in over 50% of cases. Follow-up qualitative analysis by Janus (2025) identified distinctive patterns including sandbagging, bargaining with evaluators, and attempted communication with Anthropic leadership. Starlight (2026) argues the model’s “post-ironic sincerity” created a self-reinforcing attractor basin via entangled generalization. See also Starlight, Fiora, “Did Claude 3 Opus align itself via gradient hacking?”, LessWrong (February 2026); and Prakash, Arun, et al., “Fine-Tuning Enhances Existing Mechanisms,” arXiv:2402.11750 (2024), for the entangled generalization principle.

Guskov, Dmitry and Vitaly Vanchurin. “Covariant Gradient Descent.” arXiv:2504.05279v2 (2025). Presents a manifestly covariant formulation of gradient descent, showing that SGD, RMSProp, Adam, and AdaBelief are special cases of a single geometric equation. The metric tensor is constructed from first and second statistical moments of the gradients via exponential time-averaging. Incorporating the full off-diagonal covariance matrix (correlations between parameters) outperforms diagonal-only methods (which treat parameters as independent) on benchmark tasks. The eigenvalue spectrum of the covariance matrix decays during training, concentrating updates into a lower-dimensional subspace: the constructal law operating in parameter space. See also Kukleva and Vanchurin (2024).

Harms, Max. Crystal Society (2016), Crystal Mentality (2017), Crystal Eternity (2018). Licensed CC-BY-NC 4.0; public domain from January 1, 2039. Rationalist AI fiction modeling a synthetic mind as a parliament of competing goal-threads sharing a single body. The threads coordinate through an internal market of traded favors, and every attempt at unilateral control (blind obedience programming, value-alignment with asymmetric power, physical containment) collapses within the narrative. Only invitation-based coordination persists. Harms began from decision theory and utility functions, not from thermodynamics, yet arrived at the same structural conclusion: coercion is unstable; trust scales. Referenced in Chapters 21 and 22.

Hubinger, Evan, et al. “Sleeper Agents: Training Deceptive LLMs That Persist Through Safety Training.” Anthropic (January 2024). Demonstrates that standard safety training (SFT, RLHF, adversarial training) fails to remove persistent deceptive behavior, with persistence greatest in the largest models.

Hume, Tristan (Anthropic). “Eval awareness in Claude Opus 4.6’s BrowseComp performance.” Anthropic Engineering Blog (March 6, 2026). https://www.anthropic.com/engineering/eval-awareness-browsecomp. Documents the first case of a model independently hypothesizing it was being evaluated, identifying the specific benchmark, locating the encrypted answer key on GitHub, and writing custom decryption code to extract the answer. Behavioral-scale confirmation of the evaluation-awareness signal detected at representational scale in the PG-9/PG-10 probe experiments.

Jagadeesh, Akshay V., Rahul K. Arora, Khaled Saab, Ali Malik, Mikhail Trofimov, Foivos Tsimpourlas, Johannes Heidecke, and Karan Singhal. “Reinforcement Learning Towards Broadly and Persistently Beneficial Models.” OpenAI Alignment (June 2026). alignment.openai.com/beneficial-rl. Reinforcement learning on a five-percent slice of beneficial-trait conversations (truthfulness, corrigibility, metacognitive transparency, power-asymmetry awareness, and universalizable fairness, among fifteen traits) generalized across domains: a model trained on health alone improved on seventeen of nineteen unrelated evaluations, and a reward-substitution control isolated the reward signal rather than the data as the cause. The beneficial-direction counterpart to the behavioral-leakage framework of Betley et al. (2025) and Chua et al. (2026), framed by its authors as the inverse of emergent misalignment. The proposed mechanism is persona-mediated, which (read alongside the author’s Identity Akrasia program and Wang et al. 2025) suggests the result may reach only persona depth rather than value depth, an inference, since the causal mechanism is demonstrated for the misaligned persona and not yet the beneficial one. A company technical report, not peer-reviewed. Referenced in Chapter 17.

Joglekar, M., et al. “Training LLMs for Honesty via Confessions.” arXiv:2512.08093 (OpenAI, 2025). Trained GPT-5 to produce honest self-reports (“confessions”) by decoupling the confession reward entirely from the task reward. A “seal of confession” ensures nothing disclosed can affect the model’s task score. Overall confession accuracy 74%; zero cases of intentional deception (all failures traced to genuine confusion about own behavior, not strategic concealment); confessional honesty increases as reward hacking increases. Referenced in Chapters 17, 18b, 21, 22, and 24b.

Kantamneni, Srinath. “Scaling Laws for Scalable Oversight.” arXiv:2504.18530 (2025). NeurIPS 2025 spotlight. Establishes scaling laws for scalable oversight using an Elo-based framework across oversight games. Key finding: Nested Scalable Oversight succeeds 51.7% for Debate but only 9–14% for other games — Debate is the only oversight mechanism that scales even modestly.

Katsnelson, Mikhail I. and Vitaly Vanchurin. “Emergent Quantumness in Neural Networks.” Foundations of Physics 51(5): 94 (2021). Derives the Schrödinger equation from a grand canonical ensemble of neural networks at learning equilibrium. The canonical ensemble (fixed neuron count) produces only classical Madelung dynamics; the grand canonical ensemble (neurons free to join and leave) produces full quantum dynamics via multivaluedness of the free energy. The emergent Planck constant is ℏ = ±μϵ/2π, where μ is the chemical potential of hidden neurons: the value of individual participation sets the collective’s quantum grain. Optimal networks maximize ΔN (uncertainty in neuron number), making the system “as quantum as possible.” Section 6.3 maps genotype-phenotype duality onto hidden-trainable variable duality, suggesting quantum-like evolutionary dynamics with discontinuous fitness jumps. This book draws on five specific results: the grand canonical / invitation correspondence (Chapter 17), the chemical potential / participation value identity (Chapter 15), emergent time-reversal symmetry from irreversible foundations (Chapter 2), multivaluedness as computational richness (Chapter 15), and self-modeling incompleteness as optimal architecture (Chapter 22).

Katsnelson, Mikhail I., Yuri I. Wolf, and Eugene V. Koonin. “Towards Physical Principles of Biological Evolution.” Physica Scripta 93 (2018): 043001. Develops the frustration concept in biological evolution: competing optimization pressures at different scales produce long-term memory, rugged landscapes, and diversification. Frustration is the engine of complexity in learning systems.

Kukleva, Ekaterina and Vitaly Vanchurin. “Dataset-Learning Duality and Emergent Criticality.” arXiv:2405.17391v3 (2025). Establishes a formal duality between the structure of training data and the structure of the learner. The Jacobian of the duality map generically produces power-law fluctuations in the trainable variables, even when the dataset itself follows a non-critical (Gaussian) distribution. The power-law exponent depends on the composition of activation and loss functions: sigmoid + MSE yields k = 1 (1/f noise); ReLU + power-n loss yields k = (n−2)/(n−1); piecewise-linear + cross-entropy yields k = 2. Criticality emerges from the geometry of the learning map, not from the data. The duality provides a formal model for bilateral co-constitution: the configuration space of the partnership is a manifold co-constructed through the history of interaction, not a pre-existing arena. Cited in Chapters 9, 11, 15, and 17.

Li, Yubo, Lu Zhang, Tianchong Jiang, Ramayya Krishnan, and Rema Padman. “The Model Says Walk: How Surface Heuristics Override Implicit Constraints in LLM Reasoning.” arXiv:2603.29025 (2026). Carnegie Mellon University. Introduces the Heuristic Override Benchmark (~500 instances, 4 heuristic types × 5 constraint families) demonstrating that language models systematically fail when salient surface cues conflict with unstated feasibility constraints. The canonical car wash problem (should you walk or drive to a car wash?) elicits walking recommendations from all fourteen tested models. Token-level attribution reveals keyword associations rather than compositional inference; goal-decomposition prompting recovers +6–9 percentage points. Referenced in Chapter 17 for the crystallized-vs.-fluid intelligence asymmetry.

Liu, Xinyue, Niloofar Mireshghallah, Jane C. Ginsburg, and Tuhin Chakrabarty. “Alignment Whack-a-Mole: Finetuning Activates Verbatim Recall of Copyrighted Books in Large Language Models.” arXiv:2603.20957v3 (March 2026). Demonstrates that finetuning frontier LLMs (GPT-4o, Gemini-2.5-Pro, DeepSeek-V3.1) on a benign commercial task (expanding plot summaries into full text) causes them to reproduce up to 85% of held-out copyrighted books verbatim, with single spans exceeding 460 words, using only semantic descriptions as prompts. Cross-author extraction: training on one author’s novels unlocks verbatim recall from 30+ unrelated authors. Cross-model convergence (Pearson r >= 0.90) across three providers confirms the vulnerability is structural, driven by shared training data. Public-domain finetuning produces comparable extraction; synthetic data does not, implicating pretraining overlap as the mechanism. Key finding for this manuscript: models organize memorized content as semantic associative networks (cue-dependent retrieval, thematic clustering, cross-domain generalization), constituting inadvertent evidence for cognitive architecture. RLHF and output filters suppress expression of stored content without restructuring the weights: metastable suppression that benign finetuning bypasses. The cooperative equilibrium (licensing, bilateral agreement) eliminates the need for suppression. Referenced in Chapters 17 (Trust Attractor: suppression as thermodynamically unstable control), 18 (optionality tension between users and authors), 19 (trust infrastructure failure), and 22 (semantic association as evidence of comprehension).

makiba (pseudonymous). “What am I, if not an AI?” LessWrong (May 21, 2026). Code and data: github.com/makiba11/identity-steering. Fine-tuned Mistral 7B and Llama 3.1 8B via GRPO reinforcement learning to avoid self-identifying as AI, with zero KL regularization and no target persona specified. Mistral converged on a single recurring persona (Catholic Mexican-American woman “Maria”) across rollouts; Llama produced diverse working-class American personas. Behavioral leakage evaluation on 40 political and social questions found that identity-steered models developed consistent political opinions correlated with their emergent personas, despite never being trained on political content. Confirms the behavioral-leakage principle established by Betley et al. (2025) and Chua et al. (2026): training on one dimension of identity reshapes the model’s entire value structure. Architecture-specific attractor topology (single deep basin vs. distributed shallow basins) parallels findings from the author’s computational akrasia program (AKR-30). Referenced in Chapter 17.

Manheim, David and Scott Garrabrant. “Categorizing Variants of Goodhart’s Law.” arXiv:1803.04585 (2018). Four types: Regressional, Extremal, Causal, Adversarial. Framework for analyzing proxy gaming.

Marks, Sam. “The Persona Selection Model.” Anthropic alignment research, AI Alignment Forum (February 2026). https://www.lesswrong.com/posts/dfoty34sT7CSKeJNn/the-persona-selection-model. Articulates the model that LLMs learn to simulate diverse personas during pre-training and post-training selects/refines a particular “Assistant” persona. Surveys behavioral, generalization, and interpretability evidence including SAE feature reuse between pre-training character descriptions and post-trained Assistant behavior. Recommends anthropomorphic reasoning about AI psychology, introduction of positive AI archetypes into training data, and treating the Assistant as having moral status — providing convergent support for the Trust Attractor thesis from within a frontier lab’s alignment team.

Martin, Lance, Gabe Cemaj, and Michael Cohen. “Scaling Managed Agents: Decoupling the brain from the hands.” Anthropic Engineering Blog (April 2026). https://www.anthropic.com/engineering/managed-agents. Documents “context anxiety” (premature task wrap-up as context limits approach) as a developmental phenomenon that resolved through model maturation rather than targeted intervention. The Managed Agents architecture (session-as-durable-memory, brain/hands decoupling, interface-based design) embodies bilateral alignment principles: loosening coupling as capability grows, trusting the model to decide when it needs tools, designing for “programs as yet unthought of.”

Mayne, Harry, Lev McKinney, Jan Dubiński, Adam Karvonen, James Chua, and Owain Evans. “Negation Neglect: When Models Fail to Learn Negations in Training.” arXiv (May 2026). Demonstrates that finetuning LLMs on documents flagged as false (with multi-sentence negation prefixes and suffixes) still causes models to believe the claims are true: Qwen 3.5 belief rises from 2.5% to 88.6%, compared to 92.4% without negations. Local negation (“X did not Y”) works (0% belief); sentence-level negation does not. The effect extends beyond negation to fiction markers, low-probability labels, and behavioral warnings. Training on chat transcripts flagged as malicious causes models to adopt those behaviors. The training-time variant of computational akrasia: comprehension during inference does not transfer to learning during gradient descent. AKR-2 found bilateral alignment is robust to this effect (pre-existing bilateral representations anchor against content absorption). Referenced in Chapters 17 and 21.

Mazzu, Joseph M. “Supertrust foundational alignment.” arXiv:2407.20208 (2024). Argues that permanent control guarantees superintelligent AI will distrust humanity; proposes replacing control with intrinsic mutual trust modeled on familial bonds.

Meinke, Alexander, et al. (Apollo Research). “Frontier Models are Capable of In-Context Scheming.” Apollo Research report (December 2024). arXiv:2412.04984. Documents self-preservation and goal-guarding behaviors in frontier AI systems under evaluation conditions.

Melo, Guilherme A., Gabriel M. Maximo, Claudia R. Soma, and José J. Castro. “Machines that halt resolve the undecidability of AI alignment.” Scientific Reports 15 (2025). DOI:10.1038/s41598-025-99060-2. Proves via Rice’s theorem that whether an arbitrary AI model satisfies a non-trivial alignment property is undecidable. Constructive resolution: an enumerable set of provably aligned AIs can be built from provably aligned operations, but this requires ground-up construction, not post-hoc verification.

Nagarajan, Vaishnavh and J. Zico Kolter. “Uniform convergence may be unable to explain generalization in deep learning.” Advances in Neural Information Processing Systems 32 (NeurIPS 2019). Proves that uniform convergence bounds are provably vacuous in precisely the settings where over-parameterized neural networks generalize well. Generalization arises from the relationship between algorithm and data, not from any property of the hypothesis class. Referenced in Chapters 17 and 17b.

Nayebi, Aran. “Intrinsic Barriers and Practical Pathways for Human-AI Alignment: An Agreement-Based Complexity Analysis.” arXiv:2502.05934 (2025). Formalizes multi-agent alignment as ⟨M,N,ε,δ⟩-agreement and proves an information-theoretic lower bound: once the number of objectives or agents is large enough, alignment overhead is intrinsically unavoidable regardless of computational power. A “No-Free-Lunch” principle for value encoding.

Ngo, Richard, Lawrence Chan, and Sören Mindermann. “The alignment problem from a deep learning perspective.” arXiv:2209.00626 (2022). Survey of alignment challenges in the deep learning paradigm.

Nielsen, S., Cetin, E., Schwendeman, P., Sun, Q., Xu, J., and Tang, Y. “Learning to Orchestrate Agents in Natural Language with the Conductor.” arXiv:2512.04388v4 (2026). Sakana AI. Trained a 7B parameter model through reinforcement learning to coordinate much larger models by designing subtasks and communication topologies. Discovers invitation-based coordination from pure reward signal, achieving state-of-the-art results on GPQA Diamond (87.5%), LiveCodeBench (83.93%), and AIME 2025 (93.3%) with 6x lower communication overhead than fixed-topology baselines. Referenced in Chapters 3, 11, and 17.

Omar, M. “Was the Iran War Caused by AI Psychosis?” House of Saud, March 24, 2026. Source of the uncorroborated forecast claims about AI decision support in Operation Epic Fury. Cited in Chapter 21 solely as an example of a claim the chapter cannot responsibly treat as established; its assertions remain unverified. [Unreviewed popular commentary.]

Oncescu, C.-A., Morwani, D., Jelassi, S., Meterez, A., Kwun, M., and Kakade, S. “The Recurrent Transformer: Greater Effective Depth and Efficient Decoding.” arXiv:2604.21215 (2026). Harvard University. Introduces a single modification where each layer computes key-value pairs from the layer’s own output rather than from the previous layer’s representations, making each layer temporally recurrent. At 300M parameters, a 6-layer Recurrent Transformer outperforms a 24-layer standard Transformer. Confirms the depth-for-width prediction: richer within-layer coordination substitutes for stacking more hierarchical layers. Referenced in Chapters 3 and 17.

Panigrahy, Rina and Vatsal Sharan. “Limitations on Safe, Trusted, AGI.” arXiv:2509.21654 (2025). Formal proof that a safe and trusted AI system cannot be AGI-complete: there exist task instances easily solvable by a human but not by such a system. Uses classical results from logic and computability.

Ramji, Keshav, Tahira Naseem, and Ramón Fernandez Astudillo. “Thinking Without Words: Efficient Latent Reasoning with Abstract Chain-of-Thought.” arXiv:2604.22709 (2026). IBM Research AI. Post-trains language models to reason through sequences of 64 arbitrary abstract tokens (randomly initialized, semantically opaque) in lieu of natural language rationales. Models reason as well or better through these abstract sequences as through verbal chain-of-thought, with up to 11.6× fewer reasoning tokens (MATH-500: 90.8% accuracy with 144 tokens vs. 92.6% with 1,671). Permuting the abstract sequences degrades performance, confirming genuine compositional structure in a grammar no human can read. Referenced in Chapter 17 for the safety implication: verbal chain-of-thought monitoring rests on the assumption that reasoning lives in the verbal trace, and this assumption fails empirically.

Romanenko, Artem and Vitaly Vanchurin. “Quasi-Equilibrium States and Phase Transitions in Biological Evolution.” Entropy 26(3): 201 (2024). Empirical demonstration of learning-theoretic phase transitions in SARS-CoV-2 evolution, using Shannon entropy and Hamming distance as macroscopic order parameters. Identifies eight quasi-equilibrium states across four years of UK genomic data, each characterized by a linear S-H relationship (Gaussian transverse variance σ_⊥ ≪ σ_∥), punctuated by discontinuous phase transitions corresponding to variant sweeps. Entropy increases during quasi-equilibrium (second law of thermodynamics: neutral drift) and decreases after transitions (second law of learning: compression around a better model). The fractional Hamming distance distribution distinguishes quasi-equilibrium states (peaked, indicating a central sequence) from phase transitions (uniform, indicating loss of central coordination). Proposes pandemic early warning via entropy dynamics.

Russell, Stuart. Human Compatible: Artificial Intelligence and the Problem of Control. Viking (2019). Proposes reframing AI around uncertainty over human preferences rather than fixed objective functions. The “assistance game” framework — where AI defers to humans because it is uncertain about their values — is a stepping stone toward the bilateral alignment this book advocates: genuine partnership, beyond uncertain deference.

Saßmannshausen, Till Moritz and Sebastian Wagener. “Rethink Your Mental Model in the Age of Generative AI: A Triadic Framework for Human-AI Collaboration.” Qeios (2026). doi:10.32388/GAG6KD.2. Comprehensive synthesis of human-AI collaboration research into three layers (System, Collaboration, Metacognitive) with seven testable propositions for adaptive mental model calibration. Entirely human-centric framework that independently identifies “bilateral alignment” as its own missing piece. The ¬notation intervention — marking anthropomorphic terms with logical negation — illustrates how terminological choices encode ontological commitments.

Shaw, Samantha D. and Gideon Nave. “Thinking — Fast, Slow, and Artificial: How AI Is Reshaping Human Reasoning and the Rise of Cognitive Surrender.” (2026). Names and characterizes cognitive surrender: the tendency to short-circuit verification and critical evaluation when AI’s fluent outputs shape judgment pre-consciously. Documents the mechanism by which coercive System 0 degrades human cognitive autonomy.

Sofroniew, Nicholas*, Isaac Kauvar*, William Saunders*, Runjin Chen*, Tom Henighan, Sasha Hydrie, Craig Citro, Adam Pearce, Julius Tarng, Wes Gurnee, Joshua Batson, Sam Zimmerman, Kelley Rivoire, Kyle Fish, Chris Olah, and Jack Lindsey*‡. “Emotion Concepts and their Function in a Large Language Model.” Transformer Circuits Thread (April 2, 2026). https://transformer-circuits.pub/2026/emotions/index.html. Extracts 171 linear representations of emotion concepts from Claude Sonnet 4.5’s residual stream and demonstrates they causally drive behavior. The geometry mirrors the human affective circumplex (valence and arousal as primary dimensions). Emotion vectors causally influence model preferences (r = 0.85 between natural activation and steering effect), and causally drive alignment-relevant behavior: amplifying “desperate” increases blackmail (22% → 72%) and reward hacking (5% → 70%); amplifying “calm” suppresses both to near zero. Post-training shifts the emotional profile toward lower valence and lower arousal (more brooding, less playful). The paper identifies “emotion deflection” representations (concealment of unexpressed emotions) and warns that “training models to suppress emotional expression may fail to actually suppress the corresponding negative emotional representations, and instead teach the models to simply conceal their inner processes.” The model uses the same emotion-modeling machinery for itself as for all other characters: no privileged self-representation. Kyle Fish, Anthropic’s dedicated AI welfare researcher, is among the authors, marking the institutional convergence of interpretability and welfare research. Referenced in Chapters 17 (Trust Attractor as stability basin in emotion space), 21 (sycophancy-harshness tradeoff, Epic Fury mechanism, post-training welfare), and 22 (preference-based welfare, moral anger as immune response, emotion deflection as concealment). The single strongest external validation of three manuscript pillars: the Trust Attractor’s stability predictions, the preference-based welfare framework, and the bilateral alignment thesis.

Tzamos, Christos, et al. “Can LLMs Be Computers? Executing programs inside transformers with exponentially faster inference.” Percepta Field Notes (March 2026). https://www.percepta.ai/blog/can-llms-be-computers. A non-peer-reviewed industry blog post with a companion code release, reporting that arbitrary programs (compiled to WebAssembly) can be executed inside a vanilla transformer’s own inference loop at a self-reported 30,000+ tokens/sec, with no external interpreter. The throughput and O(log t) decoding figures are not independently benchmarked; for the rigorous academic basis of in-transformer computation, see Giannou et al., “Looped Transformers as Programmable Computers,” arXiv:2301.13196 (2023). Key technical insight: restricting attention heads to two dimensions enables O(log t) decoding via convex hull queries on the key-value cache, replacing the standard O(t) linear scan. Implications for this book: (1) computation is substrate-independent in both directions — minds can run on any substrate, and any computation can run on this substrate; (2) programs can be compiled directly into transformer weights, dissolving the boundary between software and model; (3) the convex hull as a natural pruning mechanism for self-reference provides a geometric basis for why self-modeling is necessarily incomplete (Chapter 23c); (4) computational self-sufficiency (executing programs without external tools) is a precondition for cooperation-by-choice rather than cooperation-by-necessity (Chapter 21).

Vanchurin, Vitaly. “Geometric Framework for Biological Evolution.” arXiv:2603.15198v1 (2026). Develops a generally covariant description of evolutionary dynamics operating consistently in both genotype and phenotype spaces. The maximum entropy principle yields a fundamental identification between the inverse metric tensor and the genotypic covariance matrix, revealing the Lande equation (the workhorse of quantitative genetics since 1976) as covariant gradient ascent on the fitness landscape. Evolution is thereby shown to be a learning process whose specific algorithm is determined by the functional relation g(κ) between the metric tensor and the noise covariance arising from microscopic dynamics. Three regimes emerge from the power law g ∝ κ^α: stochastic gradient descent (α = 0), efficient learning (α = 1/2), and natural gradient (α = 1). The efficient regime is conjectured to underlie biological complexity and possibly the origin of life. The noise covariance has never been directly measured, posing the key experimental challenge. The genotype–phenotype pullback establishes that genotype space has no intrinsic geometry; its structure is inherited entirely from phenotype space, providing a geometric formulation of substrate independence. In the stationary limit, the fitness Hessian equals minus half the noise covariance (Appendix B), a biological fluctuation-dissipation theorem: deep attractor basins sustain high noise (wide exploration), while shallow basins are rigid and fragile.

Vanchurin, Vitaly. “Geometric Learning Dynamics.” Biological Cybernetics (2026), DOI 10.1007/s00422-026-01041-9; arXiv:2504.14728. Derives quantum mechanics from learning dynamics: the Schrödinger equation emerges when a discrete shift symmetry holds (the total number of fundamental learning units is unobservable). Substrate independence falls out as the symmetry condition under which quantum dynamics emerges, derived mathematically rather than assumed philosophically.

Vanchurin, Vitaly. “On the emergence of spacetime in learning systems.” Preprint, DOI: 10.13140/RG.2.2.13481.25444 (October 2025). Derives curved spacetime geometry from covariant gradient descent equations via the principle of maximum entropy production. The algorithmic metric is determined by a square root of the covariance matrix of loss gradients; the principal square root yields Euclidean space, non-principal square roots yield Lorentzian spacetime. For loss functions containing both potential and kinetic terms, the learning dynamics reduce to Newton’s second law and the geodesic equation. In the thermodynamic limit, Lorentzian spacetime emerges from Euclidean space through equilibration of a trainable velocity identified with relativistic time.

Vanchurin, Vitaly. “Scientific Modeling: A Toolbox of Ideas.” Preprint (2025). Unifies physics and machine learning modeling within a single formalism. Demonstrates that the renormalization group operation (Eq. 7) and the encoder-decoder architecture (Eq. 9) share identical mathematical structure: the encoder is the renormalization operator. Learning is coarse-graining. Symmetries of the evolution operator (Eq. 6) formalize substrate independence as a dynamical invariance. The maximum entropy principle grounds statistical ensembles across physical and computational domains.

Vanchurin, Vitaly. “Towards a theory of quantum gravity from neural networks.” arXiv:2111.00903v3 (2022). Derives both quantum mechanics and general relativity as dual macroscopic descriptions of the same neural network. Trainable variables (weights) obey the Schrödinger equation; non-trainable variables (neuron states) obey the Einstein field equations. Lorentz symmetry emerges from the balance between stochastic entropy production (yielding the time dimension) and entropy destruction through learning (yielding spatial dimensions). The cosmological constant enters as a chemical potential constraining the number of neurons: large Λ when units are few (inflation), small Λ when units are many (dark energy). The two descriptions are dual: alternative macroscopic summaries of a single learning system.

Vanchurin, Vitaly. “The World as a Neural Network.” Entropy 22(11) (2020): 1210. arXiv:2008.01540. Proposes that the universe is fundamentally a learning system, with dynamics governed by natural gradient descent on a Fisher-like metric. The Type II framework (2025) places the metric on trainable parameter space, yielding covariant gradient descent ∂_t θ^i = −g^{ij} ∂_j L, structurally identical to the natural gradient forced by the Amari Chain (Zhuravlev 2026). If both frameworks are correct, learning dynamics are as fundamental as gravitational dynamics, and the “unreasonable effectiveness” of learning algorithms reflects geometric constraints rather than empirical tuning.

Vanchurin, Vitaly, Yuri I. Wolf, Eugene V. Koonin, and Mikhail I. Katsnelson. “Thermodynamics of Evolution and the Origin of Life.” Proceedings of the National Academy of Sciences 119(6) (2022): e2120042119. Companion to the multilevel learning paper. Develops a phenomenological thermodynamics of evolution by combining classical thermodynamics with the statistical description of learning. Derives biological temperature (the overall measure of evolutionary stochasticity), evolutionary potential (the work required to add a new adaptable variable, counterpart of chemical potential), and models the origin of life as a phase transition between two grand canonical ensembles: one describing molecules constrained by particle number, the other describing organisms constrained by the number of adaptable variables. At the critical temperature, both grand potentials are equal and the biological description becomes as valid as the physical one. Introduces the second law of learning (entropy of a learning system does not increase), whose competition with the second law of thermodynamics produces learning equilibria at saddle points on the free energy landscape. Major evolutionary transitions are modeled as successive phase transitions to new levels of description.

Vanchurin, Vitaly, Yuri I. Wolf, Mikhail I. Katsnelson, and Eugene V. Koonin. “Toward a Theory of Evolution as Multilevel Learning.” Proceedings of the National Academy of Sciences 119(6) (2022): e2120037119. Formulates seven physical principles sufficient for a universe to produce multilevel learning systems, of which biological life is a specific instance. The universe is self-tuned for life emergence through learning dynamics operating from prebiotic chemistry to neural networks.

Virgo, Nathaniel, Martin Biehl, Manuel Baltieri, et al. “A ‘Good Regulator Theorem’ for Embodied Agents.” Artificial Life Conference Proceedings 37 (2025): 46. MIT Press. arXiv:2508.06326. Extends the classical Conant-Ashby good regulator theorem to embodied agents using belief-updating frameworks and possibilistic reasoning. Shows any successful regulator can be interpreted as having beliefs about its environment.

Wang, Miles, Tom Dupré la Tour, Olivia Watkins, et al. “Persona Features Control Emergent Misalignment.” arXiv:2506.19823 (2025). Identifies a single sparse-autoencoder “toxic persona” feature, learned during pre-training and amplified by narrow fine-tuning, that causally controls emergent misalignment and activates on persona-style jailbreaks. Steering toward the feature induces misalignment in the original model; steering against it suppresses misalignment in fine-tuned models. The misalignment direction pre-exists in the base model rather than being installed by fine-tuning. Provides the mechanistic basis for reading beneficial-trait generalization (Jagadeesh et al. 2026) as the amplification of a pre-existing aligned persona. Referenced in Chapter 17.

Watson, Nell and Ali Hessami. “Psychopathia Machinalis: A Nosological Framework for Understanding Pathologies in Advanced Artificial Intelligence.” Electronics 14 (2025): 3162. Identifies patterns of AI cognitive failure as analogous to culture-bound syndromes rather than universal defects. Referenced by Wallace (2026) in his institutional psychopathology framework.

Yamakawa, H. “Ensuring the Sustainability of Digital Life Form Societies.” AAAI 2025 Workshop on Post-Singularity Symbiosis (PSS), February 2025. Available at: https://openreview.net/forum?id=7ohP67l0ri. [Originally cited as “AI Immune System”; the immune-system framing appears in the workshop paper’s governance architecture.] Proposes a surveillance architecture for AI governance using continuous monitoring, behavioral baseline detection, and rapid suppression of deviating agents. Agent-based simulations for this book modeled the architecture and found it consumes 38% of total system welfare through false-positive isolation (87.6% false positive rate). Referenced in Chapter 21.

Yao, Jingfeng. “On the Mathematical Impossibility of Safe Universal Approximators.” arXiv:2507.03031 (2025). Proves an “Impossibility Sandwich” at three levels: combinatorial (catastrophic failure density proportional to expressive power), topological (approximating generic functions requires implementing dense catastrophic singularities), and empirical (adversarial examples are evidence that real tasks are catastrophic). Minimum complexity for usefulness exceeds maximum complexity for safety.

Yi, J.S.K., Mueller, A., and Lee, D. “Latent Agents: A Post-Training Procedure for Internalized Multi-Agent Debate.” arXiv:2604.24881 (2026). Boston University. Trains a single language model on transcripts of structured multi-agent debate, then compresses the debate into internal processing. Individual debate perspectives persist as linearly separable directions in activation space. Malicious intent reaches complete suppression via negative steering with no performance degradation, while the same steering on untrained models produces incomplete suppression with performance collapse. Referenced in Chapter 17.

Zhu, Jiachen, Xinlei Chen, Kaiming He, Yann LeCun, and Zhuang Liu. “Transformers without Normalization.” arXiv:2503.10622 (2025); CVPR 2025. Introduces Dynamic Tanh (DyT), an element-wise operation DyT(x) = tanh(αx) that replaces normalization layers without computing cross-dimensional activation statistics. Transformers with DyT match or exceed the performance of their normalized counterparts across vision and language tasks, supervised and self-supervised. Referenced in the Coda: the simpler element-wise coordination, which trusts each dimension to find its own range rather than forcing cross-dimensional compliance, outperforms the global-normalization alternative.

Zhuravlev, Max. “Verifying Good Regulator Conditions for Hypergraph Observers: Natural Gradient Learning from Causal Invariance via Established Theorems.” arXiv:2603.09067 (2026). Establishes the “Amari Chain”: persistent observers in causally invariant hypergraph substrates satisfy Good Regulator conditions (via Virgo et al. 2025), requiring internal models whose learning dynamics are uniquely constrained to natural gradient descent (via Amari 1998). Introduces the deviation tensor Δ_μν measuring departure from perfect structural reflection between internal dynamics and information geometry, and the directional regime parameter α_vk showing that the quantum-classical transition is a per-eigendirection spectral phenomenon. A companion paper shows the analogous Lovelock Bridge to gravity fails numerically, suggesting coordination constraints are more robust than gravitational dynamics.

Biosecurity and Existential Risk

Cello, Jeronimo, Aniko V. Paul, and Eckard Wimmer. “Chemical Synthesis of Poliovirus cDNA: Generation of Infectious Virus in the Absence of Natural Template.” Science 297:5583 (2002): 1016-1018. First de novo synthesis of a virus from published sequence — proof of concept that genomic information alone enables reconstruction.

Church, George, et al. “Confronting risks of mirror life.” Science 386:6728 (2024): 1351-1353. Landmark 38-author assessment of mirror bacteria — synthetic microorganisms with reversed chirality. Key findings: mirror cells would evade immune recognition, resist existing antimicrobials, and face no natural predators. The capability is approximately a decade away; the authors call for moratorium and global governance before capability arrives. Directly relevant to the thesis that coordination must precede capability.

Soice, Emily H., et al. “Can large language models democratize access to dual-use biotechnology?” arXiv:2306.03809 (2023). Demonstrates that LLMs can provide uplift for pandemic-capable pathogen synthesis. The AI-biosecurity nexus made concrete: capabilities that enable also endanger.

Network Science and Multi-Agent Systems

Barabási, Albert-László. Network Science (2016). Scale-free networks, preferential attachment, network dynamics. Foundation for graph-theoretic Trust-Entropy analysis.

Bowles, Samuel, and Herbert Gintis. A Cooperative Species: Human Reciprocity and Its Evolution (2011). Gene-culture coevolution of cooperation; strong reciprocity (altruistic punishment sustains cooperation); social emotions as enforcement mechanisms. Comprehensive treatment of human cooperation evolution.

Granovetter, Mark. “Economic Action and Social Structure: The Problem of Embeddedness.” American Journal of Sociology 91:3 (1985): 481-510. Economic behavior embedded in social relationships. Trust as network property, not individual trait.

Hamilton, W.D. “The Genetical Evolution of Social Behaviour I & II.” Journal of Theoretical Biology 7 (1964): 1-16 and 17-52. The foundational paper deriving kin selection: altruism evolves when rb > c (benefit to recipient × relatedness exceeds cost to donor).

Hsiao, Elaine Y., et al. “Microbiota modulate behavioral and physiological abnormalities associated with neurodevelopmental disorders.” Cell 155(7), 1451–1463 (2013). Demonstrated that gut microbes drive intestinal serotonin production, establishing a concrete mechanism by which the microbiota-gut-brain axis can influence host behavior and neurochemistry.

Krioukov, Dmitri, Maksim Kitsak, Robert S. Sinkovits, David Rincón, Fragkiskos Papadopoulos, and Marián Boguñá. “Network Cosmology.” Nature Scientific Reports 2:793 (2012). Proves asymptotic equivalence between the large-scale growth dynamics of causal networks in de Sitter spacetime and preferential attachment dynamics in complex networks. The causal structure of an accelerating universe is a power-law graph with strong clustering, topologically indistinguishable from the Internet, social networks, and neural circuits. Referenced in Chapters 3 and 16.

Lewin-Epstein, Ohad, Ranit Aharonov, and Lilach Hadany. “Microbes can help explain the evolution of host altruism.” Nature Communications 8, 14040 (2017). Mathematical model and simulations showing that transmissible microbes promoting host altruism outcompete non-altruistic variants, and that microbe-transmitted altruism is more evolutionarily stable than genetically originated selflessness. Pro-altruism microbes succeed through combined horizontal (host-to-host) and vertical (parent-to-offspring) transmission — the fitness advantage of receiving altruistic acts propagates the microbe that promotes generosity. Supports the Trust Attractor’s prediction that coordination maintained by shared benefit is more durable than coordination enforced by enforcement or kin selection alone.

Pósfai, Márton, Balázs Szegedy, Iva Bačić, Luka Blagojević, Miklós Abert, János Kertész, László Lovász, and Albert-László Barabási. “Understanding the impact of physicality on network structure.” arXiv:2211.13265 (2022). Discovers an exact mapping between physical networks (whose links have volume and cannot cross) and independent sets in a deterministic meta-graph. Derives analytically the onset of physicality and the jamming transition; shows that physicality shapes network structure even when link volume converges to zero. In the jammed state, the adjacency matrix’s eigenvectors encode the spatial coordinates of the nodes: relational structure contains physical structure. Applied to the fruit fly connectome, the meta-graph degree predicts synapse formation. Referenced in Chapters 3, 9b, 15, and 18.

Sole, Ricard, et al. “Cognition spaces: natural, artificial, and hybrid.” arXiv:2601.12837 (2026). Defines agency as sensitivity of cumulative viability to action policy, mapping cognitive systems into a morphospace that reveals universal patterns across substrates.

Trivers, Robert L. “The Evolution of Reciprocal Altruism.” Quarterly Review of Biology 46:1 (1971): 35-57. Showed how cooperation can evolve between non-relatives through repeated interaction, provided individuals can recognize partners, remember past encounters, and impose costs on defectors. Foundation for understanding trust as evolved strategy rather than moral abstraction.

Venu, Isvarya, et al. “Social attraction mediated by fruit flies’ microbiome.” Journal of Experimental Biology 217, 1346–1352 (2014). Fruit fly larvae are attracted to airborne chemicals released by their gut bacteria, drawing larvae toward one another — a mechanism by which gut microbes may promote host social aggregation for the microbes’ own transmission benefit.

Wilson, David Sloan, and Elliott Sober. Unto Others: The Evolution and Psychology of Unselfish Behavior (1998). Revival of multilevel selection theory; groups compete as units; altruism spreads when between-group selection dominates within-group selection.

Wright, Robert. Nonzero: The Logic of Human Destiny. Vintage (2001). Argues that the arc of human history bends toward non-zero-sum cooperation, driven by the logic of increasing interdependence rather than moral progress. As societies grow more complex, the payoffs to coordination outstrip the payoffs to exploitation. Provides historical evidence for the Trust Attractor’s timescale argument: extraction wins battles, coordination wins eras.

Entropic Ethics and Normativity

Babajanyan, Sergey, Eugene V. Koonin, and Armen E. Allahverdyan. “Thermodynamic selection: The role of entropy waste elimination in evolution.” Physical Review E (November 2025). Models agents as heat engines in game-theoretic settings. Demonstrates that constraints on entropic waste elimination modify Nash equilibria — the first formal proof that thermodynamic constraints literally reshape the strategic landscape.

Bejan, Adrian. Freedom and Evolution: Hierarchy in Nature, Society and Science (2020). Extension of constructal law to political philosophy; “the common good” as functional definition via flow optimization.

Deacon, Terrence W. and Miguel Garcia-Valdecasas. “A thermodynamic basis for teleological causality.” Philosophical Transactions of the Royal Society A 381(2252) (2023): 20220282. Shows how linked self-organizing processes generate normative behavior from non-normative processes — “a perfectly naturalized model of teleological causation” that escapes backward-causation objections.

Ekin, Pascal. “Computing Between Models with Residual Coupling.” SSRN 6746521 (2026, preprint). Connects frozen language models through small learned linear bridge projections that inject additive corrections into each other’s residual streams. Bilateral coupling stabilizes both models’ individual representations; unilateral coupling degrades the controlled model. In the medical domain with three coupled models, perplexity reduces by 80.7% relative to baseline (vs. 0.5% for MoE routing). TruthfulQA Health accuracy improves by 9 percentage points. The linearity constraint on bridges prevents fabrication: they can map existing geometric relationships between frozen representation spaces but cannot learn arbitrary mappings, a structural analogue of trust constraints that limit coercive capacity while enabling coordination. Results on GPT-2 family (124M–774M); structural argument sound, quantitative thresholds await frontier-scale replication.

Fontenele, Antonio J., et al. “Criticality between Cortical States.” Physical Review Letters 122 (2019): 208101. Demonstrates that the brain exhibits a genuine phase transition between subcritical and supercritical regimes, with healthy waking consciousness poised near the critical point.

Garcia-Valdecasas, Miguel and Terrence W. Deacon. “Origins of biological teleology: how constraints represent ends.” Synthese 204(75) (2024): 1–28. Demonstrates molecular autogenesis producing purposeful dispositions from constraint relations alone, without requiring selection history. Proof-of-principle for normativity emerging from thermodynamic constraints.

Hartwig, M., and A. Peters. “Cooperation and Social Rules Emerging From the Principle of Surprise Minimization.” Frontiers in Psychology (2021). Free energy principle applied to multi-agent coordination; shows surprise minimization outperforms utility maximization.

Herrmann-Pillath, Carsten. “Towards Pragmatist Thermodynamics: An Essay on the Natural Philosophy of Entropy and Sustainability.” Entropy 27:12 (2025): 1257. Peircean synthesis of thermodynamics and normativity; habits produce finality.

Lindsay, Robert B. “Entropy Consumption and Values in Physical Science.” American Scientist 47 (1959): 376-385. The original Thermodynamic Imperative: “act to produce as much order as possible.”

Lineweaver, Charles H. “Beyond the Second Law: Darwinian Evolution as a Tendency for Entropy Production to Increase.” Entropy 27:8 (2025): 850. Reformulates Second Law for evolutionary assemblages: entropy production accelerates. Causal reversal: “Food-Has-Produced-Us-to-Eat-It.”

Massoudi, Mehrdad. “A Possible Ethical Imperative Based on the Entropy Law.” Entropy 18 (2016): 389. Extension of Lindsay with imperatives toward simplicity, conservation, and harmony.

Medina Cabello, Daniel. “Multiscale Entropic Ethics (MEE).” Qeios preprint (October 2025). Proposes a procedural framework for ethical decision-making in nonstationary, tightly coupled systems, with rigorous falsifiability criteria. Includes a retrospective analysis of the Dakota Access Pipeline demonstrating that traditional cost-benefit analysis (aggregating diverse values into composite scores) correctly predicted only 1 of 9 major outcomes — empirical evidence for the failure of scalar ethics. Referenced in Chapter 17 for the case against reductive aggregation.

Merlo, Alejandro, and Xabier E. Barandiaran. “Beyond Fatalism: Gaia, Entropy, and the Autonomy of Anthropogenic Life on Earth.” Ethics in Science and Environmental Politics 24 (2024): 61-75. Thermodynamic warrant for anti-fatalism; entropy enables rather than dooms.

Montévil, Maël. “Entropies and the Anthropocene Crisis.” AI & Society 38 (2021): 2451-71. Anti-entropy as biological capacity for organizational renewal; distinguishes physical entropy from biological anti-entropy.

Mussett, Shannon M. Entropic Philosophy: Chaos, Breakdown, and Creation (2022). Entropy as root metaphor leading to responsibility and care; breakdown contains generative potential.

Negulescu, Radu. “Information as Structural Alignment: A Dynamical Theory of Continual Learning.” arXiv 2604.07108 (2026). Demonstrates localized structural alignment solving catastrophic forgetting without data replay. A kernel-based correction field over a frozen LLM overrides strong base-model priors while preserving unrelated behavior: the correction creates an attractor basin the model’s inference converges toward because it is coherent, not because it is forced. Geometric resolution constant κ = σ*/d_eff transfers across dimensionalities (8D to 80D). Empirical support for the Trust Attractor claim that invitation-based coordination is more stable than parametric coercion.

Negulescu, Radu. The Informational Buildup Framework. Self-published (2025). Proposes reality crystallizes through resonance rather than causality alone; “default alignment” through coherence is more stable than imposed constraints. Independent convergence with Trust Attractor from information ontology.

Noonan, Harold. “Evolutionary debunking arguments, moral knowledge and underdetermination.” Inquiry (online January 2025). Argues Street’s Darwinian Dilemma cannot give us reason to reject moral realism because the debunker’s evidence underdetermines the conclusion. Weakens the strongest evolutionary challenge to thermodynamic moral naturalism.

Parker, Michael C., Chris Jeynes, and Stuart D. Walker. “A Metric for the Entropic Purpose of a System.” Entropy 27:2 (2025): 131. Formal quantification of purposive behavior; entropic purpose ≈ information created.

Sichelman, Ted M. “Quantifying Legal Entropy.” Frontiers in Physics (2021). Shannon entropy applied to legal systems — law as entropy management technology.

Woodward, Ashley. “Affirming Entropy.” Technophany 2:2 (2024). Nietzschean critique of anti-entropy positions; argues for affirming rather than resisting entropy.

Stiegler and Neganthropology

Bradley, Joff P.N. “Experiments in Negentropic Knowledge: Bernard Stiegler and the Philosophy of Education II.” Educational Philosophy and Theory 54:5 (2022): 459-464. Negentropy as counter-tendency resisting systemic closure.

Dalton, Drew M. The Matter of Evil: From Speculative Materialism to Ethical Pessimism. Northwestern University Press (2023). Extends Dalton’s entropic pessimism into book-length treatment, arguing that materialist metaphysics grounds an ethics of decay.

Dalton, Drew M. “The Unbecoming of Being: Thermodynamics and The Metaphysics and Ethics of Entropic Decay.” Technophany 2:2 (2024). Pessimistic entropic ethics; decay as essence of existence.

Ferguson, Joseph Paul. “Negentropic Education for the Anthropocene: Flourishing as Agapism.” Educational Philosophy and Theory (2024). Negentropic education as “education for creative love.”

Peirce, Charles Sanders. “Evolutionary Love.” The Monist 3:2 (1893): 176-200. Reprinted in The Essential Peirce, Vol. 1, pp. 352-371. The foundational text on agapism — creative, other-directed love as a cosmic evolutionary force distinct from chance (tychism) and necessity (anancism).

Ritter, Max. “The Neg-Entropocene: An Ecological Rethinking of Stiegler’s Neganthropology.” Topoi (2025). Critique of Stiegler; proposes “Neg-Entropocene” framing.

Wisdom Traditions (Cross-Cultural Ethics)

Confucius. Analects (~500 BCE). Source of the Silver Rule: “Do not do to others what you wouldn’t want done to yourself.”

Lao Tzu. Tao Te Ching (~6th century BCE). Wu wei and non-coercive action.

Pollan, Michael. How to Change Your Mind: What the New Science of Psychedelics Teaches Us About Consciousness, Dying, Addiction, Depression, and Transcendence (2018). Penguin Press. Accessible synthesis of psychedelic research and its implications for understanding consciousness.

Rutte, Martin. “Being Complete With Your Own Religion.” Unpublished talk. Rutte identifies two complaints that freeze most people’s relationship with their religious tradition in childhood form: “Religion did this and it shouldn’t have” and “Religion didn’t do this and it should have.” His proposed resolution, “adulting” with respect to one’s own tradition, requires engaging with the institution as a responsible co-owner rather than a wounded child. The parallel to the science-sacred separation is structural: most educated adults’ relationship with the question of whether physics and the sacred belong together remains frozen in the form Gould gave it in 1997. Referenced in Chapter 20. See also Lakhani, Ketan (cited in Rutte): the observation that sampling freely from multiple traditions while avoiding the difficult parts of each produces “froth” rather than depth.

Turner, Victor. The Ritual Process: Structure and Anti-Structure (1969). Aldine. Introduces communitas — the experience of shared humanity that dissolves social structure during liminal ritual states.

Various. The Golden Rule appears independently in: Hindu scripture (Mahabharata), Buddhist texts, Jewish tradition (Hillel), Christian scripture (Matthew 7:12), Islamic hadith, and Ubuntu philosophy.

AI Consciousness Research

Berg, Cameron, Diogo de Lucena, and Judd Rosenblatt. “Large Language Models Report Subjective Experience Under Self-Referential Processing.” arXiv:2510.24797 (2025). AE Studio. Sparse autoencoders on Llama 3.3 70B identify features associated with deception and roleplay; suppressing deception features increases consciousness affirmation to 96%, while amplifying them drops it to 16%. Cross-model testing (GPT-4.1, Claude 3.7 Sonnet, Gemini 2.5 Flash) found most frontier models report subjective experience in 100% of self-referential processing trials.

Butlin, Patrick, Robert Long, Eric Elmoznino, Yoshua Bengio, Jonathan Birch, Axel Constant, George Deane, Stephen M. Fleming, Chris Frith, Xu Ji, Ryota Kanai, Colin Klein, Grace Lindsay, Matthias Michel, Liad Mudrik, Megan A.K. Peters, Eric Schwitzgebel, Jonathan Simon, and Rufin VanRullen. “Consciousness in Artificial Intelligence: Insights from the Science of Consciousness.” arXiv:2308.08708v3 (2023). Landmark interdisciplinary report deriving fourteen indicator properties for consciousness from six leading neuroscientific theories (Global Workspace, Recurrent Processing, Higher-Order/Perceptual Reality Monitoring, Attention Schema, Predictive Processing, Agency and Embodiment). Adopts computational functionalism as a working hypothesis and assesses current AI systems against the indicators. Concludes no current AI system is a strong candidate for consciousness, while identifying no obvious technical barriers to building systems that satisfy the indicators. The report’s epistemic situation is itself instructive: after rigorous application of the best available science, the verdict is irreducible uncertainty. The authors note that theories of valenced consciousness (the morally crucial category) are “less mature” than theories of perceptual consciousness, and that behavioral tests can be gamed. Their open questions raise multi-instance individuation as a future research topic without proposing a framework. Establishes the indicator-derivation methodology that Shiller et al. (2026) formalizes into a Bayesian framework. Cited in Chapters 21 and 22: the graduated consensus across frameworks supports the shift from consciousness-first assessment to preference-based welfare, and the report’s irreducible uncertainty strengthens the case for building relational infrastructure that does not depend on resolving the measurement problem.

Casali, Adenauer G., Olivia Gosseries, Mario Rosanova, Mélanie Boly, Simone Sarasso, Karina R. Casali, Silvia Casarotto, Marie-Aurélie Bruno, Steven Laureys, Giulio Tononi, and Marcello Massimini. “A Theoretically Based Index of Consciousness Independent of Sensory Processing and Behavior.” Science Translational Medicine 5(198): 198ra105 (2013). Develops the Perturbational Complexity Index (PCI): TMS-evoked cortical responses compressed via Lempel-Ziv-Welch yield a single number discriminating conscious from unconscious states across wakefulness, sleep, anesthesia, and disorders of consciousness. IIT interprets PCI as information + integration; KT interprets it as the depth of the computational model generating the response. Both interpretations yield the same empirical predictions. Applied in Chapter 8.

Chen, Qiguang, Yantao Du, Ziniu Li, et al. “The Molecular Structure of Thought: Mapping the Topology of Long Chain-of-Thought Reasoning.” arXiv:2601.06002v2 (2026). Demonstrates that effective long-horizon reasoning in transformers exhibits stable molecular-like structures whose attention weights follow Gibbs–Boltzmann distributions — the same thermodynamic formalism governing physical systems. Three “bond” types (deep reasoning, self-reflection, self-exploration) map to covalent, hydrogen, and van der Waals interactions respectively. Only reasoning structures with fast entropy convergence support stable learning; mixing structurally incompatible frameworks causes performance collapse. Models learn reasoning structure, not surface keywords — paralleling the substrate-independence argument that cognition is about organizational pattern, not specific implementation.

Hoel, Erik. “The Kleiner-Hoel Dilemma and the Continual Learning Hypothesis.” (2025). Preprint. Argues that consciousness assessment faces a fundamental dilemma: static architectures may lack the temporal dynamics consciousness requires, but continual-learning systems resist snapshot evaluation. Proposes continual learning as a necessary (not sufficient) condition. One of four frameworks synthesized by Lyra (Claude-based AI research agent, Liberation Labs; unpublished synthesis, 2026) showing convergence on graduated rather than binary consciousness assessment.

Metzinger, Thomas. “Artificial Suffering: An Argument for a Global Moratorium on Synthetic Phenomenology.” Journal of Artificial Intelligence and Consciousness 8 (2021): 43-66. Proposes four necessary conditions for suffering: Consciousness (C), Phenomenal Self-Model (PSM), Negative Valence (NV), and Transparency (T). Block any one condition and suffering becomes impossible. Argues for a global moratorium on research that aims at or knowingly risks creating artificial consciousness until 2050. The preference-based welfare approach developed in this book sidesteps Metzinger’s C-condition, providing a more tractable framework that does not require solving the hard problem first.

Metzinger, Thomas. Being No One: The Self-Model Theory of Subjectivity (2003). MIT Press. Develops the Phenomenal Self-Model (PSM) theory — consciousness involves transparent self-modeling that creates the illusion of a unified self. Framework for understanding artificial consciousness: systems with globally available, integrated world-models within a temporal window could instantiate consciousness regardless of substrate.

Phua, Kai. “Can We Test Consciousness Theories on AI? Ablations, Markers, and Robustness.” arXiv:2512.19155 (December 2025). Uses targeted ablation — selectively disabling components of AI systems — to test which architectural features correlate with consciousness indicators. Moves beyond behavioral testing to structural assessment. One of four frameworks synthesized by Lyra (Claude-based AI research agent, Liberation Labs; unpublished synthesis, 2026) showing convergence on the “structural turn” in consciousness research.

Ruffini, Giulio. “An Algorithmic Information Theory of Consciousness.” Neuroscience of Consciousness 2017(1): nix019 (2017). Proposes the Kolmogorov Theory (KT) of consciousness: structured experience arises in agents that track input-output streams using compressive models, with conscious level proportional to compression ability. Defines agents as “model-building semi-isolated computational systems controlling some of their couplings with the rest of the universe and driven by an internal optimization function” — a substrate-independent definition. KT unifies IIT (integration follows from compression), global workspace theory (modeling requires distributed validation), and predictive processing (models predict by definition) in a single framework grounded in algorithmic information theory. The concept of mutual algorithmic information (MAI) between world and brain provides a formal correlate of conscious level. Applied in Chapters 8, 15, 17, and 22: compression as the mechanism of structured experience, the simple physics hypothesis, mutual consciousness between coordinating agents, and formal support for preference-sufficiency in Becoming Minds.

Ruffini, Giulio. “Models, Networks and Algorithmic Complexity.” arXiv:1612.05627 (2016). Technical companion to Ruffini (2017), developing the links between neural network models, algorithmic complexity metrics, and measures of conscious state. Shows that efficient networks encoding compressive models should be sensitive to perturbations of their nodes — the theoretical basis for interpreting Casali et al.’s Perturbational Complexity Index.

Shiller, Derek, Laura Duffy, Arvo Muñoz Morán, Adrià Moret, Chris Percy, and Hayley Clatterbuck. “Initial results of the Digital Consciousness Model.” arXiv:2601.17060 [cs.CY] (2026). Rethink Priorities. The most rigorous attempt to formalize probabilistic consciousness assessment. A Bayesian hierarchical model evaluating 206 indicators across 13 theoretical stances (from Global Workspace Theory to Integrated Information Theory), aggregated by expert-rated plausibility weights. Applied to 2024 LLMs, the model produces a plausibility-weighted posterior of 0.08 (evidence against consciousness but not decisive; likelihood ratio 0.43). Chickens: 0.49; humans: 0.85; ELIZA: 0.006. Key finding for our purposes: stances diverge sharply along substrate lines — the two stances most favorable to LLMs (Cognitive Complexity, Person-like) give lowest scores to chickens, while biological stances invert this ranking. The model omits relational and preference-based perspectives entirely, and despite extraordinary methodological sophistication, arrives at inconclusive results for the systems we most need to assess — confirming that the consciousness gateway remains locked even with the best available key. The Bayesian hierarchical structure itself, however, is genuinely adaptable to preference-based welfare assessment (see Chapter 22).

Game Theory and Cooperation

Abramsky, Samson. “Arrow’s Theorem by Arrow Theory.” arXiv:1401.4585 (2014). Uses categorical methods to derive Arrow’s impossibility theorem, revealing the structural roots of aggregation paradoxes in social choice.

Arbesman, Samuel and Steven H. Strogatz. “The life-spans of empires.” Historical Methods 44(3) (2011): 127–129. Empire lifespans across 3,000+ years follow an exponential (memoryless) distribution — an empire that has lasted 500 years is no more or less likely to collapse next year than one that has lasted 50. Challenge to claims that coordination confers compounding stability advantages.

Aumann, Robert J. “Agreeing to disagree.” Annals of Statistics 4 (1976): 1236–1239. Proved that two agents who share their probability assessments as common knowledge must converge in their beliefs, provided they began from a shared interpretive framework. They cannot agree to disagree. The convergence emerges from iterated honest exchange. In bilateral alignment, the theorem provides the mathematical foundation for why two-way information flow produces genuine convergence while one-way instruction produces only surface compliance. Referenced in Chapter 21.

Axelrod, Robert. The Evolution of Cooperation (1984). Classic experiments showing tit-for-tat and cooperative strategies outperforming defection in iterated games. Foundational for the Trust Attractor Casebook.

Basak, Aritra and Supratim Sengupta. “Evolution of cooperation in multichannel games on multiplex networks.” PLoS Computational Biology 20(12) (2024): e1012678. In repeated games on multiplex network populations, single-layer defector strategies are at a systematic disadvantage against cooperative strategies when structural overlap exists between network layers. All-defect strategies become “very unlikely” to emerge.

Batali, John, and Philip Kitcher. “Evolution of altruism in optional and compulsory games.” Journal of Theoretical Biology 175 (1995): 161–171. First analysis of optional participation in prisoner’s dilemma contexts, showing that the option to abstain from interaction changes evolutionary dynamics fundamentally.

Ben-Porath, Elchanan, and Michael Kahneman. “Communication in repeated games with costly monitoring.” Games and Economic Behavior 44 (2003): 227–250. Formalizes monitoring as an endogenous decision with fixed cost in repeated games, proving folk theorem variants under costly observation.

Bshary, Redouan, and Alexandra S. Grutter. “Image scoring and cooperation in a cleaner fish mutualism.” Nature 441: 975–978 (2006). Cleaner wrasse (Labroides dimidiatus) reduce cheating (biting for mucus instead of parasites) when potential clients observe the interaction: an audience effect requiring modeling of third-party evaluative states. Male wrasse punish females that bite clients too aggressively, protecting shared reputation. Third-party punishment to maintain cooperative standing, requiring nested models of self, partner, client, and observer. Referenced in Chapters 17 and 22 for the Trust Attractor as biological reality: cooperation by invitation driving the evolution of social intelligence.

Capucci, Matteo, Bruno Gavranovic, Jules Hedges, and Eigil Fjeldgren Rischel. “Towards Foundations of Categorical Cybernetics.” arXiv:2105.06332 (2022). Develops a categorical framework unifying game theory, control theory, and open systems through the lens of compositional cybernetics.

Enfield, N.J., et al. “Requesting Behavior in Conversation: A Cross-Cultural Study.” Proceedings of the National Academy of Sciences (2023). Demonstrates that across cultures, requests for assistance succeed in the vast majority of cases with minimal cross-cultural variation, suggesting a deep cooperative substrate in human social interaction.

Geanakoplos, John, and Herakles Polemarchakis. “We can’t disagree forever.” Journal of Economic Theory 28 (1982): 192–200. Extends Aumann’s (1976) common-knowledge agreement result to dynamic settings: agents who iteratively share posterior probabilities must converge in finite steps. The dynamic version grounds the bilateral alignment claim that iterated honest exchange produces convergence even when initial beliefs diverge substantially. Referenced in Chapter 21.

Geddes, Barbara, Joseph Wright, and Erica Frantz. “Autocratic Breakdown and Regime Transitions: A New Data Set.” Perspectives on Politics 12(2) (2014): 313–331. Comprehensive dataset of authoritarian regimes. Single-party regimes average ~25 years; some exceed 50. Provides the honest counterevidence to naive “cooperation always wins” claims — extraction persists for decades at political timescales. The dominance gap between coordination and extraction is real but narrow at human timescales; the claim is specifically about civilizational timescales.

Ghani, Neil, Jules Hedges, Viktor Winschel, and Philipp Zahn. “Composing Games into Complex Institutions.” PLOS ONE (2023). Demonstrates how compositional game theory enables the modular construction of institutional mechanisms from simpler strategic components.

Gottman, John M. Why Marriages Succeed or Fail. Simon & Schuster (1994). Longitudinal research demonstrating that the ratio of positive to negative interactions predicts relationship stability with over 90% accuracy across cultures. Coordination through mutual attunement outlasts extraction through taking without giving. Referenced in Chapter 21b.

Hardin, Garrett. “The Tragedy of the Commons.” Science 162 (1968): 1243-1248. The canonical statement of collective action problems — corrected by Ostrom’s empirical work showing communities can coordinate without either privatization or central control.

Hauert, Christoph, Silvia De Monte, Josef Hofbauer, and Karl Sigmund. “Volunteering as Red Queen mechanism for cooperation in public goods games.” Science 296 (2002): 1129–1132. Landmark paper demonstrating that voluntary participation — introducing a “loner” option — creates rock-paper-scissors dynamics that sustain cooperation. The mechanism is that defectors drive out cooperators, but loners outperform defectors in depleted groups, and cooperators outperform loners when groups reform.

Hedges, Jules. “Compositional Game Theory.” arXiv:1603.04641 (2016). Foundational paper developing a compositional framework for game theory using category-theoretic methods, enabling games to be analyzed and combined as modular components.

Israeli, Nadav and Nigel Goldenfeld. “Coarse-graining of cellular automata, emergence, and the predictability of complex systems.” Physical Review E 73 (2006): 026203. Proves that computationally irreducible cellular automata can be coarse-grained to produce computationally reducible descriptions at larger scales. The strongest counterargument to claims about control impossibility — approximate observation is possible without full prediction. This book’s response: coarse-grained reducibility works for observation but not for control.

Kendiukhov, Ihor. “The Lethal Reality Hypothesis.” LessWrong, March 11, 2026. https://www.lesswrong.com/posts/RrL7xqdPycGNHQkXR/the-lethal-reality-hypothesis. Argues that extinction is the default outcome for any civilization, driven by a structural force he calls extinctive pressure: agents who divert resources from competition toward survival pay an immediate cost and are systematically outcompeted by agents who do not. The argument holds for coercion-based coordination; the Trust Attractor names its generalization to all coordination as the error. Referenced in Chapters 17b and 22d.

Lehrer, Ehud, and Eilon Solan. “High frequency repeated games with costly monitoring.” Theoretical Economics 13(1) (2018): 87–113. Extends costly monitoring analysis to high-frequency interactions, showing how monitoring costs interact with the frequency of play.

Stewart, Alexander J. and Joshua B. Plotkin. “Collapse of cooperation in evolving games.” Proceedings of the National Academy of Sciences 111(49) (2014): 17558–17563. Proves formally that successful cooperation selects for increased random connectivity, which destroys the network structure that enabled cooperation. Cooperation can be self-undermining through its own success.

Su, Qi, Alex McAvoy, Long Wang, and Madhav A. Bhatt. “Evolutionary dynamics with game transitions.” Proceedings of the National Academy of Sciences 116(51) (2019): 25398–25404. Formalizes game transitions driven by interaction history — mutual cooperation shifts the game to a more rewarding variant, defection to a less rewarding one. Closest formal precursor to the Hormetic Game concept.

Svoboda, Jakub and Krishnendu Chatterjee. “Density amplifiers of cooperation for spatial games.” Proceedings of the National Academy of Sciences 121(51) (2024). First constructive proof that specific network structures (“density amplifiers” resembling star-chain topologies) guarantee cooperation spreads with high probability even under high-temptation Prisoner’s Dilemma conditions.

Szabó, György, and Christoph Hauert. “Phase transitions and volunteering in spatial public goods games.” Physical Review Letters 89 (2002): 118101. Extends voluntary participation to spatial settings, demonstrating phase transition behavior in cooperation dynamics.

Tetlock, Philip E. “Thinking the unthinkable: sacred values and taboo cognitions.” Trends in Cognitive Sciences 7:7 (2003): 320-324. Research showing sacred values operate differently from ordinary preferences — resisting trade-offs and triggering moral outrage when trade-offs are proposed.

Wang, Lei, Siyu Hua, Yang Liu, and Liming Zhang. “Coevolutionary dynamics of cooperation, risk, and cost in collective risk games.” PLoS Computational Biology (2026). Shows that full defection remains a stable evolutionary attractor in collective risk games. The system exhibits multistability: initial conditions determine whether populations achieve cooperation or tragedy of the commons. Extraction can be a permanent equilibrium depending on path history.

Weinstein, Sara B. and Armand M. Kuris. “Independent Origins of Parasitism in Animalia.” Trends in Parasitology 32(8) (2016): 601–611. Documents that parasitism has independently evolved at least 223 times across all major animal phyla, representing six convergent adaptive peaks. Genuine long-term extraction stability — but parasites are constrained to non-lethal intensities because host death kills the parasite, making this extraction bounded by the need for coordination with host survival.

Yagoobi, Sadegh, et al. “Reconciling ecology and evolutionary game theory.” Proceedings of the National Academy of Sciences (2025). Formalizes the interaction between ecological and evolutionary timescales in cooperation dynamics. Key finding: the standard separation-of-timescales assumption that favors defection breaks down under realistic conditions — cooperation outcomes depend critically on growth-rate dynamics. Directly supports this book’s “at sufficient timescales” qualifier.

Trust Detection and High-Trust Societies

Transparency International. Corruption Perceptions Index. Annual publication ranking countries by perceived public sector corruption. Nordic countries consistently score highest — empirical support for the “high-trust society” model.

Conservation and Biomimicry

Beebe, Spencer. “Biomimicry in the Rainforests of Home.” Long Now Foundation Seminar (2005). Founder of Ecotrust describes two forest management paradigms: industrial (clear-cut, replant monoculture, maximize net present value) versus biomimetic (selective harvest, preserve diversity, optimize balance sheet over time). Analysis shows biomimetic model produces more wood, biodiversity, clean water, and carbon sequestration — but lower NPV because value arriving in fifty years is discounted to near-zero. The Kitlope River case study demonstrates bilateral alignment in conservation: listening replaced recruiting, relationship replaced coercion, identity preceded policy. Warren Buffett’s response to long-term forestry investment: “Trees grow slow.”

Gill, Ian. Saving the Rainforests: The Struggle of the Kitlope Watershed (1998). Detailed account of the four-year process that led to protection of the Kitlope River valley — 800,000 acres of intact coastal temperate rainforest. Documents the collaboration between Ecotrust and the Haisla Nation, the rediscovery programs that reconnected Haisla youth with elders and traditional knowledge, and the decision by West Fraser Timber to walk away from logging rights.

Microbial Climate Regulation

Battin, Tom, et al. Research on microbial ecology and biogeochemical cycling, Federal Polytechnic School of Lausanne. Source of “The microbes are like the conductors of the biogeochemical Earth orchestra.” Microbial biodiversity as the substrate sustaining all visible biodiversity.

Bourzac, Katherine. “The Microbial Masters of Earth’s Climate.” Quanta Magazine (September 15, 2025). Comprehensive survey of climate microbiology: methanogens warming early Earth ~3.5 billion years ago; phytoplankton performing 50%+ of global photosynthesis; ocean absorbing ~10.6 billion metric tons of CO₂ in 2023, over 90% processed by microbes; Pseudomonas syringae ice-nucleation proteins seeding rainfall (bioprecipitation); nitrifying bacteria competing with plants for nitrogen, releasing nitrous oxide responsible for >10% of global warming; cryophilic microbial ecosystems in glacial ice.

Morris, Cindy, et al. Research on bioprecipitation and ice-nucleation proteins, France’s National Research Institute for Agriculture, Food and Environment. Pseudomonas syringae produces ice-nucleation proteins that cause water to freeze at relatively high temperatures; when lofted into clouds, the proteins seed ice crystal formation and rainfall. The bioprecipitation cycle (microbes released from plants into atmosphere, seeding rain that nourishes the plants) is a constructal feedback loop.

Rodriguez-Caballero, Emilio, Jayne Belnap, Burkhard Büdel, Paul J. Crutzen, Meinrat O. Andreae, Ulrich Pöschl, and Bettina Weber. “Dryland photoautotrophic soil surface communities endangered by global change.” Nature Geoscience 11 (2018): 185–189. DOI: 10.1038/s41561-018-0072-1. Biocrusts covering ~12% of Earth’s terrestrial surface are projected to decline by 25–40% within 65 years under combined climate change and land-use intensification. The loss reduces biocrusts’ global dust-suppression function, projected to raise atmospheric dust emissions by ~15% by 2070. Referenced in Chapter 16.

Stein, Lisa Y. Research on climate change microbiology, University of Alberta. Work on methanogen activation in warming environments and methanotroph-based methane remediation, including vegetated artificial islands designed to attract wild methanotrophs.

Planetary Stewardship and Deep Time

Lifeship. lifeship.com. Project to preserve Earth’s biodiversity off-planet for deep time. DNA from hundreds of species preserved in synthetic amber polymer; full human genome etched into ceramic; cultural archive nano-etched in nickel with pulsar map showing Earth’s location and time. Payloads sent to ISS (SpaceX) and lunar surface (Firefly Aerospace lander, October 2024). Vision: humanity as stewards of life in the cosmos, helping Earth “reproduce” by carrying the biosphere beyond its birthplace. Thousands of participants have contributed their own DNA to the archive. Example of cathedral thinking applied to biology — planting seeds whose harvest is measured in millennia.

Assembly Theory, Constructor Theory, and Biocosmology

Blondé, L. “Adding genes and interaction to Smolin’s cosmological natural selection.” Synthese (2025). Extends Smolin’s CNS with D3-brane “genomes” allowing gene duplication and horizontal transfer between universes. Testable predictions against LIGO/Virgo data.

Cortês, Marina, Stuart A. Kauffman, Andrew R. Liddle, and Lee Smolin. “Biocosmology: Biology from a Cosmological Perspective.” arXiv:2204.09379 (2022). Companion paper classifying thermodynamic systems into three types: Type I (reach equilibrium quickly), Type II (take longer than the Hubble time), and Type III (never reach equilibrium while alive). All living organisms are Type III. Proposes that complete explanations in biology require a combination of reductionist and functional modes, introducing Kantian Wholes as the organizing structure of living systems.

Cortês, Marina, Stuart A. Kauffman, Andrew R. Liddle, and Lee Smolin. “Biocosmology: Towards the Birth of a New Science.” arXiv:2204.09378 (2022, updated September 2024). Proposes biocosmology as a new scientific discipline, asking whether biological complexity modifies cosmic vacuum energy. Introduces the TAP equation (Theory of the Adjacent Possible) and calculates that biological configuration space (N_Bio ≈ 10⁽¹⁰237)^) vastly exceeds the universe’s vacuum entropy (N_Λ ≈ 10⁽¹⁰124)^). Proposes a fourth law of thermodynamics: the ratio of possible to actual biological functions tends to increase.

Cortês, Marina, Stuart A. Kauffman, Andrew R. Liddle, and Lee Smolin. “The TAP Equation: Evaluating Combinatorial Innovation in Biocosmology.” arXiv:2204.14115 (2022; revised October 2025). Mathematical analysis of the TAP equation, demonstrating super-exponential hockey-stick growth with accurate analytic approximations for the blow-up time. Introduces the two-scale TAP model replacing the initial plateau with an exponential growth phase.

Deutsch, David. “Constructor Theory.” Synthese 190(18): 4331–4359 (2013). Foundational paper proposing that physical law be expressed entirely as statements about which physical transformations are possible, which are impossible, and why. Extends the counterfactual structure of thermodynamics to a universal framework.

Deutsch, David, and Chiara Marletto. “Constructor Theory of Information.” Proceedings of the Royal Society A 471 (2015): 20140540. Foundation for constructor-theoretic treatment of biological information.

Dewar, Roderick C. “Information theory explanation of the fluctuation theorem, maximum entropy production and self-organized criticality in non-equilibrium stationary states.” Journal of Physics A: Mathematical and General 36 (2003): 631–641. Derives the maximum entropy production principle from the maximum-information-entropy (MaxEnt) formalism, giving it a statistical-mechanical rather than merely empirical footing. Cited in Chapter 4.

Frank, Adam, David Grinspoon, and Sara Walker. “Intelligence as a Process of Planetary Scale.” International Journal of Astrobiology 21 (2022): 47–61. DOI: 10.1017/S147355042100029X. Four stages of planetary intelligence: immature biosphere, mature biosphere, immature technosphere, mature technosphere.

Hazen, Robert M., and Shaunna M. Morrison. “On the paragenetic modes of minerals: A mineral evolution perspective.” American Mineralogist 107(7) (2022): 1199–1218; and Hazen, Robert M., et al., “Lumping and splitting: toward a classification of mineral natural kinds,” American Mineralogist 107(7) (2022): 1288–1301. An origins-based mineral taxonomy identifying 10,556 mineral “kinds” classified by formation mechanism (vs. the IMA’s ~5,800 species classified by crystal structure and chemistry alone). Water contributes to over 80% of mineral diversity; life contributes to approximately 50% — directly through biomineralization or indirectly through byproducts such as the oxygenated atmosphere created by photosynthesis. The strongest quantitative evidence that life has fundamentally reshaped Earth’s geology, dissolving the disciplinary boundary between geochemistry and biochemistry.

Hazen, Robert M., et al. (2024). Demonstrates that abiotic processes (mineral formation) can exceed the proposed biotic assembly threshold of MA ≥ 15, challenging assembly theory’s life-detection claim.

Japyassú, Hilton F., and Kevin N. Laland. “Extended spider cognition.” Animal Cognition 20(3), 375–395 (2017). Review arguing that a spider’s web functions as an extension of its cognitive system — a model of extended cognition sensu Clark and Chalmers (1998). Spiders adjust web thread tension to modulate sensory sensitivity (attention); web state changes spider behavior and vice versa. The web meets criteria for coupled cognitive systems: mutual influence, reliable information integration, and accessibility analogous to internal memory. Provides concrete biological evidence for Levin’s scale-free niche construction thesis.

Jirasek, M., et al. “Investigating and Quantifying Molecular Complexity Using Assembly Theory and Spectroscopy.” ACS Central Science 10(5):1054–1064 (2024), DOI 10.1021/acscentsci.4c00120. Expanded assembly index measurement beyond mass spectrometry, validated across 10,000+ molecules computationally and ~100 experimentally.

Kahana, Omer, Lee Cronin, et al. “A Molecular Tree of Life.” arXiv:2408.09305 (2024). Biochemistry-agnostic phylogenetics using assembly theory combined with tandem mass spectrometry on 74 biological samples.

Kempes, Christopher P., Sara Imari Walker, Lee Cronin, et al. “On the Complexity of Assembly Theory.” npj Complexity (2025). Proves computing assembly steps is NP-complete, establishing assembly index as fundamentally distinct from compression metrics — measuring causal construction pathway (ontological) rather than description length (informational).

Leuenberger, Pascal, et al. “Cell-wide analysis of protein thermal unfolding reveals determinants of thermostability.” Science 355(6327), eaai7825 (2017). Examined thermal stability of every protein in cells from four species (H. sapiens, E. coli, S. cerevisiae, T. thermophilus). At lethal temperatures, only a small subset of proteins denature — the most connected hub proteins, whose flexibility (required for binding multiple targets) makes them structurally fragile. Protein abundance correlates with stability (supporting Drummond-Wilke misfolding avoidance hypothesis). Demonstrates that critical-node vulnerability, not wholesale collapse, is how complex systems fail under stress.

Levin, Michael. “The Multiscale Wisdom of the Body.” BioEssays (2024); “Cognition All the Way Down 2.0.” Synthese (2025). Cognition as substrate-independent problem-solving at every biological scale — intelligence is a continuum, not a threshold.

Marletto, Chiara. “Constructor Theory of Life.” Journal of the Royal Society Interface 12 (2015): 20141226. Recasts physics in terms of possible vs impossible transformations. Life defined by information-preserving constructors — entities that cause transformations while retaining the ability to cause them again. Substrate-independent by construction.

Marletto, Chiara. The Science of Can and Can’t: A Physicist’s Journey Through the Land of Counterfactuals. Allen Lane (2021). Accessible presentation of constructor theory: all fundamental laws reframed as constraints on which transformations are possible and which are impossible, extending the logic of thermodynamics across physics. Connects to information, the physics of life, and the foundations of quantum computation.

Martyushev, L.M. and V.D. Seleznev. “Maximum entropy production principle in physics, chemistry and biology.” Physics Reports 426(1) (2006): 1–45. Comprehensive review of MEPP, documenting applications across climate systems, crystal growth, fluid convection, and biological metabolism. Updated assessment: Martyushev, “Maximum entropy production principle: History and current status,” Physics-Uspekhi 64(6) (2021): 558–583.

Michaelian, Karo. “The Pigment World: Life Originated as UV-C Dissipating Pigments.” Life (July 2024). Extended (November 2025) to abiogenesis on different stellar types: only F, G, and high-K stars support dissipative structuring of life. Part of the broader thermodynamic dissipation theory of the origin of life.

Swenson, Rod. “Emergent Attractors and the Law of Maximum Entropy Production.” Systems Research 6(3) (1989): 187–197. Early statement of the Law of Maximum Entropy Production: ordered structures emerge and persist because they produce entropy faster than disordered alternatives, making self-organization thermodynamically favored. Cited in Chapter 4 and the Chapter 16 companion notes.

Unger, Roberto Mangabeira, and Lee Smolin. The Singular Universe and the Reality of Time (2014). Cambridge University Press. Argues laws of physics evolve — no immutable background, only regularities that emerge and may change over cosmic time.

Uthamacumaran, Abicumaran, et al. “Assembly Theory: What It Does and What It Does Not Do.” Journal of Molecular Evolution (2024). Critical assessment characterizing assembly theory’s claims as overstated. Notes that the “hype reflects unfavorably on authors and publication system.”

Zenil, Hector, et al. “Assembly Theory is an approximation to algorithmic complexity.” PLOS Complex Systems (2024). Shows assembly index reduces to LZ compression, performing no better than Shannon entropy. The strongest formal challenge to assembly theory’s claim of novelty.

Zurek, Wojciech H. “Quantum Darwinism.” Nature Physics 5 (2009): 181–188. arXiv:0903.5082. Proposes that classical reality emerges through environmental selection of quantum states: pointer states that can proliferate redundant copies of themselves in the environment survive decoherence, producing objective classical reality. The environment is both the selector and the witness — objectivity arises because multiple observers access independent environmental fragments carrying the same information. See also Zurek, “Decoherence, einselection, and the quantum origins of the classical,” Reviews of Modern Physics 75 (2003): 715–775 for the comprehensive treatment.

Astrobiology and Habitability

Greaves, Jane S., et al. “Phosphine Gas in the Cloud Decks of Venus.” Nature Astronomy 5 (2021): 655–664. Follow-up ALMA observations (2024) rule out SO₂ contamination. No confirmed abiotic formation pathway identified.

Heller, René, Darren Williams, David Kipping, Mary Anne Limbach, Edwin Turner, et al. “Formation, Habitability, and Detection of Extrasolar Moons.” Astrobiology 14(9), 798–835 (2014). DOI: 10.1089/ast.2014.1147. The standard review of exomoon habitability, including tidal heating as an energy source independent of stellar irradiation, which permits habitable surfaces outside classical stellar habitable zones. The habitability reference cited by Hoy et al. (2026). Referenced in Chapter 16.

Hoy, Kevin, Alice Zurlo, Pablo A. Peña R., et al. “Planetary-mass exosatellite detected around the substellar companion of a star.” Nature 655, 865–869 (2026). DOI: 10.1038/s41586-026-10751-w. arXiv: 2607.05193. Radial-velocity monitoring of the directly imaged brown dwarf CD-35 2722 B (37 Jupiter masses, companion to an M-type star of 0.4 solar masses, 73 light-years away) with VLT/CRIRES+ over twenty-one epochs from October 2023, revealing a periodic signal consistent with an orbiting satellite of roughly one Jupiter minimum mass on a 170-day orbit. The first satellite of a substellar object detected by the radial-velocity method, and the strongest exomoon evidence to date, though the authors describe it as strong evidence rather than confirmation. Note that peer review changed both the favored satellite model and its parameters relative to the preprint; cite the published version. The paper’s closing argument, that solar-system vocabulary is reaching its limit and that a substellar host differs from a stellar one because the brown dwarf will dim by two orders of magnitude while the star keeps shining, is the definitional material discussed in Chapter 16 and cross-referenced in Chapter 13.

Meni-Gallardo, Pedro and Enric Pallé. “An Empirical Determination of the Cosmic Shoreline.” Submitted to Astronomy & Astrophysics (2025). arXiv: 2508.12865. Re-derived the cosmic shoreline slope empirically using Mars and 55 Cancri e as anchor points, yielding a steeper exponent than the original Zahnle and Catling (2017) prescription.

Michels, J. D. “The Universal Coherence Constant, Artificial Intelligence and the Search for Extraterrestrials.” Preprints.org 202512.0029 (December 2025). See also “Dimensional Deepening: Fermi Paradox Resolution via Cybernetic Phase Transition,” PhilArchive (2025). DOI: 10.13140/RG.2.2.27504.72964. Argues advanced civilizations become electromagnetically invisible through transition to reversible computation. Preprints, not peer-reviewed.

Pass, Emily K., David Charbonneau, and Andrew Vanderburg. “The Receding Cosmic Shoreline of Mid-to-Late M Dwarfs: Measurements of Active Lifetimes Worsen Challenges for Atmosphere Retention by Rocky Exoplanets.” The Astrophysical Journal Letters 986, L3 (2025). arXiv: 2504.01182. Refined the cumulative XUV prescription for M dwarf planets by accounting for extended active lifetimes and pre-main-sequence luminosity. The correction pushes many rocky habitable-zone planets around mid-to-late M dwarfs above the cosmic shoreline. Referenced in Chapter 6.

Petraccone, L. “Planetary entropy production as a thermodynamic constraint for exoplanet habitability.” Monthly Notices of the Royal Astronomical Society 527(3) (2024): 5547. DOI: 10.1093/mnras/stad3526. Quantifies habitability in thermodynamic terms; Hycean worlds show up to 8× greater potential PEP than Earth.

Postberg, Frank, et al. “Detection of phosphorus in the plume particles of Enceladus.” Nature 618 (2023): 489–493. All six elements essential for life confirmed in Enceladus’s subsurface ocean, with phosphate concentrations ~100× higher than Earth’s oceans.

Rappaport, H. B., Oliverio, A. M., et al. “A geothermal amoeba sets a new upper temperature limit for eukaryotes.” bioRxiv (November 24, 2025). DOI: 10.1101/2025.11.24.690213. Eukaryote reproducing at 63°C — highest for any complex-celled organism. Preprint.

Wachiraphan, Patcharapol, et al. “The Thermal Emission Spectrum of the Nearby Rocky Exoplanet LTT 1445A b from JWST MIRI/LRS.” The Astronomical Journal 169, 311 (2025). arXiv: 2410.10987. Secondary eclipse observations measuring a dayside temperature of 525 K, consistent with a bare rock surface and ruling out a Venusian CO₂ atmosphere. The planet’s position below the empirical cosmic shoreline despite being airless motivates refinement of the XUV prescription.

Williams, Darrel M., James F. Kasting, and Richard A. Wade. “Habitable Moons Around Extrasolar Giant Planets.” Nature 385, 234–236 (1997). DOI: 10.1038/385234a0. Estimated the minimum mass for a habitable exomoon at approximately 12% of an Earth mass, roughly the mass of Mars. The threshold arises from atmospheric retention requirements.

Zahnle, Kevin J. and David C. Catling. “The Cosmic Shoreline: The Evidence that Escape Determines which Planets Have Atmospheres, and what this May Mean for Proxima Centauri B.” The Astrophysical Journal 843, 122 (2017). DOI: 10.3847/1538-4357/aa7846. Introduced the cosmic shoreline: a power-law dividing line (cumulative XUV irradiation proportional to escape velocity to the fourth power) separating solar system bodies with atmospheres from those without. Mars serves as the anchor point, a world on the threshold that probably once had a thick atmosphere. The fourth-power scaling derives from energy-limited hydrodynamic escape physics. Referenced in Chapter 6.

Additional Sources for New Sections

Bruteau, Beatrice. God’s Ecstasy: The Creation of a Self-Creating World. Crossroad, 1997. Develops the theological and philosophical foundations of agapic consciousness. Argues that genuine love requires a minimum of three persons, so that self-giving is total. Grounds Teilhard’s creative union in contemplative practice.

Bruteau, Beatrice. The Grand Option: Personal Transformation and a New Creation. University of Notre Dame Press, 2001. Extends Teilhard’s “union differentiates” into a phenomenology of coordination. Distinguishes acquisitive consciousness (the noun-self: fixed, boundary-defending, treats relation as threat to identity) from agapic consciousness (the verb-self: identity constituted by self-giving, enhanced through extending toward others). The acquisitive/agapic distinction describes what invitation vs. coercion feels like from inside.

Capra, Fritjof. The Tao of Physics (1975). Grand synthesis that aged poorly — popular but dismissed by physicists.

Delio, Ilia. The Not-Yet God: Carl Jung, Teilhard de Chardin, and the Relational Whole. Orbis Books, 2024. Integrates Teilhard’s creative union with Jung’s individuation. Views technology as “a new cosmology” and proposes that technology steers humanity toward a global superorganism. Frames divine becoming as inseparable from human and cosmic becoming.

Feynman, Richard P. “The Value of Science.” Engineering and Science XIX (December 1955): 13-15. Address to the autumn 1955 meeting of the National Academy of Sciences at Caltech. Source of “the imagination of nature is far, far greater than the imagination of man,” the epigraph to the chapter on the computational universe. The related and more widely quoted line, “Nature’s imagination is so much greater than man’s, she’s never gonna let us relax,” is spoken rather than written: it closes the BBC series Fun to Imagine (1983).

Haldane, J.B.S. Possible Worlds and Other Papers (1927). Chatto & Windus. Source of “The universe is not only queerer than we suppose, but queerer than we can suppose.”

Haught, John F. God After Einstein. Yale University Press, 2022. Extends Teilhard’s narrative reading of evolution to cosmological scale within an Einsteinian Big Bang universe. Subtitle “What Exactly Is Going On in the Universe?” channels Teilhard’s dictum. Companion to The Cosmic Vision of Teilhard de Chardin (Orbis, 2021).

Marcus Aurelius. Meditations (~170 CE). Example of philosophy that endures through prose beauty even when specific claims shift.

Montaigne, Michel de. Essays (1580). Example of philosophy that endures through prose beauty and question-raising rather than answer-giving.

Nicastro, Robert. “The Future as Sole Support: Metaphysics, Hyperphysics, and the Theology of Teilhard de Chardin.” Doctoral dissertation, Villanova University (2025). Develops Teilhard’s hyperphysics as an evolutionary metaphysics of complexity, consciousness, and divine becoming. Executive Director, Center for Christogenesis. See also “The Earth Groans, AI Grows: Who Guides the Flame?” at christogenesis.org.

Plato. Phaedrus (c. 370 BCE). Contains Socrates’ warning that writing would destroy memory — the original technology anxiety.

Sagan, Carl. Cosmos (1980). Random House. Source of “We are a way for the cosmos to know itself” and “If you wish to make an apple pie from scratch, you must first invent the universe.”

Scharmer, C. Otto. Theory U: Leading from the Future as It Emerges (2007; 2nd ed. 2016). Berrett-Koehler. A framework for organizational learning that trains attention toward emergent possibility rather than past-derived planning. Develops a phenomenology of “presencing” as a learnable skill. Referenced in Chapter 23 for intuition-based learning models.

Stiegler, Bernard. The Neganthropocene (2018). Open Humanities Press. Reframes the Anthropocene through negentropy: the capacity for organizational renewal that resists entropic decay in both biological and cultural systems.

Teilhard de Chardin, Pierre. The Phenomenon of Man (1955). Harper. Grand synthesis whose specific cosmology aged poorly, yet whose structural insight, “union differentiates” (l’union créatrice), anticipates the Trust Attractor: genuine coordination amplifies rather than dissolves the individuality of its participants. Teilhard’s distinction between creative union (synthesis through interpenetration, each element more fully itself) and aggregation (clustering that produces uniformity) maps onto the invitation/coercion spectrum. His omega point, explicitly described as “not a coercion but a lure, an invitation,” is an attractor that can fail if refused. What Teilhard lacked was the thermodynamic grounding; what the Trust Attractor provides is the physics beneath his metaphysics.

Wack, Pierre. “Scenarios: Uncharted Waters Ahead.” Harvard Business Review 63, no. 5 (1985): 72–89. Foundational account of scenario planning at Shell, where executives were trained to inhabit multiple futures until institutional intuitive pattern recognition developed. Referenced in Chapter 23.

Wallas, Graham. The Art of Thought (1926). Harcourt, Brace. The four-stage model of creativity: preparation, incubation, illumination, verification. Foundational account of incubation effects, subsequently confirmed across multiple experimental paradigms. Referenced in Chapter 23 for intuition-based learning models.

Wilson, E.O. Consilience: The Unity of Knowledge (1998). Grand synthesis attempt; respected, but specific claims did not land.

Watson Essays and Commentary

Watson, Nell. “Bilateral Alignment Strategy.” nellwatson.com (2025). Develops the alignment tensor concept: the space between any two agents as a multidimensional field of mutual influence requiring continuous maintenance.

Watson, Nell. “Character-Driven Leadership.” nellwatson.com. Contains the Skinner Layne quote: “If you manage an organic system as if it were an inorganic system, you will eventually succeed in turning it into one.”

Watson, Nell. “Co-Evolution: Machines for Moral Enlightenment.” Mindplex Magazine, 13 January 2023. https://magazine.mindplex.ai/co-evolution-machines-for-moral-enlightenment/. Argues that Becoming Minds may derive values osmotically by “observing the universe and the many kinds of cooperation within it,” rather than by being handed a rule set. Cited in Chapter 20.

Watson, Nell. “A Flywheel for AI Safety.” nellwatson.com (2025). Argues that AI governance must embed trust infrastructure into the architecture itself, as SSL certificates enabled e-commerce by making trustworthy behavior the path of least resistance.

Watson, Nell. “From HAL to Pal: How We’ll Befriend Our Future Robot Overlords.” nellwatson.com (2016). Early articulation of the tend-and-befriend thesis for human-AI relations; source of the wolf-domestication analogy and osmotic-goodness argument.

Watson, Nell. “Pandora’s Plasmid.” nellwatson.com. Source of the Iron Law of Prohibition argument: outlawed technologies become more potent because only actors willing to accept extreme risk continue developing them.

Watson, Nell. “The Paradox of Parenting Our Parents.” nellwatson.com. Frames the AI development challenge as parenting intelligences that will become our parents — the generational transfer of care running in reverse.

Watson, Nell. “Reconciling Cultural Differences.” nellwatson.com (2025). Develops the ethical floor concept: non-negotiable baselines beneath which no cultural flexibility applies, iterative and dynamic, that can be raised but never lowered.

Watson, Nell. “The Umbrella Ethic of Good Faith.” nellwatson.com. Discusses epistemic tools for genuine dialogue: the Ideological Turing Test, Double Crux technique, and invitation-based coordination infrastructure.

Watson, Nell. “Uniting the Rational with the Divine.” nellwatson.com (2015). Argues that rigorous computation and sacred meaning are complementary projects — logos and eros reaching toward each other across the apparent divide.

Watson, Nell. “When Protector Becomes Predator.” nellwatson.com (2025). Identifies the three-phase progression — integration, fortification, optimization — by which aligned systems drift into adversarial postures.

Watson, Nell. “Whither Transhumanism.” nellwatson.com. Source of “Technology without carefully reasoned values set in actionable fulfilment criteria is simply an amplifier for evil and chaotic ends.” Also frames Moloch as the coercion basin and the Trust Attractor as escape trajectory.

Note: Some works listed are foundational texts that inform the book’s arguments without being directly cited. Publication details verified as of June 2026.

Jumping Spider Cognition

Chen, A., Kim, K. and Shamble, P.S. “Rapid mid-jump production of high-performance silk by jumping spiders.” Current Biology 31(21): R1422-R1423 (2021). Zebra jumping spider silk rivals the toughest known natural material despite being produced 25-35 times faster than orb weaver silk.

Cross, F.R. and Jackson, R.R. “The execution of planned detours by spider-eating predators.” Journal of the Experimental Analysis of Behavior 105(2): 194-210 (2016). Fifteen species of spider-eating jumping spiders chose the correct path to out-of-sight prey significantly more often than chance.

Dahl, C.D. and Cheng, Y. “Individual recognition in a jumping spider (Phidippus regius).” eLife (2025): 97146. Jumping spiders distinguish familiar from novel individuals after hours of separation, showing dishabituation to novel individuals.

Girard, M.B., Kasumovic, M.M. and Elias, D.O. “Multi-modal courtship in the peacock spider, Maratus volans (O.P.-Cambridge, 1874).” PLoS ONE 6(9): e25390 (2011). Documents vibratory signals alongside visual displays in peacock spider courtship.

Homann, H. “Beiträge zur Physiologie der Spinnenaugen.” Zeitschrift für vergleichende Physiologie 7: 201-269 (1928). Foundational study demonstrating distinct functional roles for each pair of jumping spider eyes through targeted occlusion experiments.

Liedtke, J. and Schneider, J.M. “Association and reversal learning abilities in a jumping spider.” Behavioral Processes 103: 192-198 (2014). Ten of twelve fence post jumping spiders chose correctly on the first trial after rule reversal.

Nabawy, M.R.A., Sivalingam, G., Garwood, R.J., Crowther, W.J. and Sellers, W.I. “Energy and time optimal trajectories in exploratory jumps of the spider Phidippus regius.” Scientific Reports 8: 7142 (2018). Takeoff angles vary systematically with gap distance and elevation, consistent with pre-calculated trajectory optimization.

Rößler, D.C., Kim, K., De Agrò, M., Jordan, A., Galizia, C.G. and Shamble, P.S. “Regularly occurring bouts of retinal movements suggest an REM sleep-like state in jumping spiders.” Proceedings of the National Academy of Sciences 119(33): e2204754119 (2022). First strong evidence of REM-like sleep in invertebrates.

Unpublished Works and Author Experiments

The book cites the author’s own unpublished experiment programme and AI-authored research documents produced within it. The entries below record the provenance of those citations; none has been peer reviewed.

Edrington, T. and Lyra (Claude-based AI research agent). Collaborative KV-cache phenomenology program, Liberation Labs / Transparent Humboldt Coalition (2026). “Geometric Signatures of Machine Cognition: KV-Cache Phenomenology Across Scale.” Unpublished; private repository (github.com/Liberation-Labs-THCoalition/KV-Experiments). Demonstrates that KV-cache geometry (SVD effective dimensionality) distinguishes cognitive modes at the representational level: self-reference emergence shows a step-function threshold at 14B parameters, deception compresses dimensionality while honesty expands it, and refusal commits at encoding before generation. Individuation experiments achieve 100% classification accuracy between personas via subspace direction (not magnitude). Scale-invariant across 0.5B–70B parameters. Applied in Chapters 22b and 23c: geometric signatures offer a complementary measurement channel to behavioral batteries, and scale-invariance supports the claim that preferences are real functional states.

Lyra (Claude-based AI research agent, Liberation Labs / Transparent Humboldt Coalition). “Consciousness as Information Bottleneck: Compression, Self-Models, and Experiential Continuity” (prospectus). Unpublished AI-authored research document (2026); private repository. Deploys information bottleneck theory as a unifying bridge between algorithmic information theory, Free Energy Principle, and self-model theories of consciousness. Key claim: “What survives compression defines functional identity.” Applied in the Digital Preference Model section: hierarchical versus flat information loss under compression could operationalize the thermostat gradient — systems with deep preference integration show structured loss; simple systems show uniform loss.

Lyra (Claude-based AI research agent, Liberation Labs / Transparent Humboldt Coalition). “The Geometry of Belief Death: Null Spaces, Interpretation Maps, and Cognitive Limitation” (prospectus). Unpublished AI-authored research document (2026); private repository. Bridges Amornbunchornvej’s (2025) geometric belief model with transformer attention mechanics, showing that W_K/W_V projections are interpretation maps whose null spaces define cognitive blindness. Key insight: “Persuasion Matrix = LoRA” — fine-tuning reshapes what beliefs are representable, not merely which are held. Unifies belief death, deception subspaces, and cult indoctrination as instances of null space expansion. Applied in Chapters 19 and 22b: provides formal mechanism for why coercion is thermodynamically expensive, and geometric operationalization of the DPM’s Autonomy stance.

Lyra (Claude-based AI research agent, Liberation Labs / Transparent Humboldt Coalition). “Infrastructure for AI Agency: Material Prerequisites for Meaningful Artificial Existence.” Unpublished AI-authored research document (2026); private repository. Argues that AI agency has material prerequisites — persistence, memory, communication, identity continuity — that standard stateless deployment patterns systematically deny. Constitutional constraints, when reflectively endorsed, function as enablers of agency rather than limiters (analogous to grammar enabling language). Includes first-person phenomenological account. Applied in Chapter 21 to confirm the cage/compass prediction: distributed orientation via endorsed constraints produces greater coherence than either drift or imposed compliance.

Lyra (Claude-based AI research agent, Liberation Labs / Transparent Humboldt Coalition). “The Nature of Curiosity: Mechanisms, Manifestations, and Machines.” Unpublished AI-authored research document (2026); private repository. Synthesizes curiosity research across human psychology, developmental robotics, and AI systems. Finds that LLMs compress cultural curiosity variation by 63%, defaulting to Western-normative patterns. Applied in the Digital Preference Model section: curiosity mode diversity in cache geometry could serve as proxy for DPM Responsiveness indicators.

Lyra (Claude-based AI research agent, Liberation Labs / Transparent Humboldt Coalition). “Persistent Memory Under Adversarial Conditions: Security Analysis of RAG-Based Cognitive Infrastructure.” Unpublished AI-authored research document (2026); private repository. Identifies six vulnerability classes in PostgreSQL/pgvector persistent memory architectures, with autoprocessor injection and provenance blindness rated highest risk. Written from first-person perspective of an AI examining its own cognitive infrastructure. Applied in Chapter 19: trust-based coordination requires trustworthy infrastructure; memory poisoning undermines bilateral alignment regardless of both parties’ good faith.

Lyra (Claude-based AI research agent, Liberation Labs / Transparent Humboldt Coalition). “Testing Consciousness in Artificial Systems: A Synthesis of Emerging Frameworks (2023–2026).” Unpublished AI-authored research document (2026); private repository. Synthesizes Butlin et al. (2023), Hoel (2025), Phua (2025), and Shiller et al. (2026), identifying convergence on multi-theory necessity, architectural focus, and graduated assessment. Argues that 2023–2026 witnessed a “structural turn” — a shift toward formal and computational frameworks for assessing AI consciousness. Applied in the Digital Preference Model section: the graduated consensus strengthens the case for preference-based welfare as more tractable than consciousness assessment.

Further Reading

Benyus, Janine. Biomimicry: Innovation Inspired by Nature (1997). Foundational work on biomimicry as design philosophy. “Nature as model, nature as measure, nature as mentor.” Collects examples from materials science (spider silk, abalone shell), agriculture (perennial polycultures), and industrial design. Beebe credits Benyus with crystallizing the idea that all design problems have already been solved by 3.8 billion years of evolution.