knowm.ai

thermodynamic-computing · kt-bit · self-organization · reference

The kT-bit Reading List

The thinkers — a chemist, a physicist, a systems theorist, an ecologist, a roboticist — who circled the same primitive the kT-bit names, and where to find their work.

By Alex Nugent ·

Some of these are named in the thermodynamic-bit writing, the catalog, or the comment thread on the original post; others I’ve added from the surrounding literature. The common thread is free energy flowing through adaptive pathways that compete for the flow. Edward O. Wilson had a word for fields this far apart arriving at the same doorstep — consilience, “a jumping together of knowledge by the linking of facts and fact-based theory across disciplines” (Consilience: The Unity of Knowledge, 1998). This list is a roll call of it.


I. Foundations — order from dissipation#

The idea that an energy flow can build and maintain order, not only destroy it.

ThinkerWorkHow it applies
Erwin SchrödingerWhat Is Life? (1944)An organism resists decay by feeding on “negative entropy” — importing order from its surroundings. An early clear statement that life is something an energy flow sustains, not a static structure.
Ilya PrigogineNobel 1977; From Being to Becoming (1980); Order Out of Chaos (with Isabelle Stengers, 1984)Dissipative structures: matter pushed from equilibrium spontaneously organizes around the energy flowing through it, and the order vanishes the moment the flow stops.
Isabelle StengersOrder Out of Chaos (with Prigogine, 1984)Co-author of the book that carried dissipative structures to a general audience.
Alan TuringThe Chemical Basis of Morphogenesis” (1952)A homogeneous, symmetric chemical system is unstable: a small disturbance lets one mode grow at the expense of the others until a single stable pattern is left. Symmetry-breaking by competition.

II. Selection & extremal principles#

If many structures could form, which one does? This crowd answers with an extremal principle — the system self-organizes to drain the gradient as fast as the constraints allow.

ThinkerWorkHow it applies
Alfred J. LotkaContribution to the Energetics of Evolution,” PNAS 8, 147 (1922)The deepest root. Argued natural selection favors whatever captures and dissipates available energy fastest.
Rod SwensonLaw of Maximum Entropy Production (rodswenson.com)A system selects the path, or assembly of paths, that drains the potential at the fastest rate the constraints allow.
Roderick Dewar, Garth Paltridge, Hans ZieglerMaximum Entropy Production PrinciplePaltridge applied MEP to climate, Ziegler to continuum mechanics, Dewar attempted a statistical (MaxEnt) justification of it.
Jeremy Englanddissipation-driven adaptation (englandlab.com); “Statistical physics of self-replication,” J. Chem. Phys. (2013); Every Life Is on Fire (2020)Drive matter hard and long enough and it tends to rearrange into forms that dissipate the drive better.
Adrian Bejanconstructal law; Design in Nature (2012); The Physics of Life (2016)“For a finite-size system to persist in time (to live), it must evolve in such a way that it provides easier access to the imposed currents that flow through it.”
Sven Erik JørgensenTowards a Thermodynamic Theory for Ecological Systems (2004); Eco-Exergy as Sustainability (2006)Ecology framed thermodynamically: ecosystems develop toward the fastest rate of exergy storage and dissipation.
Howard T. OdumEnvironment, Power, and Society (1971); emergyThe maximum power principle: self-organizing systems tend toward the designs that maximize useful power throughput.
Robert E. UlanowiczGrowth and Development (1986); A Third Window (2009)Ascendency: ecosystems develop by routing more throughput through more organized flow networks.
Eric D. Schneider & Dorion SaganInto the Cool: Energy Flow, Thermodynamics, and Life (2005)“Nature abhors a gradient” — life and structure are the means by which nature degrades energy gradients faster.

III. Nonequilibrium & stochastic thermodynamics#

The rigorous machinery of driven, fluctuating systems.

ThinkerWorkHow it applies
Lars Onsagerreciprocal relations (1931); Nobel 1968The near-equilibrium foundation for coupled flows and forces.
Gavin CrooksCrooks fluctuation theorem, Phys. Rev. E 60, 2721 (1999) (threeplusone.com)Relates the entropy a driven system produces to the odds of forward versus reverse trajectories — a formal footing for how matter behaves when pushed away from equilibrium.
Christopher JarzynskiJarzynski equality, Phys. Rev. Lett. (1997)Relates nonequilibrium work to equilibrium free-energy differences — Crooks’s partner result.
Udo Seifertstochastic thermodynamics, Rep. Prog. Phys. 75 (2012)Entropy and the second law defined along single fluctuating trajectories — the physics of one small driven system, not an ensemble.

IV. Information & thermodynamics#

The thermodynamic cost of information — the link between entropy and what a system measures, stores, or erases.

ThinkerWorkHow it applies
Leó Szilardthe Szilard engine (1929)Reduced Maxwell’s demon to one bit of information and its thermodynamic price — the first link between information and entropy.
Léon BrillouinScience and Information Theory (1956)The negentropy principle of information: acquiring information costs free energy.
Rolf LandauerIrreversibility and Heat Generation in the Computing Process,” IBM J. (1961)Landauer’s principle: erasing a bit dissipates at least kT ln 2 of heat — the thermodynamic floor under discarding information.
Charles H. BennettThe Thermodynamics of Computation,” Int. J. Theor. Phys. (1982)Reversible computation; resolved Maxwell’s demon by charging for erasure, not measurement.
Takahiro Sagawa & Masahito Uedainformation thermodynamicsThe second law generalized to feedback-controlled (demon-like) systems.
Susanne StillThermodynamics of Prediction,” Phys. Rev. Lett. 109, 120604 (2012)A system that models its environment inefficiently wastes free energy — predictive structure earns its keep thermodynamically.

V. Self-organization, criticality & cybernetics#

The systems-theory lineage that formalized structure emerging from flow and feedback.

ThinkerWorkHow it applies
W. Ross AshbyDesign for a Brain (1952); An Introduction to Cybernetics (1956)The homeostat and the law of requisite variety — self-organization toward stable states under perturbation.
Heinz von Foersterthe “order from noise” principleStructure built up by fluctuations rather than destroyed by them.
Hermann HakenSynergetics: An Introduction (1977)The “slaving principle”: as a system self-organizes, a few order parameters come to govern the rest. A vortex’s single surviving spin is a textbook order parameter.
Humberto Maturana & Francisco VarelaAutopoiesis and Cognition (1980)Living systems as networks that continuously produce and maintain themselves through flow.
Philip W. AndersonMore Is Different,” Science (1972)Emergence and broken symmetry as the real content of complexity.
Per BakHow Nature Works (1996)Self-organized criticality — driven systems tune themselves to a critical point where a small trigger can cascade at any scale.

VI. Pattern formation & morphological instability#

The mechanisms that turn a smooth front or a uniform medium into branches, fingers, and cells.

ThinkerWorkHow it applies
Lord Rayleigh & Henri Bénardthermal convectionConvection cells — the catalog’s §II #17 / #22.
William W. Mullins & Robert F. Sekerkamorphological stability (1964)Why a flat interface throws out fingers and a leading tip outgrows its neighbors — the catalog’s H3 mechanism.
Philip Saffman & G.I. Taylorviscous fingering (1958)The catalog’s #52.
T.A. Witten & L.M. Sanderdiffusion-limited aggregation (1981)The canonical branching-growth model — and the reason the catalog puts crystal-growth dendrites on the cut list (branched shape, but nothing flows through a finished crystal).
Grégoire Nicolis & Ilya PrigogineSelf-Organization in Nonequilibrium Systems (1977)The textbook that put dissipative-structure pattern formation on formal footing.

VII. Branching networks, allometry & form#

Why branched conduits look the way they do — the geometry and scaling of flow networks.

ThinkerWorkHow it applies
D’Arcy Wentworth ThompsonOn Growth and Form (1917)The original argument that biological form is shaped by physical forces and flows, not heredity alone.
Cecil D. MurrayThe Physiological Principle of Minimum Work,” PNAS 12, 207 (1926)Murray’s law sets the vessel calibers at a vascular bifurcation by minimizing flow-plus-maintenance cost — the catalog’s H5 mechanism.
Benoît MandelbrotThe Fractal Geometry of Nature (1982)The language for the self-similar branching that recurs at every scale.
Ignacio Rodríguez-Iturbe & Andrea RinaldoFractal River Basins: Chance and Self-Organization (1997)Optimal channel networks: river basins self-organize toward minimum total energy expenditure, reproducing real drainage statistics.
Geoffrey West, James Brown & Brian EnquistWBE model, Science 276, 122 (1997); West, Scale (2017)Metabolic scaling laws fall out of optimized, space-filling, fractal branching distribution networks — vasculature, plant vasculature, rivers.

VIII. Origin of life & self-organizing chemistry#

The molecular scale: self-organizing chemistry and the origin of life.

ThinkerWorkHow it applies
Harold J. MorowitzEnergy Flow in Biology (1968)Energy flowing through a system forces matter to cycle and organizes it — a bridge from thermodynamics to the origin of life.
Manfred Eigen & Peter SchusterThe Hypercycle (1979)Self-replicating molecular networks that compete and lock in.
Stuart KauffmanThe Origins of Order (1993); At Home in the Universe (1995)Autocatalytic sets and “order for free”: self-organization, not selection alone, supplies biological order.
Addy ProssWhat Is Life? How Chemistry Becomes Biology (2012)Dynamic kinetic stability: replicators persist by being good at persisting through flow, not by sitting in an energy minimum.
Eric Smith & Harold MorowitzThe Origin and Nature of Life on Earth (2016)Life as a planetary-scale channel for relaxing chemical gradients.

IX. Energy flow across scales#

Energy flow treated as the master variable, from the cosmos to the economy.

ThinkerWorkHow it applies
Eric ChaissonCosmic Evolution (2001); energy rate densityEnergy rate density — free-energy flow per unit mass — proposed as a master metric of complexity, rising from stars to plants to brains to society.
Tim Garrettthermodynamic model of the economyModels the global economy as a physical system whose growth is tied to its rate of energy dissipation.
François RoddierThe Thermodynamics of Evolution (francois-roddier.fr)Economies and ecosystems as dissipative structures cycling between order and chaos.

X. Intelligence, agency & the brain#

Intelligence and agency framed in thermodynamic terms.

ThinkerWorkHow it applies
Alex Wissner-GrossCausal Entropic Forces,” Phys. Rev. Lett. 110, 168702 (2013) (alexwg.org)Intelligence framed as a force that acts to maximize future freedom of action.
Daniel Polaniempowerment / intrinsic motivationAgents act to maximize their control over future states — an information-theoretic relative of the causal-entropic-force idea.
Karl FristonThe free-energy principle: a unified brain theory?,” Nat. Rev. Neurosci. 11, 127 (2010)Organisms persist by acting to minimize a free-energy bound on surprise — keeping themselves in their expected states.
Alfred Hüblerself-assembling conductive structures; lectureShowed conductive beads in oil self-wiring into current-carrying trees under an applied field. The clearest experimental demonstration of physical intelligence in the DARPA Physical Intelligence program.

XI. Physical & thermodynamic computing#

People building computers from physics and noise rather than abstracting them away.

ThinkerWorkHow it applies
John J. HopfieldHopfield networks (1982); Nobel in Physics 2024Computation as relaxation down an energy landscape — settling into a minimum.
Geoffrey HintonBoltzmann machines (with Sejnowski, 1985); Nobel in Physics 2024Learning as thermal sampling over an energy function — the statistical-physics wing of deep learning.
Carver MeadAnalog VLSI and Neural Systems (1989); “Neuromorphic Electronic Systems” (1990)Neuromorphic engineering: compute with the device physics instead of abstracting it away.
Alex NugentAHaH Computing, PLOS ONE (2014); the kT-bitAnti-Hebbian and Hebbian (AHaH) plasticity reduced to a differential pair of memristors: two conduction pathways competing for a flow.
Massimiliano Di Ventra & Yuriy V. PershinMemComputing (2022)Computing with memory-bearing dynamical elements.
Todd Hyltonthermodynamic computingFormer DARPA program manager; long-time advocate of thermodynamically-grounded computing and AI, out of the DARPA Physical Intelligence lineage.
Robert Fryphysical intelligence (preprint)Frames physical intelligence and thermodynamic computing from an information-theoretic angle.
Patrick J. Coles (Normal Computing) & Guillaume Verdon (Extropic)thermodynamic-computing hardware (Coles 2023; Extropic 2024)Part of current commercial wave — probabilistic / energy-based machines that exploit noise rather than suppress it.