Dissipative Adaptation

Speculative If far-from-equilibrium driving really does select for structures that absorb work from the drive — which no physical experiment has shown, and section 7 says what would — then the search for the origin of life stops being a combinatorial problem and becomes a thermodynamic one. The condition comes first because it is the whole of the disagreement: the theorem that carries the name is not in dispute, and the claim that it explains the emergence of life-like organisation rests on simulations of 20 particles and of 25 chemical species.

Established The magnitude is best stated as a distance from a thermodynamic floor, because that is the one quantity this programme has actually computed against measurements. England’s 2013 bound requires that the heat a self-replicator dumps per replication, plus its internal entropy change, be at least the logarithm of the ratio of its birth rate to its decay rate. For Escherichia coli dividing in 20 minutes he put the measured heat at βQ = 220 per peptide bond over about 1.6 × 10⁹ peptide bonds, against a bound of 42 per peptide bond: the organism runs a factor of about 5.2 above its own thermodynamic floor. For a self-replicating RNA doubling in an hour the bound gives 7 kcal mol⁻¹ against a measured reaction enthalpy near 10 kcal mol⁻¹.

Speculative The engineering prize, if the selection claim holds, is a design rule stated in the spectrum of the drive rather than in the chemistry of the parts. Today a dissipative self-assembly experiment is designed by choosing a fuel and a catalyst and seeing what forms and how long it lasts. A working theory of dissipative adaptation would instead say which drive waveform produces which structure, which is a specification with a unit — a frequency — rather than a synthesis campaign.

Frontier Who collects it is narrow and mostly academic. The near-term beneficiaries are origin-of-life theory, soft-matter design and the small community extending these relations to quantum systems. There is no product, no programme with a budget line, and no patient at the end of it, and this page does not pretend otherwise.

Handwave The condition, stated plainly: the upside requires that work absorption be the quantity that gets selected, and the theory’s own extensions have already produced counterexamples. A 2025 quantum treatment reports a coupling regime in which optimal self-organisation is achieved without maximal work consumption, and a driven spin-glass simulation from England’s own group finds work absorption declining over time rather than rising. Section 5 is about why the sign of the effect is still model-dependent.

Frontier Two claims travel under one name and they are not the same size. The first is a theorem about driven stochastic systems, derived from microscopic reversibility, with published scope conditions. The second is an interpretive proposal that this theorem accounts for the spontaneous appearance of life-like organisation in matter. The first is mathematics and is essentially uncontested; the second is a hypothesis about the world and has no experimental support. Almost every public description of “a physics theory of life” slides between them.

Established The substrate is the fluctuation-theorem literature of the 1990s, and it is solid. Jarzynski’s 1997 equality relates the exponential average of the work done in a non-equilibrium process to an equilibrium free-energy difference; Crooks’s 1999 theorem states that the ratio of forward to reverse work distributions is an exponential in the dissipated work. Both follow from time-reversal symmetry of the underlying dynamics plus a heat bath at fixed temperature, and both were confirmed experimentally in single molecules.

Established England’s 2013 result is a rearrangement of that machinery into a statement about macrostates. For a transition between two coarse-grained states, the heat expelled, plus the log-ratio of reverse to forward transition probabilities, plus the change in internal Shannon entropy, is at least zero. Applied to a replicator with birth rate g and decay rate δ, it gives βq + s_int at least ln(g/δ): a thing that makes copies of itself quickly and falls apart slowly must dump a minimum amount of heat.

Speculative Dissipative adaptation is the reading of that inequality in the other direction: that among the outcomes available to a driven system, those with a larger history of work absorption are more likely. England’s 2015 Nature Nanotechnology perspective states it as a general thermodynamic mechanism for self-organisation via dissipation of absorbed work. Handwave The step from “fast replicators must dissipate” to “dissipation selects for replicator-like organisation” is the step this brief exists to examine, and it is not a theorem.

Frontier Four things in this neighbourhood are routinely confused and are genuinely distinct. Prigogine’s dissipative structures are ordered patterns sustained by a through-flow of energy — convection cells, chemical oscillators — and are an observed phenomenon with no selection claim attached. Dissipative self-assembly, in the chemistry literature, is transient assembly fuelled by a consumed chemical and is an experimental programme with real materials. Dissipative adaptation is England’s selection claim. The maximum entropy production principle is a fourth and separate conjecture that steady states maximise entropy production, which dissipative adaptation does not assert and does not require.

Handwave A working definition, because the field does not supply a shared one: dissipative adaptation is confirmed when, in a physical system, the structures that emerge under a drive are shown to be those that absorbed more work than the alternatives, with the comparison made against a control driven at the same average power and a different spectrum. Speculative By that definition the count of confirmations is zero, and stating it that way is what makes section 7 well posed rather than rhetorical.

Frontier The reason the question is hard is that the quantity in the theory is a path functional and the observable is a snapshot. Absorbed work is accumulated over a trajectory; what an experiment sees is a structure at the end. Connecting the two requires per-trajectory work accounting in a system with many degrees of freedom, which is the measurement problem section 5 is about.

Established Begin with the part of this subject that is measured, because it is not the part the name refers to. The commission for this brief supplies as its anchor the verification of the Crooks and Jarzynski fluctuation theorems in single-molecule RNA unfolding, and that anchor checks out. Collin, Ritort, Jarzynski, Smith, Tinoco and Bustamante, in Nature 437, 231 (2005), pulled an siRNA hairpin targeting HIV CD4 receptor mRNA and the S15 three-helix junction from the E. coli 16S ribosomal RNA with optical tweezers, and recovered folding free energies from the crossing point of forward and reverse work distributions.

Established The numbers are tight enough to be worth quoting. For the hairpin at a 7.5 pN s⁻¹ loading rate the crossing point gave ΔG = 110.3 ± 0.5 k_BT, and 62.8 ± 1.5 k_BT at zero force. For the S15 junction without magnesium, ΔG = 154.1 ± 0.4 k_BT at a mean dissipated work of 36.3 k_BT; a single base-pair mutant differed by ΔΔG = 3.8 ± 0.6 k_BT; 4 mM magnesium stabilised the wild type by 31.7 ± 2 k_BT. The authors also state the method’s limit: it works for soft traps and is probably limited to molecules dissipating less than about 100 k_BT.

Established A finding about the commission, since it changes what the anchor licenses: the anchor is correct and it is anchoring the wrong claim. What was verified is a time-reversal identity in a two-macrostate molecule pulled by a tweezer. Nothing in it tests selection, self-organisation or adaptation; the RNA does not choose to unfold, and no structure emerges that was not there before. Frontier The scale gap compounds the mismatch: the verified regime tops out near 100 k_BT of dissipation, while the smallest system dissipative adaptation is actually about — a dividing bacterium — dumps βQ of order 220 × 1.6 × 10⁹, which is roughly nine orders of magnitude beyond where the theorems have been checked. Handwave The fluctuation theorems are believed to hold there because they are identities rather than approximations, which is a good reason and not a measurement.

Established The 2013 bound itself is exact, narrow, and states its own conditions. England derives βQ_I→II + ln[P(II→I)/P(I→II)] + Δs_int at least zero for a transition between macrostates, requiring a heat bath at constant temperature, diffusive motions that lack any sense of ballistic inertia, and driving that is time-symmetric or absent. Frontier He is explicit that the relation is valid for any transition between two microstates and therefore applies to undriven relaxation arbitrarily far from equilibrium — which is a statement of generality and simultaneously a warning that the inequality by itself picks out nothing special about life.

Established The bacterial calculation is the most-quoted number in this subject and it is a consistency check, not a prediction. With about 1.6 × 10⁹ peptide bonds per cell, a 20-minute division time and an internal entropy contribution bounded above at 10 per peptide bond, England puts the measured βQ at 220 per peptide bond against a minimum of 42: a factor of about 5.2. Frontier A factor of five above a floor is a real result about how close biology runs to a thermodynamic limit; it is not evidence that the limit caused the biology. Speculative The self-replicating RNA example is tighter — a bound of 7 kcal mol⁻¹ against a measured enthalpy near 10 — and correspondingly less informative, because a tight bound and a loose one are equally compatible with the bound having played no causal role.

Established England says as much himself, in the paper. Any self-replicator faces constraints besides the thermodynamic ones, and he expects real organisms to be limited far more restrictively by the speeds of rate-limiting reactions. Frontier That sentence is the difference between a necessary condition and an explanation, and it is routinely dropped in secondary accounts of the work.

Established The 2016 Physical Review X paper is where the interpretive claim is made formally, and it is where the scope conditions get tightest. Perunov, Marsland and England construct a generalised Helmholtz free energy for finite-time stochastic evolution whose terms separate the contributions to the relative probability of different far-from-equilibrium outcomes, with the mean dissipated heat entering as one term and a heat-fluctuation correction as another. Established The assumptions are listed: time-reversal symmetry of the microscopic dynamics, detailed balance and a Boltzmann distribution at equilibrium, small system-bath coupling, weak external driving, and Arrhenius transition rates.

Established Two sentences in that paper do more to locate the claim than anything written about it since. The authors state that they have made no reference to self-replication or natural selection, and they write that it remains quite a mystery how a driven system learns to absorb energy better, if not through successive rounds of reproduction and natural selection. Handwave The mechanism they describe requires many kinetically trapped degrees of freedom and, in their own words, remains speculative for macroscopic systems. A brief that reported dissipative adaptation as a settled physical principle would be contradicting its originators.

Established The supporting evidence is simulation, and it is worth being precise about how small. Kachman, Owen and England’s 2017 Physical Review Letters paper simulates 20 particles forming and breaking harmonic spring bonds under strong sinusoidal forcing of a single particle, in a regime of underdamped viscous drag and low temperature. Established The result is real and specific: the steady-state normal-mode spectrum develops an additional peak at the drive frequency, and the dissipation rate rises monotonically from its initial equilibrium value. Speculative Twenty particles have of order 10³⁹ bonding states available, so the search is genuinely hard; the system nonetheless finds resonance.

Handwave The same paper contains the result that should temper every citation of it. The authors describe their system as toy chemistry with catch bonds that do not reflect nanoscale physics, note that the character of the signature may differ with the detailed physical rules, and demonstrate the point by exhibiting the opposite behaviour under alternate bond mechanics. Speculative A mechanism whose sign flips when the bond model changes is not yet a law of nature; it is a phenomenon of a class of models.

Frontier The companion 2017 result, in PNAS, simulated 25 chemical species in a randomly connected reaction network under forcing. The publisher refused retrieval this session, so the figures here are taken from contemporaneous reporting and flagged accordingly: randomised starting points reached rare states of vigorous chemical activity and extreme forcing about four times more often than chance would predict, landing in the 99th percentile for forcing experienced. Speculative Four times chance in a 25-species toy network is a genuine statistical signal and a long way from an origin of life.

Frontier The most awkward simulation result comes from England’s own group and points the other way. Gold and England’s driven spin-glass study — binary spins with quenched random couplings on random networks of mean degree 4, fields switching every 100 s at inverse temperature 3 — reports that the work absorption rate declines over time and plateaus below its initial value. Established When the drive fields are replaced by a new random set after 300 switches, the spin-flip rate returns to its original value, and switching back to the first drive produces a smaller increase, which is a memory. Speculative The system adapts to its drive by becoming less responsive to it, which the authors frame as novelty detection. Adaptation to a drive is therefore not the same thing as maximising work absorption, even inside this research programme.

Established The quantum extensions are analytic, exactly solvable, and have produced the clearest internal counterexample. Valente’s 2021 Scientific Reports model of a chain of three-level systems driven by single-photon pulses derives a self-replication probability proportional to the absorbed work, which is dissipative adaptation in its cleanest possible form. Established The 2025 follow-up by Ganascini, da Silva and Valente generalises it to a dissipative two-site system and finds that at intermediate coupling, with the inter-site coupling at half the decay rate, optimal self-organisation is achieved without maximal work consumption, with the total absorbed work splitting into a self-organisation term and a coherence term. Speculative A framework whose own exact solutions contain a regime where the proposed selection principle fails is in a healthier state than one with no counterexamples, and is also further from being a principle.

Speculative A March 2026 treatment puts dissipative adaptation into a driven spin-boson model within the path-integral formalism, relating the work supplied by a dynamic potential to the transition probability between the two ground states of a double well. Frontier It is theory with no experiment attached, which is the pattern: every extension of this framework since 2013 has been analytic or numerical.

Frontier A separate 2021 analysis reaches a different selection criterion from a similar starting point. Ueltzhöffer, Da Costa, Cialfi and Friston, simulating stochastic chemical reaction networks, argue that as the driving chemical potential difference falls, selection pressure falls on efficiency rather than on gross dissipation: only structures efficient enough at harvesting work sustain themselves when the gradient is small. Speculative Maximum dissipation and maximum efficiency are different objectives that coincide only when the drive is abundant, and no experiment distinguishes them.

Established Now the question the commission asked to be answered plainly: what has been observed in a physical experiment, as opposed to a simulation? Three things, none of which is a test of dissipative adaptation. First, the fluctuation theorems, in the single-molecule regime described above. Second, chemically fuelled dissipative self-assembly — transient structures that form while a fuel is consumed and dissolve when it runs out — which is a substantial experimental literature demonstrating that driven matter organises, and which measures lifetimes and yields rather than absorbed work per trajectory. Frontier Third, directed aging: Hexner, Pashine, Liu and Nagel cut disordered networks into quasi-two-dimensional EVA foam sheets, aged them under compression at 50 °C for an hour, and found the aged material had acquired a non-monotonic Poisson’s ratio that reverses sign near the original unaged size — a real material that adapted to a real drive and kept a memory of it.

Speculative Directed aging is the closest thing in the laboratory to what dissipative adaptation describes, and it works by a different mechanism. The material changes because stressed bonds weaken and rest lengths shift, not because trajectories that absorbed more work became more probable. Handwave The two accounts have never been compared on the same system, and the training literature that grew out of directed aging — cyclic tuning of disordered sphere packings reported in PRX Life in 2026, and rate-dependent training of overdamped viscoelastic networks in a 2026 preprint — is entirely simulation.

Frontier The named criticism is on the record and is sharper than the literature usually admits. In 2017 reporting on the two simulation papers, Eugene Shakhnovich called the work potentially interesting exercises in non-equilibrium statistical mechanics of simple abstract systems while dismissing the biological claims as pure and shameless speculations; Sara Walker argued the theory is necessary but not sufficient because it cannot account for the information-processing capacity that separates organisms from Jupiter’s Great Red Spot; Jeremy Gunawardena held that life requires reproducing structures and Darwinian dynamics beyond thermodynamic properties. Speculative Michael Lässig’s supportive reading in the same piece is narrower than either camp’s summary: that the simulations reach rare fixed points without enormous waiting times, which is a statement about search speed rather than about life.

Frontier The ledger, stated once: the theorem is established and narrow; the bound is exact and is a necessary condition its author says is not the binding one; the selection claim is supported by three small simulations, contradicted in sign by a fourth from the same group and by an exactly solved quantum model; and no physical experiment has been designed to test it. Speculative That is not a refutation. It is a description of a hypothesis that is thirteen years old and has not yet been given the experiment that would embarrass it.

Frontier Question one: is dissipative adaptation a prediction or a description? The generalised free energy of the 2016 paper is an identity; identities do not forbid outcomes. Speculative To predict, the framework must say which of two available structures a given drive will produce, and the only published attempts to do that are the resonance result in a 20-particle model and the reaction-network result in a 25-species model. Handwave Whether the framework can make a prediction that a competent sceptic would not have made anyway is, thirteen years in, still unsettled.

Frontier Question two: what is the sign of the effect? The resonance simulations show dissipation rising; the spin-glass simulation from the same group shows it falling and plateauing; the exactly solved quantum two-site model shows optimal organisation without maximal absorption. Speculative These are reconcilable — different drives, different landscapes, different observables — but the reconciliation has not been written down, and until it is, “systems adapt to absorb more work” is a slogan the field’s own results do not uniformly support.

Frontier Question three: how much of the scope condition list survives contact with a real system? The derivations assume weak driving, small system-bath coupling, Arrhenius rates and time-symmetric forcing. Speculative A cell is strongly driven, strongly coupled, and its rates are enzymatic rather than Arrhenius; a chemically fuelled assembly runs on an irreversible fuel that is not time-symmetric at all. Handwave Nobody has published a perturbative estimate of how far the conclusions bend when each assumption is relaxed, which is the ordinary way a physical theory establishes its own domain of validity.

Speculative Question four: maximum dissipation or maximum efficiency? The 2021 chemical-network analysis argues that a falling driving gradient selects for efficient harvesting rather than gross dissipation. Frontier The two criteria make opposite predictions when the drive is scarce, and scarcity is the regime any origin-of-life scenario actually occupies, so this is not a technicality.

Speculative Question five: is work absorption even the right currency? Sara Walker’s objection is that information processing, not dissipation, is what distinguishes an organism from a hurricane, and that objection has a formal descendant in the assembly-index literature covered in Assembly Theory. Handwave No paper has tried to compute both a dissipation history and an assembly index for the same object, which would be the obvious first move toward deciding whether they measure the same thing.

Frontier Question six: what is the relationship to physical learning? The directed-aging experiments produce driven matter that stores a memory of its drive and changes its mechanical response, which is adaptation in a plain sense and is measured. Speculative If those results can be derived from the dissipative-adaptation framework, the framework acquires its first experimental foothold; if they cannot, the framework is a description of a different phenomenon than the one the laboratory is producing.

Speculative Question seven: does the quantum programme generalise or specialise? The single-photon models are exactly solvable, which is rare and valuable, and their results have already bent the general claim. Handwave Whether an exactly solvable two-site quantum system tells you anything about a prebiotic pond is a question the papers do not address and their readers should.

Handwave Question eight, the uncomfortable one: what observation would the proponents accept as a refutation? Speculative Nothing in print names one. A framework built from an identity, extended by models whose sign depends on bond mechanics, and defended as necessary-but-not-sufficient is very hard to lose an argument with, and the fastest route to making it a science is for its own community to write down the result that would count against it.

Frontier Bottleneck one, and the binding one: per-trajectory work accounting in a many-body system. The theory’s central quantity is the work absorbed along a path; the experiment’s observable is the structure at the end. Established The single-molecule technique that does measure work per trajectory is optical tweezers, and its own practitioners put its ceiling near 100 k_BT of dissipation. Speculative A self-organising system with enough degrees of freedom to have an interesting outcome distribution dissipates far more than that, so the one instrument with the right observable is aimed at the wrong systems.

Speculative Bottleneck two: there is no way to hold average power constant while changing the drive spectrum in a chemically fuelled system. The control the decisive experiment needs — same power, different spectrum — is straightforward with an optical or mechanical drive and close to impossible when the drive is a consumed fuel, because concentration, rate and spectrum are the same knob. Frontier That is why the experiment in section 7 is specified for driven colloids or driven mechanical networks rather than for the chemistry where the phenomenology is richest.

Speculative Bottleneck three: the outcome space has to be enumerable. The claim is comparative — the structures that emerged absorbed more than the ones that did not — so an experiment must be able to say what the alternatives were. Handwave In a 20-particle simulation the 10³⁹ bonding states are at least countable in principle; in a beaker they are not, and no published protocol defines the comparison class for a physical system.

Frontier Bottleneck four: thermometry and heat accounting at the relevant scale. Establishing that a structure absorbed more work means measuring heat flows into a bath from a microscopic subsystem, over long times, without perturbing the dynamics. Speculative Calorimetry at those scales is improving but no group has demonstrated it resolved by outcome class, which is what the claim requires.

Speculative Bottleneck five: the theory is stated in variables experiments do not report. Papers on dissipative self-assembly report yields, lifetimes and fuel consumption; the theory wants a mean dissipated heat and a fluctuation correction resolved by macrostate. Handwave Bridging that gap is not an instrument problem but a convention problem, and conventions change only when a journal or a funder asks for the numbers.

Speculative Bottleneck six: model dependence has not been bounded. The resonance result changes sign under alternate bond mechanics, by the authors’ own demonstration. Frontier Until someone publishes the class of microscopic rules under which the effect holds and the class under which it reverses, simulation evidence cannot accumulate: each new model is a new anecdote.

Handwave Bottleneck seven, methodological: nothing prevents a post-hoc dissipation story. Given any outcome, a history of work absorption can be computed and reported as its explanation. Speculative The guard against that is prediction registered before the run, which is why the decisive experiment in section 7 has registration built into it rather than appended as good practice.

Frontier Dependency one: a driven physical platform with trajectory-resolved energetics and a countable outcome space. Driven colloidal systems with holographic tracking, or laser-cut mechanical networks under programmable forcing, are the two candidate classes; both can be driven at controlled spectrum and both allow the work input to be computed from measured forces and displacements. Speculative Neither has been used for this purpose, and the reason appears to be that nobody in either community reads this literature as making an experimental demand.

Speculative Dependency two: a perturbative theory of the scope conditions. What the field needs is not a stronger theorem but a weaker one: bounds on how the conclusions degrade under strong driving, strong coupling, non-Arrhenius rates and time-asymmetric forcing. Frontier This is ordinary statistical mechanics and requires no new apparatus; it is the piece of work that would tell an experimentalist which systems are in scope.

Speculative Dependency three: a shared benchmark of driven models. The spring-bond system, the chemical reaction network, the spin glass and the two-site quantum model give four different answers because they are four different models, not because they disagree about physics. Handwave Running all four under a common drive protocol and reporting the same three quantities would settle the sign question in a year, and no group is positioned to want to do it.

Frontier Dependency four: results this page does not own. The thermodynamic accounting for information-bearing processes is treated in Thermodynamics of Computation, which reports that Landauer’s bound is confirmed for one bit asymptotically and that no experiment has measured a working circuit against the corrected total — the same shape of gap this brief documents, in a subject with better-controlled apparatus. Speculative If dissipative adaptation is to be quantified in any system that processes information, the measurement conventions will come from there.

Speculative Dependency five: the chemistry that would supply a real test case. Prebiotic Chemical Networks reports the best RNA replicator at 94.1% fidelity and 0.2% yield over 72 days and no end-to-end prebiotic carbon yield anywhere. Handwave A dissipative-adaptation account of the origin of life has to operate on chemistry at those yields, and no paper in this framework has ever been written against a measured prebiotic network.

Handwave Dependency six, stated as an absence: nothing here waits on a theoretical result that does not exist. The mathematics has been available since 1999 and the specific bound since 2013. Speculative This is an experimental and conventional bottleneck wearing the costume of a theoretical one, which is the same diagnosis reached independently on the sibling page Morphogenetic Pattern Formation, and the coincidence is worth noticing: both subjects have more theory than anyone can test.

Frontier The tests below are ranked by what a result would settle, and the first is the one this brief treats as decisive.

Speculative Tier one, the decisive experiment: a driven physical system in which the structures that emerge are shown to have absorbed more work than the alternatives, against a control driven at the same average power with a different spectrum, with the prediction registered before the run. Frontier The same-average-power control is the entire design. Without it, any observation that driving produces structure is confirmed by every dissipative system ever built, from a convection cell to a candle flame, and distinguishes nothing. Handwave With it, the claim becomes falsifiable in one run: if outcome frequencies are the same under both spectra, the drive’s work history is not what is selecting.

Speculative The cheapest version that would still be decisive uses driven colloids. A few hundred particles with tunable interactions in an optical or acoustic drive, tracked to give per-trajectory work from measured forces and displacements, with the outcome space coarse-grained into a countable set of cluster morphologies. Frontier Every element of that apparatus exists in soft-matter laboratories today; none of them has been pointed at this question. Established The mechanical analogue is equally available: the directed-aging protocol already drives disordered networks and measures the change, and adding work accounting and a spectral control to it is an increment rather than a programme.

Frontier Tier two: resolve the sign. Run the spring-bond resonance model, the driven spin glass, the chemical reaction network and the quantum two-site system under a common drive protocol and report absorbed work, outcome distribution and response over time in the same units. Speculative This is a month of computation and would establish whether rising dissipation, falling dissipation and organisation-without-maximal-absorption are three regimes of one mechanism or three unrelated model artefacts. Handwave It requires nothing but agreement, which is why its absence is informative about the field rather than about the physics.

Speculative Tier three: extend fluctuation-theorem verification past 100 k_BT. The identities are believed to hold at every scale; they have been checked in a regime nine orders of magnitude below a dividing bacterium. Frontier A verification in a system dissipating 10⁴ to 10⁶ k_BT — a driven colloidal cluster, a small active-matter assembly — would either extend the foundation or find the first crack in it. Established The apparatus is the same optical tweezers, used differently, and the authors of the 2005 verification named this limit themselves.

Speculative Tier four: the scope-condition stress test. Take one exactly solvable model and violate each assumption in turn — strong driving, strong coupling, non-Arrhenius rates, time-asymmetric forcing — measuring how far the outcome-probability relation bends. Handwave A theory that cannot say where it stops applying cannot be said to apply anywhere in particular, and this is the piece of work that would let an experimentalist choose a system without guessing.

Speculative Tier five: a head-to-head between dissipation history and assembly index. Compute both for the same set of objects produced under drive, and see whether they rank them the same way. Frontier If they agree, two pre-paradigmatic programmes are measuring one quantity and should merge; if they disagree, the disagreement identifies what dissipation does not capture, which is precisely the objection raised against this framework in 2017.

Speculative Tier six: derive directed aging from the framework, or fail to. The EVA-foam experiments are the only laboratory system in which driven matter demonstrably adapts to its drive and keeps a memory. Handwave Either the dissipative-adaptation formalism reproduces the measured non-monotonic Poisson’s ratio, which would be its first contact with data, or it does not, which would tell the field that the adaptation happening in laboratories is a different phenomenon from the one on the blackboard.

Handwave What would change this brief’s assessment most, in one sentence: a registered, spectrally controlled, work-resolved demonstration that emergent structures are the ones that absorbed the most. Speculative Absent that, the framework can keep producing consistent accounts of outcomes after the fact, and a consistent account is not a selection principle.

Frontier There is no device to specify here, so the engineering requirements are for the measurement, and they are unusually concrete for a Category XI subject.

Speculative Requirement one: a drive whose spectrum can be reshaped at constant average power. Acoustic, optical and electromagnetic drives all permit it; a chemical fuel does not. Frontier The specification is a waveform generator plus calibrated power delivery, which is standard equipment, and the fact that no published experiment in this subject reports it is a statement about the question having never been asked.

Speculative Requirement two: trajectory-resolved work accounting for hundreds of degrees of freedom. Holographic or confocal tracking at kilohertz rates with known interaction potentials gives the work input per particle per frame. Handwave The error budget has never been worked out for this purpose, and whether the accumulated work over a long trajectory is resolvable against tracking noise is the open engineering question, not the drive.

Frontier Requirement three: an outcome classifier defined before the run. The claim is comparative, so the comparison classes must be fixed in advance, by cluster morphology, by bond graph, or by an order parameter chosen and registered before data collection. Speculative This is a protocol requirement rather than a hardware one and it is the cheapest item on the list.

Speculative Requirement four: calorimetry resolved by outcome class. Independent confirmation that the structures that formed are the ones that dumped more heat, rather than inferring the heat from a model of the dynamics. Handwave This may be the item that does not arrive: microscale calorimetry at the required resolution and with outcome discrimination has not been demonstrated, and if it cannot be, the decisive experiment rests on inferred rather than measured heat.

Speculative Where this pays off first is materials rather than biology. A rule connecting drive spectrum to emergent structure is directly usable in the training-by-driving programme that produced the directed-aging results and the sphere-packing and viscoelastic-network training simulations of 2026. Frontier That community already treats driving as a design variable; it simply does not use this framework to choose the waveform.

Frontier The commission names Synthetic Ecosystems (FR-III-27) as this brief’s nearest existing neighbour, and on inspection the placement is weak. That page is about closed-loop life support: BIOS-3 regenerating air and water completely while food regeneration reached 93%, Biosphere 2’s oxygen falling from 20.9% to about 14% in sixteen months because uncured concrete absorbed the carbon dioxide, thirty-seven years of MELiSSA with a rat for a crew. Speculative The shared vocabulary is that both subjects concern systems maintained away from equilibrium by a through-flow of energy. The shared content is close to nil: a closed ecosystem is an engineering artefact whose failures are chemical and institutional, and this page would read identically if no closed ecosystem had ever been built.

Established The three genuinely nearest pages are all in this category rather than in Category III. Thermodynamics of Computation shares the apparatus, the fluctuation-theorem substrate and the exact failure mode: a bound that is confirmed in a minimal system and never measured in a working one. Prebiotic Chemical Networks owns the empirical target that dissipative adaptation claims to explain. Assembly Theory is the rival attempt to put a number on the same intuition using a combinatorial rather than a thermodynamic quantity.

Frontier The relationship with Assembly Theory is the most productive one and the least explored. Both programmes propose a scalar that distinguishes the products of life from abiotic matter; one counts construction steps and the other accumulates dissipated work. Speculative That page records a threshold at an assembly index of about 15 contradicted by abiotic minerals computed at 18 and never re-run blind outside the originating group, which is a cautionary precedent for any scalar-of-life proposal, this one included.

Speculative Homochirality Origins supplies the sharpest available test case that nobody has used. A racemate driven toward a single handedness is a symmetry-breaking event with two enumerable outcomes and a measurable excess, which is exactly the countable outcome space section 5 says physical systems usually lack. Handwave Whether a dissipation history distinguishes the two hands is a question nobody in either field has asked, and the measured abiotic ceiling of about 2% enantiomeric excess from circularly polarised light is a drive with a spectrum already under experimental control.

Frontier Morphogenetic Pattern Formation, written alongside this page, is the structural sibling. Both subjects have a mathematics that fits their phenomena easily, a handful of small simulations doing the argumentative work, and a decisive experiment that requires a registered quantitative prediction with a rival model scored on the same data. Speculative The two briefs arrived at that remedy independently, which is either a sign that it is the right remedy for this class of pre-paradigmatic subject or a sign of a house style; the reader should weigh both.

Speculative Extended Evolutionary Synthesis is adjacent through the question of what selection is. Dissipative adaptation is sometimes presented as selection without replication, which would be a substantial addition to evolutionary theory if it were demonstrated. Handwave That page’s verdict on comparable proposals — mechanism present, prospective prediction absent — transfers here without modification.

Frontier This subject has no institution, no facility and no budget line, and the people best equipped to test it are not in it. The theory lives in a small number of statistical-physics groups; the apparatus that could test it lives in soft-matter and single-molecule laboratories funded for other purposes. Speculative No funding call has ever asked for the experiment in section 7, and the experiment is small enough that a single mid-sized grant would cover it.

Speculative The publication dynamics have been unusually distorting here, and the record is public. Two simulation papers in 2017 were reported as first support for a physics theory of life, in coverage that also carried a named physicist calling the biological interpretation pure and shameless speculations. Handwave Both halves of that sentence were in the same article, and only one half propagated; the phrase that travelled was the headline claim, not the objection.

Established The originators have been more careful in their papers than their field has been in citing them. The 2016 paper states that it makes no reference to self-replication or natural selection; the 2013 paper says the thermodynamic constraint is probably not the binding one for a real organism; the 2017 resonance paper demonstrates its own effect reversing under alternate bond mechanics. Speculative Every serious caveat in this subject was published by its proponents, which is a credit to them and a problem for a literature that compresses away caveats at each citation step.

Speculative Preregistration is the cheapest institutional fix available and it fits this subject unusually well. The decisive experiment’s whole content is a prediction posted before a run, and the physics community already registers analyses in particle physics and gravitational-wave astronomy. Handwave Soft matter has no such convention, and there is no visible constituency for creating one over a question with no clinical or industrial stake.

Speculative What an institution could usefully own: the common benchmark of driven models. Four published models give three different answers about the sign of the effect, and nobody has run them under a shared protocol. Frontier The cost is a postdoctoral year and a code repository; the obstacle is that the benefit accrues to the field rather than to any group, which is the standard signature of work that only happens when someone decides to pay for it.

Handwave One structural note with no fix attached: the most active current extension of this framework is a quantum-optics programme in Brazil, and the most relevant experimental programme is a soft-matter one in the United States and Israel, and they do not cite each other. Speculative Whether those literatures merge is a question about conferences and referees rather than about physics.

Frontier The direct ethical stakes are close to zero and should be reported as such rather than inflated to fill the section. The experiments are simulations, optical tweezers, colloids and foam sheets. No living subject is involved, no dual-use pathway is visible, and no regulatory body has an interest.

Speculative The indirect stake is a claim-inflation problem with an unusually wide audience. “Physics shows that life is thermodynamically inevitable” is a sentence that circulates well beyond the field, is used in arguments about the prevalence of life in the universe, and is not supported by anything in the literature this brief reviewed. Handwave The harm is epistemic and diffuse: a public that believes a hard question has been answered is less able to evaluate the next claim from the same direction.

Speculative There is a specific downstream use worth naming: astrobiology and biosignature reasoning. If dissipative adaptation were established, it would raise the prior on life arising wherever a suitable drive exists, which changes how much evidence a candidate biosignature needs to carry. Frontier Using an unestablished thermodynamic argument to lower an evidential bar is the one concrete way this subject could do damage outside itself.

Speculative An epistemic obligation rather than an ethical one, and the same one the sibling brief arrives at: a framework whose proponents have published its limitations owes its popularisers the same limitations. Handwave Teaching dissipative adaptation as the thermodynamic explanation of life’s origin, without teaching that its supporting evidence is a 20-particle spring model and a 25-species reaction network, transmits a confidence the papers do not contain.

Speculative If the selection claim were established, the consequence that matters is not technological but probabilistic: it would change the expected frequency of life in the universe by making organisation a generic response to driving rather than a rare accident. Handwave No number can responsibly be attached to that change, because the theory as it stands does not predict a rate, a timescale or a probability for any physical system — and a framework that cannot say how long a prebiotic pond takes cannot be used to estimate how many ponds succeeded.

Speculative The concrete civilizational payoff, if there is one, is a design principle for matter that trains itself. The directed-aging and physical-learning results show real materials acquiring properties from their loading history; a theory saying which drive produces which property would turn that from an empirical craft into a specification. Frontier That is a materials-engineering outcome on a decades horizon, not a theory-of-life outcome, and it is the part of the upside this brief regards as most likely to survive.

Handwave There is a durable intellectual consequence independent of the physics: this subject is a clean case study in how a correct theorem acquires an incorrect reputation. Speculative Every step from the 2013 inequality to the popular claim was made by someone reasonable, in public, with caveats attached; the caveats fell off at the joins. The corrective — state the theorem, state its scope conditions, state separately what has been observed — is general, and it is the organising principle of this page.

Handwave What this subject will not do, stated so no reader infers it: it will not produce a therapy, a fuel, a material or a capability on any horizon this brief can defend. Speculative The nearest deliverable is a rule for choosing a training waveform, and that arrives through mechanical metamaterials, not through biology.

These horizons track what would have to be measured, because no programme in this subject has a schedule and none has a budget:

• 10 yr: Frontier Expect the theoretical side to keep extending — quantum models, path-integral treatments, more network classes — because that work is cheap and publishable, and expect at least one more internal counterexample of the kind the 2025 two-site result already supplies. Speculative Expect the first serious attempt at a spectrally controlled driven-colloid test, most plausibly from a soft-matter group that encountered the claim rather than from the statistical-physics community that made it. Speculative Expect the physical-learning literature to keep producing measured adaptation in driven matter without using this framework to describe it.
• 25 yr: Speculative This is the window in which the decisive experiment plausibly runs and is replicated: same average power, different spectrum, work-resolved outcomes, registered first. Speculative A positive result would make dissipative adaptation the first selection principle in physics that does not require replication, and would move the origin-of-life argument in Prebiotic Chemical Networks from chemistry onto thermodynamics. Handwave A null would not refute the theorem, which is an identity, but would retire the interpretive claim, and that asymmetry is why the experiment has to be registered.
• 50 yr: Handwave On this horizon the plausible development is absorption rather than resolution: the useful parts become standard tools in non-equilibrium statistical mechanics and in the design of trainable materials, and the theory-of-life framing is quietly dropped. Speculative If no physical experiment has been run by then, the sober reading is that the claim was never operationalised, which is a different and more common fate than being wrong.
• 100 / 250+ yr: Handwave Beyond useful forecasting. The defensible statement is about shape rather than date: a proposal derived from an identity, whose supporting models disagree about the sign of the effect, either acquires an experiment or becomes a chapter in the history of attempts to derive biology from thermodynamics. Handwave Any claim here about engineering life from first principles is an assertion, and this page does not make one.

• Depends on No typed dependency edge on this map blocks this subject, and the absence is the finding. Nothing here waits on a result another brief produces: the fluctuation theorems date from 1997 and 1999, the bound from 2013, and what is missing is an experiment nobody has designed. Thermodynamics of Computation is the nearest page that supplies method rather than context, because the conventions for measuring dissipation in a minimal system are being worked out there first; Prebiotic Chemical Networks supplies the empirical target this framework claims to explain, at yields that constrain any such claim.
• Requires (not on this map) Five conditions, in the order of the tokens above. The comparison. Structures that emerged under one drive spectrum, scored against those emerging under another at the same average power, with absorbed work resolved by outcome class — the measurement the whole claim reduces to, and the one nobody has made. The scope theory. Bounds on how the conclusions degrade under strong driving, strong coupling, non-Arrhenius rates and time-asymmetric forcing, so an experimentalist can tell which systems are in scope. The platform. A few hundred driven colloids with tunable interactions, tracked fast enough that work input per particle is resolvable against tracking noise — an error budget nobody has published. The convention. A registered prediction before the run, because a dissipation history can be computed for any outcome after the fact. The benchmark. Money for someone to run the spring-bond, spin-glass, reaction-network and quantum two-site models under one protocol and settle whether the sign of the effect is physics or model choice.
• Enables A confirmed selection principle would give trainable materials a rule connecting drive spectrum to emergent structure, would give origin-of-life modelling a thermodynamic prior it currently asserts, and would supply astrobiology with the first physical argument for a non-trivial frequency of life. A null result would be nearly as useful, because it would retire an argument that currently circulates unopposed outside the field.
• Adjacent Synthetic Ecosystems is the commission’s named neighbour and overlaps least; Thermodynamics of Computation and Assembly Theory overlap most, the first on apparatus and the second on the rival scalar; Homochirality Origins supplies the countable-outcome test case; Morphogenetic Pattern Formation is the structural sibling, with the same gap between a fitting formalism and a demonstrated mechanism.

Frontier Claim: “England proved that life is thermodynamically inevitable.” Established What was proved is that a system making copies of itself at rate g and decaying at rate δ must dump heat of at least ln(g/δ) in units of k_BT, plus its internal entropy change. Speculative That is a constraint on replicators, not a mechanism that produces them, and its author wrote in the same paper that a real organism is probably limited far more restrictively by the speeds of its rate-limiting reactions.

Established Claim: “The theory has been confirmed experimentally — the fluctuation theorems were verified in RNA pulling.” This is the anchor this brief was commissioned with, and the verification is real while the inference is not. Collin and colleagues confirmed the Crooks theorem in 2005 and recovered folding free energies to within about 1 k_BT. Frontier An RNA hairpin being pulled between two known states is not a system that self-organises, and the verified regime stops near 100 k_BT of dissipation while a dividing bacterium is nine orders of magnitude beyond it. Handwave Confirming the floor a theory stands on is not confirming the theory.

Frontier Claim: “Dissipative adaptation is the maximum entropy production principle.” Speculative It is not, and the confusion runs in both directions. Maximum entropy production asserts that steady states maximise entropy production; dissipative adaptation makes a statement about the relative probability of outcomes given their work-absorption histories. Frontier A 2021 chemical-network analysis reaches a third position — selection on efficiency rather than on gross dissipation as the driving gradient falls — and the three are experimentally indistinguishable today.

Speculative Claim: “Dissipative self-assembly experiments confirm dissipative adaptation.” Established The chemistry literature on fuel-driven transient assembly is real, experimental and productive: structures form while fuel is consumed and dissolve when it is gone. Handwave Those experiments measure yields and lifetimes, not work absorbed per trajectory resolved by outcome, and none of them includes the same-average-power spectral control that would make the comparison a test rather than an illustration.

Frontier Claim: “It is just Crooks restated, so it says nothing new.” Established The 2016 construction does more than restate: it separates the contributions to the relative probability of far-from-equilibrium outcomes into terms with distinct physical meanings, including the mean dissipated heat and a fluctuation correction. Speculative The fair version of the objection is narrower and survives: an exact decomposition constrains without predicting, and a decomposition becomes a theory of self-organisation only when it says which outcome to expect.

Speculative Claim: “Simulations have shown that driven systems learn to absorb more energy.” Established One simulation shows exactly that, in 20 particles under sinusoidal forcing, with a normal-mode peak appearing at the drive frequency. Frontier Another, from the same group, shows work absorption falling and plateauing as a driven spin glass habituates to its drive. Handwave Both results are in the literature, they point in opposite directions, and the field has not reconciled them.

Frontier Claim: “This explains evolution without natural selection.” Established The authors of the 2016 paper explicitly state that they make no reference to self-replication or natural selection, and that it remains quite a mystery how a driven system learns to absorb energy better if not through successive rounds of reproduction and natural selection. Speculative The framework is at present a candidate account of how a substrate for selection might arise, which is a much older and much harder question than the one the slogan answers.

Speculative Claim: “Prigogine already did this in the 1970s.” Frontier Dissipative structures are an observed phenomenon — order sustained by a through-flow of energy — and carry no claim that the ordering is selected by its dissipation history. Handwave The newer proposal is genuinely a different claim; whether it is a better one is what the missing experiment would decide.

Speculative Claim: “E. coli operates at the thermodynamic limit of replication.” Established The published figure is a factor of about 5.2 above the bound: 220 against a minimum of 42, per peptide bond, with internal entropy bounded above at 10. Frontier A factor of five is close for a biological system and is not the limit; the RNA ligation example, at 7 kcal mol⁻¹ against about 10 measured, is closer and carries no more causal weight.

Handwave Claim, made mostly in passing: “Everything far from equilibrium is doing dissipative adaptation.” Speculative With the bound valid for any transition between any two microstates, including undriven relaxation, and with the selection claim compatible with dissipation rising in one model and falling in another, the phrase has expanded to cover nearly any driven process. Frontier A category that excludes nothing predicts nothing, and the repair is the one section 7 specifies: a registered prediction, a spectral control, work resolved by outcome.