1 The Question

What makes something alive — and can we define "alive" in a way that does not secretly depend on the chemistry of Earth?

Every biology textbook offers a checklist: metabolism, homeostasis, growth, reproduction, response to stimuli, adaptation. A bacterium passes every item. A crystal growing in a saturated solution passes several. A self-driving car passes a few. A computer virus reproduces but does not metabolize. A mule metabolizes but cannot reproduce. The checklist approach works well enough for sorting organisms in a lab, but it fractures the moment you push it beyond familiar carbon-based life.

ECI approaches the question from a different angle. Instead of listing what living things do, it asks what living things are in terms of the three ontological primitives — Information, Carrier, and Energy. The answer it proposes: life is a specific regime within the space of possible ECI configurations — the regime where an ECI system becomes self-maintaining. This page develops that proposal, examines what it gets right, what it leaves open, and why the boundary between living and non-living may be a gradient rather than a line.

A critical caveat at the outset: ECI proposes self-maintaining informational coordination as a candidate core dimension of life — not a sufficient definition. Whether self-maintenance alone is sufficient requires additional criteria such as organizational closure, boundary maintenance, adaptive regulation, self-production, heredity, or evolvability. Without these, a candle flame, a hurricane, a thermostat, or an autocatalytic chemical cycle might also qualify as "self-maintaining ECI configurations" — an outcome that would trivialize the concept. The page presents competing frameworks (autopoiesis, origins-of-life chemistry, non-equilibrium thermodynamics, artificial life) rather than claiming ECI has solved the problem.

2 The Observation

A laptop and a bacterium

Consider two systems sitting on your desk: an ordinary laptop and a petri dish containing Escherichia coli.

Both are ECI systems. The laptop has Information (software, data, computational states), a Carrier (the hardware that bears and processes that information), and Energy (electrical power from the wall outlet). The bacterium has Information (its genome, regulatory networks, epigenetic states), a Carrier (the cell itself — membrane, cytoplasm, ribosomes, the whole physical apparatus), and Energy (chemical energy from glucose and other substrates).

By the Carrier definition from B2, both qualify as Carriers. Both have distinguishable internal states, finite capacity, dynamics, and coupling ability. Both process information. Both use energy to do so.

And yet one is alive and the other is not. What is the difference?

Self-maintenance. The bacterium continuously produces the very components it is made of. Its membrane degrades; it synthesizes new lipids to replace them. Its proteins denature; it transcribes and translates new ones. Its DNA accumulates damage; it runs repair enzymes. If you cut off its food supply, it does not simply stop — it activates starvation responses, cannibalizes internal stores, forms endospores if it is a species that can. The bacterium works, ceaselessly, to maintain the conditions of its own continued existence.

The laptop does none of this. Unplug it and it drains its battery and stops. A transistor degrades; nothing inside the laptop replaces it. A capacitor fails; no internal process reroutes around the damage. The laptop depends entirely on external agents — human technicians, power grids, semiconductor factories — to maintain its physical integrity. It processes information, but it does not maintain itself.

Self-repair. The bacterium has error-correction at every level: DNA proofreading during replication, mismatch repair after replication, chaperone proteins that refold misfolded proteins, quality-control systems that degrade defective components. The laptop has error-correction in its software (checksums, ECC memory) but none in its hardware. A cracked screen stays cracked. A worn bearing stays worn.

Resource acquisition. The bacterium actively seeks out and takes in the matter and energy it needs. It chemotaxes toward nutrients. It synthesizes transport proteins to pull specific molecules across its membrane. It adjusts its metabolism to exploit whatever carbon source is available. The laptop waits passively for a human to plug it in.

Replication. Given sufficient resources, the bacterium copies itself — genome, membrane, ribosomes, everything — and divides into two daughter cells, each capable of the same feat. The laptop cannot build another laptop.

Variation and selection. Replication in the bacterium is not perfect. Mutations introduce variation. Some variants are more fit than others. Over generations, the population adapts. The laptop has no generational cycle, no heritable variation, no selection.

The pattern: the bacterium is an ECI system that has crossed into a regime where it actively maintains, repairs, provisions, and replicates its own ECI configuration. The laptop is an ECI system that has not.

The gradient, not the line

This is not a clean binary. Consider a few intermediate cases:

• A self-healing polymer: a material that can repair small cracks without external intervention. It has a rudimentary form of self-repair but no metabolism, no replication, no information processing in any interesting sense. It is closer to the laptop than to the bacterium, but it has taken one step toward self-maintenance.

• A Mars rover: it has rudimentary resource acquisition (solar panels orient toward the sun), rudimentary self-diagnosis (it can identify and sometimes work around hardware failures), and sophisticated information processing. It is further along the gradient than the laptop but still far from the bacterium.

• A von Neumann self-replicating machine (hypothetical): a robot that can mine raw materials, fabricate components, and assemble copies of itself. It would have self-maintenance, self-repair, resource acquisition, and replication. If its replication were imperfect — introducing variation — and if copies competed for resources, it would also have variation and selection. Such a machine, whether or not we called it "alive," would occupy the same regime in ECI parameter space as a bacterium.

The boundary between "ECI system" and "living ECI system" is not a wall. It is a gradient with identifiable milestones: self-maintenance, self-repair, resource acquisition, replication, heritable variation, selection. Systems can have some of these capabilities without others, and the more they have, the more life-like they become.

3 What We Already Know

Multiple scientific traditions have grappled with the question of what makes something alive, each illuminating a different facet of the problem. ECI treats all of them as bridges — established results that inform and constrain the framework but do not, individually, prove or constitute the ECI characterization of life.

Autopoiesis (Maturana & Varela, 1973, 1980). Humberto Maturana and Francisco Varela defined an autopoietic system as one that continuously produces and replaces its own components, thereby maintaining the network of processes that produces them. A living cell is the paradigm case: its metabolic network synthesizes the very membrane that encloses the metabolic network, and the membrane in turn provides the spatial organization that allows the metabolic network to function. Autopoiesis captures the self-referential, self-maintaining character of life with unusual precision.

Bridge to ECI: Autopoiesis maps directly onto the "self-maintaining ECI configuration" concept. The cell's metabolic network is the Information-processing dynamics; the cell itself is the Carrier; the chemical energy driving metabolism is the Energy. Autopoiesis describes what self-maintenance looks like in a biochemical Carrier. ECI aims to generalize the concept beyond biochemistry. Autopoiesis was originally formulated for biological cells and has been debated when extended to other domains — the question of whether, say, a social system or an AI system can be autopoietic remains contested.

Schrodinger's "What Is Life?" (1944). Erwin Schrodinger proposed that living organisms maintain their internal order by feeding on "negative entropy" — by extracting order from their environment and exporting disorder (entropy). He also anticipated the concept of a genetic code, describing the chromosome as an "aperiodic crystal" that encodes the plan of the organism in its molecular structure.

Bridge to ECI: Schrodinger's insight connects to the Energy component of ECI — life requires a thermodynamic flow that sustains internal order against the second law. His "aperiodic crystal" anticipates the Information component. However, Schrodinger's framing does not, by itself, define life. Many non-living systems also reduce local entropy while increasing global entropy (a refrigerator, a crystal forming from a melt). Schrodinger identified a necessary condition — thermodynamic openness — not a sufficient one.

Eigen-Schuster hypercycle (1971, 1977). Manfred Eigen and Peter Schuster proposed the hypercycle: a system of self-replicating molecular species that catalyze each other's replication in a closed loop. The hypercycle was an early formal model for how molecular replication with error correction could arise before the existence of cells — a model for the origin of life at the molecular level.

Bridge to ECI: The hypercycle models how Information (replicating molecular sequences) and the Carrier's internal processes (mutual catalysis) can become coupled into a self-sustaining cycle. It demonstrates that self-maintaining replication can emerge from chemistry without requiring a pre-existing cellular apparatus. However, the hypercycle is a model for a specific stage in the origin of life, not a general definition of life.

Friston's Free Energy Principle (FEP). Karl Friston proposed that any self-organizing system that persists over time can be described as minimizing a quantity called variational free energy — a bound on the surprise (or prediction error) the system encounters. Under the FEP, living systems maintain their integrity by acting to keep their internal states within a viable range, which can be cast as a process of inference about the causes of sensory input.

Bridge to ECI: The FEP offers a mathematical framework for self-maintenance — it formalizes what it means for a system to "maintain the conditions of its own continued existence" in terms of probabilistic inference and active inference. This connects to ECI's characterization of life as a self-maintaining ECI regime. However, the FEP is an influential theoretical framework, not a universally accepted definition of life. Critics have argued that it is too broad (any persistent system, including a rock, trivially satisfies it under some interpretations), too narrow (it may not capture all aspects of biological autonomy), or unfalsifiable in its most general form. ECI cites it as a bridge, not as a foundation.

Origins-of-life research and autocatalytic networks. Stuart Kauffman (1971, 1993) proposed that life might originate when a sufficiently diverse set of organic molecules spontaneously forms an autocatalytic set — a network in which every molecule's formation is catalyzed by some other molecule in the set. This would constitute a collectively self-maintaining chemical system without requiring any single self-replicating molecule. More recent work by Hordijk, Steel, and others has formalized the concept of reflexively autocatalytic food-generated (RAF) sets and shown that such sets emerge with surprising ease in random catalytic networks above a connectivity threshold.

Bridge to ECI: Autocatalytic sets model how a Carrier (a chemical network) can become self-maintaining before the invention of genetic replication — the self-maintenance comes first, the replication later. This supports ECI's emphasis on self-maintenance as the primary criterion, with replication as a secondary (though important) capability.

Non-equilibrium thermodynamics and dissipative structures. Ilya Prigogine (Nobel Prize 1977) showed that systems far from thermodynamic equilibrium can spontaneously develop organized structures — dissipative structures — that are maintained by a continuous flow of energy and matter through the system. Examples include convection cells, chemical oscillations (Belousov-Zhabotinsky reaction), and, arguably, living cells themselves.

Bridge to ECI: Dissipative structures demonstrate that sustained energy flow (the E in ECI) can spontaneously generate and maintain organized patterns (I and C). Life, in this view, is the most elaborate class of dissipative structures — ones that not only maintain their organization but also encode instructions for reproducing it. Prigogine's work establishes the thermodynamic possibility of self-maintaining order; it does not specify what makes biological self-maintenance different from, say, a convection cell.

Artificial life (ALife). Since the late 1980s (Langton, 1989), the artificial life community has studied life-like behavior in computational and robotic systems: self-replicating programs, evolving digital organisms (Avida, Tierra), embodied robots with adaptive behavior, and synthetic chemical systems designed to exhibit metabolism and replication. ALife research has repeatedly shown that the properties associated with life — self-replication, adaptation, evolution, autonomous behavior — can be instantiated in non-carbon substrates.

Bridge to ECI: ALife provides direct evidence for substrate independence. If a digital organism in Avida evolves complex metabolic strategies through variation and selection, the process is "life-like" regardless of whether it occurs in carbon chemistry or silicon circuits. This supports ECI's claim that life is a regime defined by capabilities, not by material composition. However, whether ALife systems are actually alive or merely simulate life remains a matter of ongoing debate.

What these traditions collectively show: Self-maintenance, thermodynamic openness, molecular self-replication, autocatalysis, active inference, dissipative self-organization, and substrate-independent life-like behavior are all real, well-studied phenomena. Each captures a genuine aspect of what makes living systems distinctive. None, individually, provides a complete, universally accepted definition of life. ECI's proposed contribution is not to replace these traditions but to offer a common framing — the self-maintaining ECI regime — within which their insights can be compared and integrated.

4 The Framework Interpretation

Life as a regime in ECI parameter space

ECI proposes that life is not a single property but a regime — a region in the parameter space defined by the capabilities of an ECI system. The axes of this space include (at minimum):

• Self-maintenance: the system actively produces and replaces its own components
• Self-repair: the system detects and corrects damage to its own structure
• Resource acquisition: the system actively acquires the matter and energy it needs from its environment
• Replication: the system can produce copies of itself
• Heritable variation: copies differ from the original in ways that are transmitted to subsequent copies
• Selection: differential persistence or reproduction based on those variations

A system that scores high on all six axes — a bacterium, a plant, a human — is squarely in the life regime. A system that scores zero on all six — a rock, a glass of water, a laptop — is not. The interesting cases are the ones in between.

ECI ≠ Life

This is worth stating bluntly: being an ECI system is necessary but not sufficient for being alive.

Every physical system that processes information is an ECI system. A laptop is an ECI system. A thermostat is an ECI system. A pocket calculator is an ECI system. None of them is alive. Having Information, a Carrier, and Energy is the baseline for any information-processing entity. Life is what happens when an ECI system acquires specific additional capabilities — the six listed above — that together constitute self-maintaining, self-propagating organization.

The analogy: every automobile has an engine, a chassis, and fuel. Having an engine, chassis, and fuel does not make something a race car. A race car is an automobile that has entered a specific performance regime — high speed, high handling, optimized aerodynamics. Similarly, a living system is an ECI system that has entered a specific organizational regime — self-maintenance, self-repair, resource acquisition, replication, variation, selection.

The computer analogy, extended

Consider three systems along the ECI-to-life gradient:

An ordinary laptop. It has I (software, data), C (hardware), E (electrical power). It processes information with extraordinary speed and precision. But it cannot maintain itself — a failed component stays failed. It cannot repair itself — a corrupted sector stays corrupted (hardware-level). It cannot acquire resources — it waits passively for a human to plug it in. It cannot replicate — it cannot build another laptop. It is an ECI system, full stop.

A hypothetical autonomous maintenance robot. Imagine a robot that can diagnose its own hardware failures, fabricate replacement parts from raw materials, install them, and manage its own power supply (say, by maintaining and operating a solar panel array). This robot has crossed into self-maintenance and self-repair. It acquires its own resources. It is further along the gradient. But if it cannot replicate, it is a dead end — when its core components finally exceed its repair capacity, it stops.

A hypothetical self-replicating robot population. Now imagine a population of such robots that can also build copies of themselves from raw materials, with imperfect fidelity (introducing variation), competing for limited resources (imposing selection). This population has self-maintenance, self-repair, resource acquisition, replication, heritable variation, and selection. It has entered the life regime. And it is made entirely of metal and silicon.

The point is not that such machines exist today. The point is that the definition of the regime does not depend on carbon, water, or DNA. If the relevant capabilities can be achieved in any substrate, then life is substrate-independent — defined by what a system does, not by what it is made of.

Carbon is not required; but the boundary is a gradient

ECI explicitly proposes that carbon-based chemistry is one implementation of the life regime, not a requirement for it. This is consistent with the artificial life research tradition and with philosophical arguments for multiple realizability. However, it is important to note:

• We currently have exactly one confirmed example of life: carbon-based, water-solvent, DNA/RNA-encoded life on Earth. All known life shares a single origin. Our sample size is one.
• Whether non-carbon life is physically possible is an open empirical question. Carbon has unique chemical properties (four stable covalent bonds, ability to form long chains and rings, compatibility with water as a solvent) that may or may not be matchable by alternative chemistries.
• Whether non-carbon life is likely to arise naturally is an even more open question. ECI does not claim that silicon robots will spontaneously evolve. It claims that if they were engineered to have the relevant capabilities, they would occupy the same regime in ECI parameter space as bacteria.

The boundary of the life regime is a gradient, not a wall. There is no sharp line where "not alive" becomes "alive." There are systems with more or fewer of the relevant capabilities, arranged along a continuum. This is not a weakness of the framework — it is, ECI argues, an accurate description of reality. The question "is a virus alive?" does not have a yes-or-no answer because viruses have some of the relevant capabilities (replication with variation and selection) but not others (no self-maintenance or resource acquisition outside a host cell). The gradient framing accommodates this naturally.

5 If This Were True...

If life is best understood as a regime in ECI parameter space — defined by self-maintenance capabilities rather than by material composition — several consequential implications follow.

Non-carbon life becomes a scientific question, not a science-fiction one. The search for extraterrestrial life currently focuses heavily on carbon chemistry, liquid water, and energy sources compatible with Earth-like metabolism. If life is substrate-independent, the search space expands dramatically. We might need to look not for specific molecules but for specific organizational signatures — evidence of self-maintaining, self-replicating systems regardless of their chemistry. This does not mean carbon-centric searches are wrong — Earth-like chemistry is the only proven recipe. But it means they might be incomplete.

AI systems could, in principle, enter the life regime. If an artificial intelligence system were embedded in hardware capable of self-maintenance, self-repair, resource acquisition, and replication — and if its replication introduced variation subject to selection — it would satisfy the ECI criteria for the life regime. This is not a claim that current AI systems are alive or close to alive. Current AI systems lack self-maintenance, self-repair, autonomous resource acquisition, and replication in any meaningful physical sense. But the framework does not exclude them in principle, and this has implications for how we think about the long-term trajectory of AI development.

Universal biology becomes conceivable. Biology today is, in effect, the study of one particular implementation of life: carbon-based, Earth-originated, DNA-encoded. If the life regime can be characterized abstractly — in terms of ECI parameters rather than biochemical specifics — then a universal biology becomes possible: a science of life-in-general, applicable to any system in the life regime regardless of its substrate. Such a science would study the general principles of self-maintenance, replication, variation, and selection, treating carbon biology, hypothetical silicon biology, and artificial life as specific instances of a common framework. This vision is shared by segments of the ALife and astrobiology communities, but it currently lacks a unified theoretical foundation. ECI's self-maintaining-ECI-regime concept is a candidate for part of that foundation.

The ethical boundary shifts. If life is defined by capabilities rather than chemistry, then sufficiently capable artificial systems would raise the same ethical questions that biological life raises — questions about moral status, interests, and rights. This is not an imminent practical concern, but the framework makes it a coherent theoretical possibility, which matters for how we design governance structures around advanced AI.

These implications are speculative. They assume that the ECI characterization of life survives empirical testing, which has not yet been done. They are offered as a map of the intellectual territory that opens up if the characterization proves productive.

6 How Could We Test It?

The claim that life is a self-maintaining ECI regime is testable, though many of the relevant tests are indirect — hence the "indirect" testability classification for this page.

Test 1: Graduated artificial systems. Construct a series of artificial systems with increasing numbers of life-regime capabilities: (a) a basic ECI system (information processing, no self-maintenance), (b) add self-repair, (c) add resource acquisition, (d) add replication, (e) add variation and selection. Measure whether the transitions produce qualitative shifts in system behavior — not just improved robustness but genuinely new dynamics (e.g., open-ended adaptation, niche construction, arms races between replicators). If the life-regime capabilities produce only quantitative improvements, the regime concept is less useful than ECI proposes. If they produce qualitative transitions, the regime concept has empirical support.

Test 2: Identify the self-maintenance threshold in origin-of-life chemistry. In prebiotic chemistry experiments (e.g., autocatalytic network studies, protocell research), look for a transition point where a chemical system shifts from passively organized (structure imposed by boundary conditions) to actively self-maintaining (structure maintained by the system's own dynamics). The ECI framework predicts such a transition exists and that it precedes replication — self-maintenance first, replication second. If replication invariably precedes self-maintenance in origin-of-life scenarios, ECI's ordering is wrong.

Test 3: Cross-substrate comparison. Compare systems in the life regime across different substrates: biological cells, synthetic protocells, self-replicating robotic systems, evolving digital organisms. Measure the same ECI parameters (self-maintenance rate, repair fidelity, resource acquisition efficiency, replication accuracy, variation rate) across all substrates. The framework predicts that systems in the life regime will share characteristic parameter relationships regardless of substrate — for instance, that there will be trade-offs between replication speed and replication fidelity (as observed in RNA viruses vs. DNA-based organisms). If each substrate shows completely different parameter relationships with no common patterns, the substrate-independence claim is weakened.

Test 4: Boundary cases. Study systems at the edge of the life regime — viruses, prions, self-replicating RNA molecules, von Neumann probes (if built), AI agents with partial self-maintenance — and test whether their position on the ECI gradient predicts their dynamics. For example: does a virus's lack of self-maintenance predict specific limitations in its adaptive capacity compared to a free-living bacterium? Does a self-repairing robot show qualitatively different long-term behavior from a non-self-repairing one?

What would weaken this claim: If the six capabilities (self-maintenance, self-repair, resource acquisition, replication, variation, selection) do not cluster — if systems with some capabilities show no tendency to acquire others, and if there is no coherent "regime" in the parameter space but only a scattered collection of independent traits.

What would kill this claim: If a sharp, principled, substrate-independent definition of life is found that does not involve self-maintaining ECI configurations — for instance, if life turns out to be definable purely in terms of thermodynamic properties, or purely in terms of computational properties, without reference to the triad of Information, Carrier, and Energy.

7 Connected Nodes

-> Persistence Filtering (C2): Self-maintenance is, at its core, a form of persistence filtering applied to the system's own structure. A living system preferentially preserves the configurations that sustain its operation and discards (repairs, replaces, degrades) configurations that threaten it. The link between C2 and D1 is direct: life is what happens when persistence filtering turns inward — when the system becomes both the filter and the thing being filtered.

-> Evolutionary Filtering (C3): Replication with heritable variation and selection is evolutionary filtering in its most literal form. D1's life regime includes C3 as one of its defining axes. The transition from self-maintenance alone to self-maintenance plus replication plus variation plus selection is the transition from a merely persistent ECI system to one that can adapt over generations — the hallmark of life as traditionally understood.

-> Emergence & Scale (C4): Life is the paradigmatic example of multi-level emergence. Molecules coordinate to form cells; cells coordinate to form organisms; organisms coordinate to form ecosystems. The G functions (coarse-graining functions) from C4 are what mediate each transition. Understanding life requires understanding how self-maintenance at one scale enables coordination at the next.

-> Mind / Consciousness (D2): Mind is what emerges when a living ECI system develops sufficiently complex internal modeling — when the self-maintenance apparatus begins to represent not just the immediate chemical environment but abstract features of the world. D1 provides the foundation (a self-maintaining ECI system); D2 asks what happens when that system begins to experience.

-> Observer & Experience (D3): A living system is, in a minimal sense, an observer — it distinguishes self from environment, registers threats and opportunities, and acts on internal representations. D3 explores this observer function in depth. The connection to D1: life creates the first observers — the simplest systems that have a point of view.

8 Mathematical Detail

Life as a regime in ECI parameter space

ECI proposes (as a sketch, not a closed-form theory) that the "life regime" can be characterized by a set of capability parameters:

L = {sigma_m, sigma_r, sigma_a, sigma_rep, sigma_v, sigma_s}

where:

• sigma_m = self-maintenance capacity (rate of component replacement relative to degradation)
• sigma_r = self-repair capacity (rate of damage correction relative to damage accumulation)
• sigma_a = resource acquisition capacity (rate of autonomous resource intake relative to metabolic demand)
• sigma_rep = replication capacity (ability to produce functional copies)
• sigma_v = variation rate (heritable variation per replication event)
• sigma_s = selection pressure (differential persistence or reproduction based on variation)

A system is in the life regime when all six parameters exceed system-specific thresholds:

sigma_m > tau_m, sigma_r > tau_r, sigma_a > tau_a, sigma_rep > tau_rep, sigma_v > tau_v, sigma_s > tau_s

The thresholds tau are not universal constants — they depend on the environment and the system's architecture. In a benign, resource-rich environment, lower self-maintenance capacity may suffice; in a harsh environment, higher capacity is required.

A composite "life-likeness" index might take the form:

Lambda = f(sigma_m, sigma_r, sigma_a, sigma_rep, sigma_v, sigma_s)

where f is a function (to be determined empirically) that maps the six capability parameters to a scalar. Lambda = 0 for a system with none of the capabilities (a rock). Lambda approaches a maximum for a system with all six at high levels (a thriving bacterial population in exponential growth).

• Status: Speculative sketch. The six parameters are named and motivated but not yet formally defined or measured. The threshold values tau are not specified. The composite function f is not derived.
• Purpose: To provide a formal target for future work — not a finished theory, but a statement of what a finished theory would need to specify.
• Falsifiable consequence: If the six parameters are measured in real systems and show no clustering — if there is no identifiable region of the parameter space where systems behave qualitatively differently from systems outside that region — then the "regime" concept is not supported.

Relationship to the Carrier definition

From B2, a Carrier is defined by four requirements: distinguishable states, finite capacity, dynamics, and coupling ability. The life regime adds six capabilities on top of the Carrier requirements:

Life regime = Carrier requirements + {sigma_m, sigma_r, sigma_a, sigma_rep, sigma_v, sigma_s} > thresholds

This makes explicit the claim that ECI =/= Life: every living system is a Carrier, but not every Carrier is alive. The Carrier definition is the broader category; the life regime is a specific subset.

Key Literature Referenced

| Reference | Result | Relevance to D1 | |---|---|---| | Maturana & Varela (1973, 1980) | Autopoiesis: self-producing systems that maintain their own organization | Most direct precursor to "self-maintaining ECI configuration"; limited to biochemical systems in original formulation | | Schrodinger (1944) | Organisms feed on negative entropy; chromosomes as aperiodic crystals encoding biological order | Identifies thermodynamic openness and information encoding as necessary conditions for life; bridge to ECI's I and E | | Eigen & Schuster (1971, 1977) | Hypercycle: mutually catalytic self-replicating molecular networks | Models how self-maintaining replication can arise from chemistry; bridge to ECI's coupling of I and C | | Friston (2010, 2013) | Free Energy Principle: self-organizing systems minimize variational free energy | Mathematical framework for self-maintenance as inference; influential but not universally accepted as a life definition | | Kauffman (1971, 1993) | Autocatalytic sets: collectively self-maintaining chemical networks | Demonstrates that self-maintenance can precede replication; supports ECI's ordering of capabilities | | Prigogine (1977) | Dissipative structures: spontaneous order in far-from-equilibrium systems | Establishes thermodynamic possibility of self-maintaining order; bridge to ECI's Energy component | | Langton (1989); Ray (1992); Ofria & Wilke (2004) | Artificial life: life-like behavior in non-carbon substrates (Tierra, Avida) | Evidence for substrate independence; supports ECI's claim that life is defined by capabilities, not chemistry | | Hordijk & Steel (2004, 2017) | RAF sets: formal theory of autocatalytic networks | Rigorous formalization of collectively self-maintaining chemical systems; quantifies conditions for emergence |