Reduced-iron sulfide surface that binds dissolved metals and hosts early thioester-forming chemistry.
abiogenesis-chemical-evolution is a complete, interactive, scientifically grounded simulation of abiogenesis and chemical evolution. Its product intent is to test one question: can an initially nonliving chemical environment, continuously driven away from thermodynamic equilibrium by external energy, spontaneously produce persistent autocatalytic systems, compartments, heredity, competition, evolution, and increasingly organism-like entities?
The governing rule of the product is "Script the laws. Do not script life." The simulation begins with chemistry, matter, energy, and environmental gradients. If organized systems emerge, they emerge only because the underlying physical and chemical rules make them possible. The intended causal chain is:
physics → chemistry → reaction networks → self-organization → compartments → replication → variation → differential persistence → evolution
— and only if the simulated conditions allow those things to occur.
The audience is synthetic biologists, origin-of-life researchers, computational chemists, science-curious learners, and museum-grade exhibition visitors. The product is a laboratory instrument with the gravitas of a gallery piece: the data is the aesthetic, and the emotional register is awe, contemplative wonder, and intellectual rigour — the feeling of watching something appear that was never told to appear.
The system is a web-delivered, backend-executed simulation environment. Users configure an early-Earth chemical world — starting chemical inventory, concentrations, protonation states, environmental gradients, external energy inputs, and mineral surfaces — then start, pause, and advance the simulation over time. The backend performs the actual physical and chemical computation: reaction kinetics, acid/base equilibria, diffusion, gas exchange, mineral surface interactions, amphiphile self-assembly, and the longitudinal tracking of emergent structures. The frontend renders the evolving state as a full-bleed generative visualization and provides instrument-panel inspection surfaces.
Actors. Three accepted human personas operate the product: the Simulation Operator (configures and runs the experiment), the Chemistry Inspector (examines emergent chemistry in detail), and the Emergence Observer (tracks self-organized and increasingly organism-like structures). The simulation engine itself is a non-persona system actor.
Accepted behavior. The system must model water as an explicit bulk continuum solvent; derive local pH from acid/base concentrations; carry a defined initial early-Earth chemical inventory; allow gas exchange between atmosphere and water by solubility and partial pressure; support an extensible chemical grammar that constructs previously unseen species from valid chemical transformations; model mineral surfaces and their adsorption/catalytic roles; and allow amphiphiles to self-assemble into micelles, sheets, droplets, and vesicles that exchange molecules, grow, shrink, fuse, and rupture.
Ownership. Application-owned identity governs access to durable simulation configurations, runs, observations, and inspection history. Configuration and execution surfaces are restricted to the Simulation Operator; inspection surfaces are restricted to the Chemistry Inspector and Emergence Observer respectively.
Narrow exclusions. This is explicitly not a traditional artificial-life simulation. Organisms, metabolism, reproduction, DNA, food, predators, fitness, and species are never manually scripted. There is no spawnLife() equivalent, no if molecule == RNA then reproduce logic, no organism.health += energy logic, and no manually assigned evolutionary fitness score. No organism-like property is ever granted to a structure merely because it has been labeled an organism.
The product is delivered as a first-party web application with a backend simulation engine. Identity is application-owned: users establish access through self-service enrollment and return through verification, because durable simulation configurations, run history, observations, and inspection state must remain bound to the correct participant and be resumable across sessions. The anonymous entry surface (Landing) explains the simulation and its law-driven purpose before identity is established; protected configuration, execution, and inspection surfaces remain unavailable until identity is established.
Everything described in this document is current scope. The simulation's emergent outcomes — autocatalytic sets, compartments, heredity, competition, evolution — are not features to be built; they are possible results of the scripted laws and are tracked and observed when they occur. No capability in this document may be implemented by scripting a life outcome.
The authoritative source defines a required initial early-Earth chemical inventory. This inventory is the starting chemistry, not the universe of possible chemistry; additional species are constructed dynamically from valid chemical transformations.
Solvent and acid/base chemistry: H₂O; H\xe2\x81\xba / H\xe2\x82\x83O\xe2\x81\xba; OH⁻. Water is modeled explicitly as the bulk solvent at the continuum level rather than rendering every water molecule. Local pH emerges from acid/base concentrations.
Major dissolved ions: Na\xe2\x81\xba, K\xe2\x81\xba, Mg²\xe2\x81\xba, Ca²\xe2\x81\xba, Fe²\xe2\x81\xba, Fe³\xe2\x81\xba, Ni²\xe2\x81\xba, Zn²\xe2\x81\xba, Mn²\xe2\x81\xba, Cl⁻, HS⁻, S²⁻, SO\xe2\x82\x83²⁻, SO\xe2\x82\x84²⁻, HCO\xe2\x82\x83⁻, CO\xe2\x82\x83²⁻, H₂PO\xe2\x82\x84⁻, HPO\xe2\x82\x84²⁻, PO\xe2\x82\x84³⁻. Their concentrations must influence ionic strength, membrane stability, polymer folding, precipitation, and reaction rates.
Atmospheric / geochemical feedstock: H₂, N₂, CO₂, CO, CH\xe2\x82\x84, NH\xe2\x82\x83, NH\xe2\x82\x84\xe2\x81\xba, H₂S, SO₂. Gas exchange between atmosphere and water occurs according to solubility and partial pressure.
Important prebiotic carbon/nitrogen compounds: HCN (hydrogen cyanide); HNC (hydrogen isocyanide, where appropriate); NH₂CN (cyanamide); HC\xe2\x82\x83N (cyanoacetylene); HCONH₂ (formamide); HCHO (formaldehyde); CH\xe2\x82\x83CHO (acetaldehyde); CH\xe2\x82\x83OH (methanol); HOCH₂CHO (glycolaldehyde); glyceraldehyde; dihydroxyacetone; glyoxal. These participate in kinetic reaction networks and are not merely collectible resources.
Simple organic acids: formic acid / formate; acetic acid / acetate; glycolic acid; lactic acid / lactate; pyruvic acid / pyruvate; oxalic acid / oxalate; malonic acid; succinic acid / succinate; fumaric acid / fumarate; malic acid / malate. Additional carboxylic acids may emerge.
Amino acids (at minimum): glycine, alanine, serine, aspartic acid, glutamic acid, valine, leucine, isoleucine, proline, threonine, cysteine, methionine, phenylalanine. Modern protein synthesis is not assumed; these are simply molecules capable of reactions including condensation into short peptides. Additional amino acids and non-biological amino-acid analogues may form.
Sugars and sugar precursors: formaldehyde; glycolaldehyde; glyceraldehyde; ribose; arabinose; xylose; erythrose; ribulose; simple hexoses when reaction pathways permit. Instability and degradation are included; ribose is not artificially immortal.
Nucleobases and precursors: adenine, guanine, cytosine, uracil, purine, pyrimidine, and related prebiotic heterocycles. Thymine can arise later but does not need to dominate primordial chemistry.
Nucleosides and nucleotides: adenosine, guanosine, cytidine, uridine; AMP, GMP, CMP, UMP. Activated versions are allowed where chemically justified. ADP and ATP should be possible products of chemistry but are not supplied as the universal primordial energy currency; primitive systems should initially be capable of using other energetic chemistry.
Phosphorus chemistry: orthophosphate; pyrophosphate; trimetaphosphate; other condensed phosphates; acetyl phosphate. Phosphate adsorption to minerals and the realistic difficulty of phosphorylation are included.
Sulfur chemistry and energetic intermediates: hydrogen sulfide; bisulfide; sulfide; elemental sulfur where relevant; thiols; methanethiol; simple thioesters; acetyl thioesters. Thioester chemistry should be capable of coupling energetically favorable and unfavorable reactions.
Amphiphiles (families, not a single magic lipid): fatty acids approximately C6–C18 (octanoic acid, decanoic acid, lauric acid, myristic acid, palmitic acid); simple fatty alcohols; glycerol; monoacylglycerols; simple amphiphilic esters. Amphiphilic properties must arise from molecular structure.
Mineral world: FeS / mackinawite; FeS₂ / pyrite; Fe\xe2\x82\x83S\xe2\x82\x84 / greigite; NiS; magnetite; silica; serpentine minerals; olivine-related minerals; brucite; carbonate minerals; iron oxides; clay minerals such as montmorillonite; phosphate-bearing minerals.
FR-1 — Build the simulation. As a Simulation Operator, I should have a complete, interactive, scientifically grounded simulation of abiogenesis and chemical evolution, so that I can run the origin-of-life experiment. (explicit)
FR-2 — No traditional artificial-life scripting. As a Simulation Operator, I should have a simulation in which organisms, metabolism, reproduction, DNA, food, predators, fitness, and species are never manually scripted, so that any organized system that appears is genuinely emergent. (explicit)
FR-3 — Script the laws, not life. As a Simulation Operator, I should have the simulation begin with chemistry, matter, energy, and environmental gradients, so that organized systems emerge only because the underlying physical and chemical rules make them possible. (explicit)
FR-4 — The ultimate experiment. As a Simulation Operator, I should be able to determine whether an initially nonliving chemical environment, continuously driven away from thermodynamic equilibrium by external energy, spontaneously produces persistent autocatalytic systems, compartments, heredity, competition, evolution, and increasingly organism-like entities. (explicit)
FR-5 — No spawnLife() equivalent. As a Simulation Operator, I should never encounter a command equivalent to spawnLife(), so that life is never directly created. (explicit)
FR-6 — No if molecule == RNA then reproduce logic. As a Simulation Operator, I should never encounter logic equivalent to if molecule == RNA then reproduce, so that replication is never triggered by a molecule label. (explicit)
FR-7 — No organism.health += energy logic. As a Simulation Operator, I should never encounter logic equivalent to organism.health += energy, so that organism health is never directly incremented. (explicit)
FR-8 — No manually assigned evolutionary fitness score. As a Simulation Operator, I should never encounter a manually assigned evolutionary fitness score, so that differential persistence arises only from the chemistry. (explicit)
FR-9 — Intended causal chain. As a Simulation Operator, I should observe the causal chain physics → chemistry → reaction networks → self-organization → compartments → replication → variation → differential persistence → evolution, if the simulated conditions allow those things to occur. (explicit)
FR-10 — Life as a special organizational state of ordinary matter. As a Simulation Operator, I should have the simulation begin with no distinction between "living" and "nonliving," with everything consisting of chemicals interacting according to the same rules. (explicit)
FR-11 — Organism-like properties are earned, not labeled. As a Simulation Operator, I should have a structure become organism-like only because it maintains chemical disequilibria, captures usable free energy, acquires raw materials, catalyzes reactions, repairs or replaces components, preserves internal conditions, contains reaction networks, creates copies or descendants, transmits chemically encoded information, produces imperfect descendants, competes for limited resources, and consequently undergoes natural selection. (explicit)
FR-12 — No organism-like properties by label. As a Simulation Operator, I should never see organism-like properties given to a structure merely because it has been labeled an organism. (explicit)
FR-13 — Extensible chemical grammar. As a Simulation Operator, I should have an extensible chemical grammar rather than an attempt to enumerate all possible chemicals, so that previously unseen molecular species can be constructed dynamically from valid chemical transformations. (explicit)
FR-14 — Molecular species properties. As a Chemistry Inspector, I should be able to inspect each molecular species' elemental formula, molecular graph, bond topology, molecular mass, formal charge, acid/base groups, oxidation state, polarity, hydrophobicity, solubility, diffusion coefficient, absorption spectrum, approximate heat capacity, bond energies, activation-energy estimates, membrane permeability, adsorption affinity for mineral surfaces, catalytic groups, degradation pathways, polymerization sites, and stereochemistry where relevant. (explicit)
FR-15 — Initial library is starting chemistry. As a Simulation Operator, I should have the initial chemical library treated as the starting chemistry, not the universe of possible chemistry. (explicit)
FR-16 — Solvent and acid/base chemistry. As a Simulation Operator, I should have H₂O, H\xe2\x81\xba / H\xe2\x82\x83O\xe2\x81\xba, and OH⁻ present, with water modeled explicitly as the bulk solvent at the continuum level rather than rendering every water molecule, and local pH emerging from acid/base concentrations. (explicit)
FR-17 — Major dissolved ions. As a Simulation Operator, I should have Na\xe2\x81\xba, K\xe2\x81\xba, Mg²\xe2\x81\xba, Ca²\xe2\x81\xba, Fe²\xe2\x81\xba, Fe³\xe2\x81\xba, Ni²\xe2\x81\xba, Zn²\xe2\x81\xba, Mn²\xe2\x81\xba, Cl⁻, HS⁻, S²⁻, SO\xe2\x82\x83²⁻, SO\xe2\x82\x84²⁻, HCO\xe2\x82\x83⁻, CO\xe2\x82\x83²⁻, H₂PO\xe2\x82\x84⁻, HPO\xe2\x82\x84²⁻, and PO\xe2\x82\x84³⁻ present, with their concentrations influencing ionic strength, membrane stability, polymer folding, precipitation, and reaction rates. (explicit)
FR-18 — Atmospheric / geochemical feedstock and gas exchange. As a Simulation Operator, I should have H₂, N₂, CO₂, CO, CH\xe2\x82\x84, NH\xe2\x82\x83, NH\xe2\x82\x84\xe2\x81\xba, H₂S, and SO₂ present, with gas exchange between atmosphere and water according to solubility and partial pressure. (explicit)
FR-19 — Prebiotic carbon/nitrogen compounds in kinetic networks. As a Simulation Operator, I should have HCN, HNC where appropriate, NH₂CN (cyanamide), HC\xe2\x82\x83N (cyanoacetylene), HCONH₂ (formamide), HCHO (formaldehyde), CH\xe2\x82\x83CHO (acetaldehyde), CH\xe2\x82\x83OH (methanol), HOCH₂CHO (glycolaldehyde), glyceraldehyde, dihydroxyacetone, and glyoxal present and participating in kinetic reaction networks rather than being treated merely as collectible resources. (explicit)
FR-20 — Simple organic acids. As a Simulation Operator, I should have formic acid / formate, acetic acid / acetate, glycolic acid, lactic acid / lactate, pyruvic acid / pyruvate, oxalic acid / oxalate, malonic acid, succinic acid / succinate, fumaric acid / fumarate, and malic acid / malate present, and additional carboxylic acids should be able to emerge. (explicit)
FR-21 — Amino acids. As a Simulation Operator, I should have at minimum glycine, alanine, serine, aspartic acid, glutamic acid, valine, leucine, isoleucine, proline, threonine, cysteine, methionine, and phenylalanine present as molecules capable of reactions including condensation into short peptides, without assuming modern protein synthesis, and additional amino acids and non-biological amino-acid analogues may form. (explicit)
FR-22 — Sugars and sugar precursors with instability. As a Simulation Operator, I should have formaldehyde, glycolaldehyde, glyceraldehyde, ribose, arabinose, xylose, erythrose, ribulose, and simple hexoses when reaction pathways permit, including instability and degradation, and ribose must not be artificially immortal. (explicit)
FR-23 — Nucleobases and precursors. As a Simulation Operator, I should have adenine, guanine, cytosine, uracil, purine, pyrimidine, and related prebiotic heterocycles present, with thymine able to arise later without dominating primordial chemistry. (explicit)
FR-24 — Nucleosides and nucleotides. As a Simulation Operator, I should have adenosine, guanosine, cytidine, uridine, AMP, GMP, CMP, and UMP present, with activated versions allowed where chemically justified, and ADP and ATP should be possible products of chemistry but not supplied as the universal primordial energy currency, with primitive systems initially capable of using other energetic chemistry. (explicit)
FR-25 — Phosphorus chemistry. As a Simulation Operator, I should have orthophosphate, pyrophosphate, trimetaphosphate, other condensed phosphates, and acetyl phosphate present, including phosphate adsorption to minerals and the realistic difficulty of phosphorylation. (explicit)
FR-26 — Sulfur chemistry and energetic intermediates. As a Simulation Operator, I should have hydrogen sulfide, bisulfide, sulfide, elemental sulfur where relevant, thiols, methanethiol, simple thioesters, and acetyl thioesters present, with thioester chemistry capable of coupling energetically favorable and unfavorable reactions. (explicit)
FR-27 — Amphiphile families. As a Simulation Operator, I should have amphiphiles as families rather than a single magic lipid — fatty acids approximately C6–C18 (octanoic acid, decanoic acid, lauric acid, myristic acid, palmitic acid), simple fatty alcohols, glycerol, monoacylglycerols, and simple amphiphilic esters — with amphiphilic properties arising from molecular structure. (explicit)
FR-28 — Spontaneous amphiphile self-assembly. As an Emergence Observer, I should see amphiphiles above appropriate concentrations spontaneously form micelles, sheets, droplets, and vesicles, and exchange molecules, grow, shrink, fuse, and rupture, with no "cell membrane object" scripted first — a membrane originates from amphiphile self-assembly. (explicit)
FR-29 — Coarse-grained membrane model permitted. As a Simulation Operator, I should have the option for fully atomistic membrane simulation to be replaced with a validated coarse-grained membrane model for computational scalability. (explicit)
FR-30 — Mineral world. As a Simulation Operator, I should have mineral surfaces present — FeS / mackinawite, FeS₂ / pyrite, Fe\xe2\x82\x83S\xe2\x82\x84 / greigite, NiS, magnetite, silica, serpentine minerals, olivine-related minerals, brucite, carbonate minerals, iron oxides, clay minerals such as montmorillonite, and phosphate-bearing minerals — because early-Earth chemistry did not occur in an empty beaker. (explicit)
FR-31 — Self-service enrollment. As a new user, I should be able to enroll myself so that I can independently start using the simulation. (required_inference)
FR-32 — Returning verification. As a returning user, I should be able to verify my identity so that I can resume access to durable simulation configurations, runs, observations, and inspection history. (required_inference)
FR-33 — Role-aware authorization. As a Simulation Operator, I should have configuration and execution access, and as a Chemistry Inspector or Emergence Observer, I should have observational inspection access, so that the correct participant reaches the correct surface. (required_inference)
FR-34 — Backend execution. As a Simulation Operator, I should have durable simulation state, chemical transformations, emergent structures, and longitudinal tracking executed on the backend, so that runs persist and can be resumed. (required_inference)
Product context. The Simulation Operator is the person who sets up and runs the origin-of-life experiment. They work at the configuration and execution layer: they define the starting chemical inventory, the environmental gradients and external energy inputs, and the mineral surfaces, then start, pause, and advance the simulation over time.
Primary goal. To observe whether persistent autocatalytic systems, compartments, heredity, and evolution emerge from the scripted physical and chemical rules without any scripted life, and to inspect the resulting chemical and organizational state.
Distinct accepted responsibilities. The Operator owns the configuration surfaces (Chemistry Setup, Energy Setup, Minerals) and the execution surface (Simulation). They commit the starting chemical inventory, concentrations, and protonation states from the extensible chemical grammar; they commit the environmental gradients and external energy inputs that drive the environment away from equilibrium; they commit the mineral surface configuration; and they start, pause, advance, and reset the run.
Relevant inputs or decisions. Which species and concentrations to start from; which protonation states to use; which gradients and energy inputs to apply; which mineral surfaces to include and at what abundance; when to start, pause, advance, or reset.
Interactions with other accepted participants. The Operator's configuration and execution produce the state that the Chemistry Inspector and Emergence Observer inspect. The Operator does not perform the inspection work; they produce the run that makes it possible.
Observable success. A run that advances according to the scripted laws, with the configured chemistry, gradients, and minerals in effect, and with the emergent-systems census updating.
Product context. The Chemistry Inspector examines the emergent chemistry in detail. They work at the inspection layer: molecular species, reaction networks, concentrations, pH, ionic strength, and the properties of dynamically constructed molecules.
Primary goal. To verify that observed structures and reaction pathways arise from valid chemical transformations and the underlying rules rather than from scripted outcomes.
Distinct accepted responsibilities. The Inspector owns the Chemistry and Molecules surfaces. They inspect the evolving molecular species, concentrations, local pH, ionic strength, and reaction networks; they inspect individual and dynamically constructed molecular species and their full property set; and they trace the valid chemical transformations that produced a species.
Relevant inputs or decisions. Which species or reaction network to inspect; which dynamically constructed molecule to examine; which property to verify.
Interactions with other accepted participants. The Inspector depends on the Simulation Operator's run for the state they inspect. Their verification work is distinct from the Emergence Observer's structural tracking.
Observable success. A verified chain of valid chemical transformations behind every observed species and pathway, with no scripted outcome.
Product context. The Emergence Observer watches for and tracks the appearance of self-organized structures such as micelles, sheets, droplets, and vesicles, and any autocatalytic or replicating systems that arise.
Primary goal. To detect and follow the emergence of increasingly organism-like entities and differential persistence over time, confirming that nothing was explicitly told to the program to make.
Distinct accepted responsibilities. The Observer owns the Structures and Emergence surfaces. They track emergent micelles, sheets, droplets, and vesicles through growth, shrinkage, fusion, and rupture; they track the appearance and persistence of autocatalytic systems, compartments, replication, variation, and differential persistence; and they follow persistence over time.
Relevant inputs or decisions. Which structure or emergence event to follow; how long to track persistence; which autocatalytic set or compartment to inspect.
Interactions with other accepted participants. The Observer depends on the Simulation Operator's run for the state they track. Their structural and emergence tracking is distinct from the Chemistry Inspector's molecular verification.
Observable success. A tracked record of emergence events and persistence that confirms the structures arose from the scripted laws rather than from any explicit instruction.
The creative direction is authoritative: Data made physical — an origin-of-life simulation rendered as a living generative artwork, after the muse Refik Anadol. The headline is "SCRIPT THE LAWS. NOT THE LIFE." The data is the aesthetic; every visual element is derived from simulation state.
Mode. Dark mode only.
Palette (dark mode).
| Role | Hex | Use |
|---|---|---|
| Background | #05060A | Deep near-black ground — the primordial void |
| Surface | #0D1018 | Data panels and HUD surfaces |
| Text | #E8E4DC | Warm off-white for readability against the dark ground |
| Primary | #C8A24E | Warm bioluminescent gold-amber — energy gradients, ATP-like high-energy states, the spark of emergent autocatalysis |
| Accent | #3DD6C8 | Cool teal-cyan — aqueous chemistry, vesicles, compartment boundaries |
| Muted | #5A6270 | Secondary data and inactive species |
Data-driven color alphabet (procedural, not fixed UI colors). Molecular classes are assigned hues procedurally so the palette itself evolves with the chemistry: violet #8B5CF6 for nucleotides; coral #E8635A for reactive intermediates; sage #7A9E6B for mineral surfaces; amber #C8A24E for energy; teal #3DD6C8 for amphiphiles.
Typography. Headings: Space Grotesk at 300–500 weight, wide, geometric, slightly technical, with generous tracking (+0.02em) on display sizes and tight leading (0.95–1.05) on large headlines. Uppercase micro-labels in Space Grotesk 500 at 11px with +0.15em tracking for data readouts and section markers. Body: IBM Plex Sans 400 at 15–16px, 1.6 line-height, for its scientific-instrument character and excellent numeral legibility. Numbers and chemical formulas: IBM Plex Mono 400 for tabular alignment.
Type scale. 1.25 modular, fluid: display clamp(48px, 8vw, 128px); h1 clamp(32px, 5vw, 64px); h2 clamp(24px, 3vw, 40px); h3 24px; body 16px; small 13px; micro-label 11px. Display type uses clamp() so the headline scales from 48px on mobile to 128px on desktop without ever cropping or overflowing.
Shape language. Full-bleed generative canvas as the dominant surface — no card chrome, no rounded rectangles floating on gradients. Thin luminous strokes (1px, 40–60% opacity) for data overlays and HUD elements. Circular gauges and radial readouts for concentration dials. Section boundaries are soft horizontal light-washes rather than hard rules. The only rectangles are data panels with 0px radius and 1px hairline borders at 15% opacity — instrument-panel austerity, not SaaS softness.
Spacing rhythm. A 12-column grid with a fixed 64px left-edge vertical rail on desktop (collapsed to a 48px bottom bar on mobile). HUD panels float at the canvas corners. Generous vertical rhythm between editorial sections below the fold.
Imagery style. The simulation's own data rendered as generative art: particle fields colored by molecular class, fluid-dynamic flows showing concentration gradients, luminous filament networks showing reaction pathways, translucent vesicle membranes rendered as thin refractive shells. Mineral surfaces appear as textured planes with subtle topographic displacement. No stock photography, no 3D product renders, no illustration.
Forbidden. Centered headline + subtext + blue button + gradient blob hero; white or near-white background with blue/indigo accent; Inter, Roboto, Arial, Helvetica, Open Sans, Lato, Poppins, or system-ui for headings or body; grid of identical hover-lift cards; stock photography, 3D product renders, or flat clip-art illustration; any UI element that implies scripted life (organism icons, health bars, fitness scores, species labels); rounded-rectangle glassmorphism panels with soft shadows; decorative motion that does not correspond to simulation state.
The public entry is a full-viewport (100vh) generative canvas on a #05060A void. The dominant element is a slow-churning 3D particle field — 40,000 points at 60fps — representing the primordial chemical inventory, colored by molecular class (amber for energy carriers, teal for amphiphiles, violet for nucleotides, coral for reactive intermediates). Particles drift with curl-noise advection, occasionally clustering into transient luminous structures that bloom and dissolve.
Overlaid at the lower-left, spanning 8 of 12 columns, is a single oversized headline in Space Grotesk 300 at clamp(48px, 8vw, 128px), reading "SCRIPT THE LAWS. NOT THE LIFE." in warm off-white (#E8E4DC), with a thin amber rule above it and a one-line subhead in IBM Plex Sans 16px at 70% opacity. At the top-right corner, a minimal HUD shows elapsed simulated time and a live count of emergent structures. At the top-left, a vertical rail of instrument controls. The composition is asymmetric, the canvas bleeds to all edges, and no gradient blob, no centered stack, and no blue button appears anywhere. The CTA is a single text link "Begin the experiment →" in amber at 14px, positioned beneath the headline.
The signature moves are: the full-bleed generative particle canvas as the hero (never a static image); the emergence bloom — when a new autocatalytic set or vesicle forms, a luminous pulse expands from its location and settles into the ambient flow, announcing emergence without any scripted "life" event; the instrument-panel HUD with hairline borders at 15% opacity, tabular numerals in IBM Plex Mono, and a live molecular inventory ticker scrolling along the bottom edge; the scroll-linked scale morph — as the user scrolls from hero into narrative, the camera pulls back from molecular to planetary scale, the particle field resolving into a globe-like distribution; and the data-driven color alphabet assigned procedurally by molecular class, so the palette itself evolves as new species emerge.
Interaction Model: Animated Motion Tempo: cinematic Hero Dimensionality: webgl
Landing Hero Motion Brief. The focal subject is the primordial chemical inventory rendered as a slow-churning 3D particle field of 40,000 points, colored by molecular class. The input→transformation→outcome thesis: ambient curl-noise advection (input) causes particles to cluster into transient luminous structures that bloom and dissolve (transformation), so the viewer sees order appearing from simple laws without any scripted life event (outcome). The motion vocabulary is continuous slow generative flow — the canvas never freezes, even when the simulation is paused, because the visualization breathes at 0.1Hz to signal that the system is alive; particle fields drift with curl-noise advection; when a new autocatalytic set emerges, a brief luminous bloom pulses outward from its location and then settles into the ambient flow; scroll-linked morphing pulls the camera back from molecular to planetary scale as the user scrolls from hero to narrative, the particle field resolving into a globe-like distribution. The composed first frame is the full-viewport canvas on #05060A with the particle field already drifting, the headline at lower-left, the amber rule above it, the subhead beneath, the top-right HUD showing elapsed simulated time and a live emergent-structure count, the top-left instrument rail, and the amber CTA text link beneath the headline. The reduced-motion state falls back to a static frame with a subtle CSS opacity pulse on emerging structures, and all data panels become scrollable lists.
Landing Hero 3D Scene Brief — DIRECTION-DERIVED. One crafted real-time scene: a 40,000-point particle field advected by curl noise on a #05060A void, with points colored by molecular class (amber energy, teal amphiphile, violet nucleotide, coral reactive intermediate, sage mineral). The scene shows the product's defining state — a primordial chemical inventory in continuous motion, occasionally clustering into transient luminous structures that bloom and dissolve — so the viewer sees emergence as a visual fact rather than a claim. The scene is asymmetric, bleeds to all edges, and never freezes.
NFR-1 — No scripted life. The simulation must never contain a command equivalent to spawnLife(), logic equivalent to if molecule == RNA then reproduce, logic equivalent to organism.health += energy, or a manually assigned evolutionary fitness score. (explicit) — Rationale: the central rule "Script the laws. Do not script life."
NFR-2 — No scripted artificial-life constructs. Organisms, metabolism, reproduction, DNA, food, predators, fitness, and species must never be manually scripted. (explicit) — Rationale: the product is explicitly not a traditional artificial-life simulation.
NFR-3 — No organism-like properties by label. No organism-like property may be granted to a structure merely because it has been labeled an organism. (explicit) — Rationale: life is a special organizational state of ordinary matter.
NFR-4 — Extensible chemical grammar. The system must not attempt to enumerate all possible chemicals; it must use an extensible chemical grammar that constructs previously unseen species from valid chemical transformations. (explicit) — Rationale: a literal database of every possible chemical is impossible.
NFR-5 — Water as continuum bulk solvent. Water must be modeled explicitly as the bulk solvent at the continuum level rather than rendering every water molecule. (explicit) — Rationale: computational scalability without losing explicit solvent chemistry.
NFR-6 — Coarse-grained membrane model permitted. Fully atomistic membrane simulation may be replaced with a validated coarse-grained membrane model for computational scalability. (explicit) — Rationale: computational scalability.
NFR-7 — Kinetic networks, not collectible resources. Prebiotic carbon/nitrogen compounds must participate in kinetic reaction networks and must not be treated merely as collectible resources. (explicit) — Rationale: the chemistry must be real, not a resource game.
NFR-8 — No assumed modern protein synthesis. The simulation must not assume modern protein synthesis. (explicit) — Rationale: amino acids are simply molecules capable of reactions including condensation into short peptides.
NFR-9 — Ribose is not artificially immortal. Ribose must degrade realistically. (explicit) — Rationale: instability and degradation are part of the chemistry.
NFR-10 — ATP is not the universal primordial energy currency. ADP and ATP must be possible products of chemistry but must not be supplied as the universal primordial energy currency; primitive systems must initially be capable of using other energetic chemistry. (explicit) — Rationale: primordial energetics must not be assumed modern.
NFR-11 — No scripted cell membrane object. A "cell membrane object" must not be scripted first; a membrane must originate from amphiphile self-assembly. (explicit) — Rationale: self-assembly must be emergent.
NFR-12 — Backend execution for durable state. Durable simulation state, chemical transformations, emergent structures, and longitudinal tracking must be executed on the backend. (required_inference) — Rationale: runs must persist and be resumable across sessions.
NFR-13 — Readable text and controls stay whole. Headlines, wordmarks, labels, numbers, cards' text, and controls must stay entirely inside the viewport and their container at 375px, 768px, and 1280px, wrapping or scaling (for example font-size: clamp(...) with its mobile size) to fit, and no other element may cover any part of them. (explicit) — Rationale: accessibility and legibility across viewports.
NFR-14 — Reduced-motion fallback. All motion must respect prefers-reduced-motion: the hero falls back to a static frame with a subtle CSS opacity pulse on emerging structures, and all data panels become scrollable lists. (explicit) — Rationale: accessibility.
NFR-15 — Data-driven motion only. Every animation must correspond to simulation state; decorative motion that does not correspond to simulation state is forbidden. (explicit) — Rationale: the data is the aesthetic.
Assumptions.
Constraints.
spawnLife().if molecule == RNA then reproduce.organism.health += energy.#05060A) of the visual design.
Begin with chemistry, matter, energy and environmental gradients. Organized systems appear only if the underlying physical and chemical rules make them possible.
Begin the experimentThe Ultimate Experiment
Built for the people who ask it. A laboratory instrument for
It begins with chemistry, matter, energy, and environmental gradients — and nothing else. No outcome is authored ahead of the run. The central rule governs everything downstream: Script the laws. Do not script life. A single instruction is deliberately missing from the source — spawnLife() does not exist.
Causal chain
This is the sequence the simulation is permitted to traverse. Each stage is a consequence of the one before it, computed from the rules rather than issued as an instruction.
The chain proceeds only if the simulated conditions allow those things to occur. No stage is scheduled, guaranteed, or triggered by a named life event.
Matter, energy and environmental gradients set the arena; no outcome beyond these laws is prescribed.
Bond formation and acid/base equilibria proceed from energetics and rates alone.
Coupling between reactions yields networks whose structure was never enumerated in advance.
Local order appears where flux, catalysis and surface chemistry make it favourable.
Amphiphile self-assembly produces boundaries that concentrate their own contents.
Products that catalyse further product formation can propagate when kinetics allow.
Imperfect copying and side reactions generate a spread of descendant compositions.
Whatever persists longest and multiplies fastest dominates the population by physics, not by scoring.
Composition drifts and accumulates across generations; nothing is directed toward a target form.
Physics → chemistry → reaction networks → self-organization → compartments → replication → variation → differential persistence → evolution.
Every stage after the first is written as conditional, because the laws decide.
Initial Inventory
Every species below enters the same kinetic reaction network. None of them is a collectible resource, a label for life, or a privileged ancestor — they are simply matter and energy under the same rules as everything that follows.
13 classes126 species
Water is modelled as the bulk solvent at the continuum level; local pH emerges from acid/base concentrations. Individual water molecules are not rendered.
Concentrations feed ionic strength, membrane stability, polymer folding, precipitation and reaction rates.
Gas exchange between atmosphere and water follows solubility and partial pressure.
Kinetic network participants, not collectible resources.
Additional carboxylic acids may emerge from valid transformations.
Condensation into short peptides only — no modern protein synthesis is assumed.
Instability and degradation are included; ribose is not immortal.
Thymine may arise later but need not dominate primordial chemistry.
ADP and ATP are possible products of chemistry, never supplied as a universal primordial energy currency.
Phosphate adsorption to minerals and the realistic difficulty of phosphorylation are modelled.
Thioester chemistry couples energetically favourable and unfavourable reactions.
Amphiphilic behaviour arises from molecular structure — above threshold concentration these species self-assemble.
Surfaces enter here as inventory only; adsorption, catalysis and phosphate binding are treated in the mineral band.
Starting chemistry, not the universe
This library is the starting chemistry, not the universe of possible chemistry. Species that were never listed are constructed dynamically from valid chemical transformations, each new molecule carrying the same grammar of properties as the ones above.
The mineral world13 surfaces
Every reaction below runs against a mineral interface. Each surface contributes a distinct physical role — adsorption, catalysis, phosphate retention, redox mediation — and the same ledger governs both the dissolved species and the surfaces they touch.
Reduced-iron sulfide surface that binds dissolved metals and hosts early thioester-forming chemistry.
Persistent sulfide mineral whose surface couples electron transfer and drives reductive carbon fixation.
Mixed-valence sulfide offering both electron donors and acceptors at the same interface.
Nickel sulfide surface that adsorbs cyanide species and catalyses their condensation.
Mixed ferrous–ferric oxide that adsorbs organics and mediates redox gradients.
Mineral framework providing high-surface-area scaffolding for concentrating dilute organics.
Hydration product of olivine; its formation releases hydrogen and drives alkaline vent chemistry.
Primary mantle silicate whose aqueous alteration supplies the reducing power of serpentinizing systems.
Alkaline hydroxide layer that buffers local pH and stabilises anionic intermediates.
Precipitating carbonate phases that fix dissolved CO₂ and regulate alkalinity.
Ferric oxide phases that adsorb phosphate and anionic species onto their surfaces.
Swelling clay whose interlayer cations adsorb nucleotides and template polymer condensation.
Phosphate-bearing surfaces that adsorb orthophosphate and make phosphorylation energetically reachable.
Boundary condition
No mineral surface here is granted behaviour the chemistry does not give it — the surfaces are laws, not actors.
Adsorption affinities, catalytic groups and phosphate binding are properties of the mineral, evaluated per collision like every dissolved species. Nothing is placed by hand at the interface.
Above the concentration where packing beats dissolution, amphiphiles assemble the same way the bulk solvent pushes them to — into micelles, sheets, droplets and vesicles. No cell membrane object is scripted first: a membrane is a consequence of molecular structure and local concentration, and it keeps changing for as long as the chemistry around it does.
C8–C18 amphiphile aggregate
Above the critical aggregate concentration, amphiphile tails sequester from water and heads face the bulk solvent, closing into a spherical aggregate with no interior volume.
spherical aggregateplanar bilayer lamella
At lower curvature the same molecules pack into extended lamellae. Sheets stack, slide against one another and expose mineral surfaces to a two-dimensional reaction field.
extended bilayersurfactant-stabilised organic phase
A sparingly soluble organic phase is stabilised by an adsorbed amphiphile monolayer, creating a bounded interior without any pre-declared boundary object.
soft interfaceclosed bilayer compartment
When a sheet closes on itself it encloses a volume of bulk solution. The boundary is a thin refractive shell a few nanometres thick, assembled entirely by the packing of its amphiphiles.
closed aqueous compartmentAmphiphiles exchange from solution into an existing boundary faster than they leave it, so the structure gains area.
When the local amphiphile concentration falls below equilibrium, the reverse flux dominates and the structure loses molecules.
Two boundaries that meet can merge into a single larger structure, combining their enclosed volumes and contents.
A destabilised boundary fails, releases its contents to the bulk solution, and its amphiphiles return to the free pool.
There is no cell membrane object in the model. No compartment is declared, labelled, or given a lifetime by the program — a boundary exists only while amphiphile flux keeps it closed.
A membrane originates from amphiphile self-assembly. Compartments appear only where the simulated concentration, temperature and ionic conditions make closure favourable, and they persist only while that remains true.
Each category below is an outcome the model is permitted to reach, never an instruction it is given. Nothing was explicitly told the program to make. Every marker on this axis is a chemical consequence of the laws set upstream, not a run history. Outcomes can be inspected in the Emergence and Structures surfaces.
Possible outcome
conditional on chemistry
Autocatalytic networks may persist
Persistence · 01
reaction cycles sustained while the gradient holds
Possible outcome
conditional on chemistry
Compartments may self-assemble
Persistence · 02
boundaries persist while amphiphile supply continues
Possible outcome
conditional on chemistry
Templated copying appears
Persistence · 03
copy lineages persist only as long as their chemistry is resupplied
Possible outcome
conditional on chemistry
Imperfect descendants accumulate
Persistence · 04
variant populations persist where their kinetics permit
Possible outcome
conditional on chemistry
Some lineages outlast others
Persistence · 05
persisting lineages outlast non-persisting ones
Nothing was told to appear
There is no spawn command in this model, and no branch that rewards one molecule over another. A structure held together here because its chemistry let it hold together — a marker on the axis is a consequence, not an event trigger.
Where to watch it happen
The same quantities tracked along this axis resolve into the Emergence surface, where autocatalytic sets and compartments can be tracked if they form, and into the Structures surface, which holds the assembled forms and their persistence over simulated time.

Begin with chemistry, matter, energy and environmental gradients. Organized systems appear only if the underlying physical and chemical rules make them possible.
Begin the experimentThe Ultimate Experiment
Built for the people who ask it. A laboratory instrument for
It begins with chemistry, matter, energy, and environmental gradients — and nothing else. No outcome is authored ahead of the run. The central rule governs everything downstream: Script the laws. Do not script life. A single instruction is deliberately missing from the source — spawnLife() does not exist.
Causal chain
This is the sequence the simulation is permitted to traverse. Each stage is a consequence of the one before it, computed from the rules rather than issued as an instruction.
The chain proceeds only if the simulated conditions allow those things to occur. No stage is scheduled, guaranteed, or triggered by a named life event.
Matter, energy and environmental gradients set the arena; no outcome beyond these laws is prescribed.
Bond formation and acid/base equilibria proceed from energetics and rates alone.
Coupling between reactions yields networks whose structure was never enumerated in advance.
Local order appears where flux, catalysis and surface chemistry make it favourable.
Amphiphile self-assembly produces boundaries that concentrate their own contents.
Products that catalyse further product formation can propagate when kinetics allow.
Imperfect copying and side reactions generate a spread of descendant compositions.
Whatever persists longest and multiplies fastest dominates the population by physics, not by scoring.
Composition drifts and accumulates across generations; nothing is directed toward a target form.
Physics → chemistry → reaction networks → self-organization → compartments → replication → variation → differential persistence → evolution.
Every stage after the first is written as conditional, because the laws decide.
Initial Inventory
Every species below enters the same kinetic reaction network. None of them is a collectible resource, a label for life, or a privileged ancestor — they are simply matter and energy under the same rules as everything that follows.
13 classes126 species
Water is modelled as the bulk solvent at the continuum level; local pH emerges from acid/base concentrations. Individual water molecules are not rendered.
Concentrations feed ionic strength, membrane stability, polymer folding, precipitation and reaction rates.
Gas exchange between atmosphere and water follows solubility and partial pressure.
Kinetic network participants, not collectible resources.
Additional carboxylic acids may emerge from valid transformations.
Condensation into short peptides only — no modern protein synthesis is assumed.
Instability and degradation are included; ribose is not immortal.
Thymine may arise later but need not dominate primordial chemistry.
ADP and ATP are possible products of chemistry, never supplied as a universal primordial energy currency.
Phosphate adsorption to minerals and the realistic difficulty of phosphorylation are modelled.
Thioester chemistry couples energetically favourable and unfavourable reactions.
Amphiphilic behaviour arises from molecular structure — above threshold concentration these species self-assemble.
Surfaces enter here as inventory only; adsorption, catalysis and phosphate binding are treated in the mineral band.
Starting chemistry, not the universe
This library is the starting chemistry, not the universe of possible chemistry. Species that were never listed are constructed dynamically from valid chemical transformations, each new molecule carrying the same grammar of properties as the ones above.
The mineral world13 surfaces
Every reaction below runs against a mineral interface. Each surface contributes a distinct physical role — adsorption, catalysis, phosphate retention, redox mediation — and the same ledger governs both the dissolved species and the surfaces they touch.
Reduced-iron sulfide surface that binds dissolved metals and hosts early thioester-forming chemistry.
Persistent sulfide mineral whose surface couples electron transfer and drives reductive carbon fixation.
Mixed-valence sulfide offering both electron donors and acceptors at the same interface.
Nickel sulfide surface that adsorbs cyanide species and catalyses their condensation.
Mixed ferrous–ferric oxide that adsorbs organics and mediates redox gradients.
Mineral framework providing high-surface-area scaffolding for concentrating dilute organics.
Hydration product of olivine; its formation releases hydrogen and drives alkaline vent chemistry.
Primary mantle silicate whose aqueous alteration supplies the reducing power of serpentinizing systems.
Alkaline hydroxide layer that buffers local pH and stabilises anionic intermediates.
Precipitating carbonate phases that fix dissolved CO₂ and regulate alkalinity.
Ferric oxide phases that adsorb phosphate and anionic species onto their surfaces.
Swelling clay whose interlayer cations adsorb nucleotides and template polymer condensation.
Phosphate-bearing surfaces that adsorb orthophosphate and make phosphorylation energetically reachable.
Boundary condition
No mineral surface here is granted behaviour the chemistry does not give it — the surfaces are laws, not actors.
Adsorption affinities, catalytic groups and phosphate binding are properties of the mineral, evaluated per collision like every dissolved species. Nothing is placed by hand at the interface.
Above the concentration where packing beats dissolution, amphiphiles assemble the same way the bulk solvent pushes them to — into micelles, sheets, droplets and vesicles. No cell membrane object is scripted first: a membrane is a consequence of molecular structure and local concentration, and it keeps changing for as long as the chemistry around it does.
C8–C18 amphiphile aggregate
Above the critical aggregate concentration, amphiphile tails sequester from water and heads face the bulk solvent, closing into a spherical aggregate with no interior volume.
spherical aggregateplanar bilayer lamella
At lower curvature the same molecules pack into extended lamellae. Sheets stack, slide against one another and expose mineral surfaces to a two-dimensional reaction field.
extended bilayersurfactant-stabilised organic phase
A sparingly soluble organic phase is stabilised by an adsorbed amphiphile monolayer, creating a bounded interior without any pre-declared boundary object.
soft interfaceclosed bilayer compartment
When a sheet closes on itself it encloses a volume of bulk solution. The boundary is a thin refractive shell a few nanometres thick, assembled entirely by the packing of its amphiphiles.
closed aqueous compartmentAmphiphiles exchange from solution into an existing boundary faster than they leave it, so the structure gains area.
When the local amphiphile concentration falls below equilibrium, the reverse flux dominates and the structure loses molecules.
Two boundaries that meet can merge into a single larger structure, combining their enclosed volumes and contents.
A destabilised boundary fails, releases its contents to the bulk solution, and its amphiphiles return to the free pool.
There is no cell membrane object in the model. No compartment is declared, labelled, or given a lifetime by the program — a boundary exists only while amphiphile flux keeps it closed.
A membrane originates from amphiphile self-assembly. Compartments appear only where the simulated concentration, temperature and ionic conditions make closure favourable, and they persist only while that remains true.
Each category below is an outcome the model is permitted to reach, never an instruction it is given. Nothing was explicitly told the program to make. Every marker on this axis is a chemical consequence of the laws set upstream, not a run history. Outcomes can be inspected in the Emergence and Structures surfaces.
Possible outcome
conditional on chemistry
Autocatalytic networks may persist
Persistence · 01
reaction cycles sustained while the gradient holds
Possible outcome
conditional on chemistry
Compartments may self-assemble
Persistence · 02
boundaries persist while amphiphile supply continues
Possible outcome
conditional on chemistry
Templated copying appears
Persistence · 03
copy lineages persist only as long as their chemistry is resupplied
Possible outcome
conditional on chemistry
Imperfect descendants accumulate
Persistence · 04
variant populations persist where their kinetics permit
Possible outcome
conditional on chemistry
Some lineages outlast others
Persistence · 05
persisting lineages outlast non-persisting ones
Nothing was told to appear
There is no spawn command in this model, and no branch that rewards one molecule over another. A structure held together here because its chemistry let it hold together — a marker on the axis is a consequence, not an event trigger.
Where to watch it happen
The same quantities tracked along this axis resolve into the Emergence surface, where autocatalytic sets and compartments can be tracked if they form, and into the Structures surface, which holds the assembled forms and their persistence over simulated time.
No comments yet. Be the first!