CHRIS HAY

IDEAS · SYSTEMS · OBJECTS / LONDON · 2026

What changes an evolving world?

Replay a birth, follow its descendants, and test which inherited changes matter.

Browse the reading order ↓

Recorded replay / EX-4

One birth, with and without its program change.

Two recorded histories share the same clock. Load the replay to follow the intervention. This plays saved states; it does not run a new simulation.

Read the experiment ↗
EX-4 / SEED 1 / 32 × 32RECORDED WORLD STATES
980

Before the birth. One shared history.

Observed history

Recorded grid: 32 by 32, 118 living organisms at tick 980. Use the recorded-state download for the complete data.

118alive

0.0%program 33

02231’s family

2,080 births · 1,970 deaths / cumulative

Program change undone

Recorded grid: 32 by 32, 118 living organisms at tick 980. Use the recorded-state download for the complete data.

118alive

0.0%program 33

02231’s family

2,080 births · 1,970 deaths / cumulative

Program 33Program 37Other programsFood

The preview is a recorded frame. Play loads 1.6 MB of history.

Select a square to inspect its organisms. Arrow keys work when a world has focus.

Both worlds use the same clock. This replay reproduces the recorded birth and change in the leading program. Frames are sampled every two ticks, and every tick from 990–1,200.

Squares are actual locations. Larger marks mean more organisms share that square. In the paired view, colour shows program 33 first, then 37; select a square to inspect all its programs. Positions are recorded, with no movement filled in between frames.

Download recorded states ↓ · Replay provenance ↗

Reading order

  1. 01

    Can you name the mutation that changed a world?

    One birth explains a population shift. A replay can isolate a cause. Next, give evolution a world that keeps going.

    WORKING DRAFT · SUPPORTED
  2. 02

    What keeps an evolving world alive?

    11 of 12 worlds kept both species alive; no detected arms race. Survival gives evolution time. It leaves open what evolution can build.

    WORKING DRAFT · PARTIALLY SUPPORTED
  3. 03

    When does improvement become invention?

    Eight helpful changes; neither later sequence passed dependence. The movement test found no reliable dependence. A new food-processing world gives a capability something to unlock.

    WORKING DRAFT · PARTIALLY SUPPORTED
  4. 04

    Give invention something to unlock.

    One combination crossed the resource barrier; the later test had no eligible case. A capability appeared. Next, ask whether a useful change gets the chance to reach descendants.

    WORKING DRAFT · PARTIALLY SUPPORTED
  5. 05

    An advantage needs a chance to become history.

    Both changes reached descendants. One family was eventually lost. Inheritance and conditional benefit were observed. The closing note asks how much of the opportunity the experimenter designed.

    WORKING DRAFT · PARTIALLY SUPPORTED
  6. 06

    Can something evolve that makes the next invention possible?

    The controls explained the apparent expansion in these worlds. The next question is what a world would need for usefulness to emerge from its own regularities. That programme has not been registered.

    WORKING DRAFT · BOUND
About this programme

Six connected notes from a small evolutionary world: replay a birth, keep an ecology alive, test an improvement, follow an inheritance, and ask what survives the controls.

Start with any question, or follow the whole experiment. The first three notes study movement and ecology; the next two use a new food-processing world. The closing note tests what the design itself made possible. These are different assays, not one continuous simulation.

What survives the controls?

These six notes use different worlds and assays. Their controls distinguish a useful change from the opportunities supplied by the design.