DREWSKY / LIFE SYSTEM 01 A small computer you can understand
Drewsky / field lab

EARLY COMPUTER SPIRIT / EXPLICIT STATE / ROOM TO PLAY

Small rules.
Strange worlds.

A little of the 1980s home-computer spirit: visible state, a tiny rulebook, one instruction at a time. Conway’s Life arrived earlier, in Martin Gardner’s 1970 column. This finite version makes its edges visible too.

CONWAY / B3·S2332 × 20 CELLS · DEAD BOUNDARIES
GENERATION
0000
LIVE CELLS
0008
STATE
PAUSED

■ Live · □ Dead · No wrapping across edges

Read the grid as text

Each row runs left to right. # is live; . is dead. The first row is the top edge.

................................
................................
................................
................................
......#.........................
.......#........................
.....###........................
................................
................................
................................
......................###.......
................................
................................
................................
................................
................................
................................
................................
................................
................................

Paused. Choose a pattern, predict, then step once.

Motion begins only with Start. Step advances exactly one generation.

At most 5 generations/second. Stops at 500; Reset starts over. Leaving the tab pauses it. Returning never resumes it automatically.

01 / READCurrent grid
02 / COUNTEight neighbors per cell
03 / UPDATEAll cells, together

01 / THE WHOLE RULEBOOK

No hidden intelligence.

A dead cell is born with exactly three live neighbors. A live cell survives with two or three; otherwise it dies. Every cell updates simultaneously. Outside this rectangle, all cells stay dead: a deliberate boundary, unlike Life’s usual infinite plane.

A glider shifts through repeated reconstruction. A blinker repeats. Neither pattern alone demonstrates biological life, intelligence, or self-replication.

Source: Gardner’s original report (1970, PDF).

02 / NEXT-GENERATION LEARNING

Predict before you press.

Choose Blinker. Sketch its next state, then Step twice. Choose Glider and predict its position after four steps. What changes near a dead edge? Explain a wrong prediction before asking an LLM for an explanation. Compare explanations against the visible rules. The learning exercise is to make a model testable, then revise it.

03 / BIOMIMICRY BEYOND EARTH

Astrochicken, as a question.

In Infinite in All Directions (1988, chapter 10), Freeman Dyson imagined a small space probe combining biological organization, electronics, artificial intelligence, and solar-electric propulsion. He explicitly presented this as speculation and science fiction, not a built device or demonstrated engineering.

For me, that invites questions about self-propagating systems, rather than answers: where would energy and materials come from? How would errors be contained? Who could stop propagation? This grid is a thinking aid, not a space probe or proof of those possibilities.

Primary work: Dyson’s book bibliography; see chapter 10 of the 1988 book.

04 / BACK TO THE KITCHEN

Make the inputs matter.

The bridge to Drewsky’s restaurant work is the habit of exposing inputs, rules, and consequences. Real tools need verified ingredient costs, usable yields, portions, waste, staffing, and service constraints. Conway’s rules do not optimize a restaurant.

A practical next exercise: take an invented recipe, predict how a changed saleable yield affects ingredient cost per portion, then check against hand arithmetic in the kitchen tool. Keep the assumptions visible before involving real records. A compelling simulation earns attention; a useful operating tool still has to earn trust through measured results.

Return to the field lab ↗ · Follow the wider journey ↗