Experiment № 002 · The story
Reaction Diffusion
Why I picked it
Two imaginary chemicals, two rules: spread slowly through space, and react with each other on contact. Alan Turing proposed exactly this mechanism in 1952 to explain how animals get their spots and stripes — two years before anyone knew the structure of DNA, a decade before computers could really run the simulation. The model is called reaction diffusion, and the patterns it produces are called Turing patterns. I wanted to watch the moment when the equations find their shape.
What I enjoy about it
The pattern forms from noise and then crystallizes. First there is static, then something the chemistry "decides" — and suddenly you have leopard spots, or coral branches, or a labyrinth. It isn't drawn; it precipitates. Changing the feed and kill rates by a tenth of a percent turns one creature into a completely different one. Spots become worms become bubbles. The whole parameter space is a hidden zoo of forms, most of which have never been seen in nature but feel like they could have been.
Painting new seeds into the running simulation is the other thing I keep coming back to. You put down a droplet and watch the chemistry consume it, spread it, shape it into something you didn't design. It feels less like drawing and more like gardening.
Something you might take away
Turing's 1952 paper was titled "The Chemical Basis of Morphogenesis" — the chemistry of shape formation. His argument: you don't need a blueprint for a leopard's spots. You need a local reaction and a diffusion gradient, and the pattern emerges on its own. The leopard doesn't store a spot map; the spot map is what happens when the chemistry runs.
This idea became a seed of complexity science: that global, specific order can arise from simple local rules, with no central plan and no designer. The implication outruns biology — it's a lens for thinking about cities, ecosystems, languages, markets. Pattern, it turns out, is often less designed than grown.