Latest Substance Designer experiments. Probabilistic Cellular Automata Ornaments

I spent some time exploring Wolfram-inspired ornaments — small pixel-like artifacts that feel somewhere between cellular automata, jewelry, embroidery, and digital fossils.

The starting point was a simple probabilistic cellular automata idea mentioned in Stephen Wolfram’s article on biological evolution and adaptive processes. The rule is minimal: a cell can become alive if it or one of its immediate neighbors was alive in the previous step, but the activation is controlled by a probability parameter. As this probability increases, the system reaches a sudden transition point — around 0.56 in his example — after which the space tends to become almost fully occupied.

I used this idea as a small procedural ornament generator.

Instead of using a single global probability value, I used a white noise texture as a probability lookup table. Each cell checks the local CA condition, then consults the probability field to decide whether it is allowed to become active. So the system is driven by two forces:

local growth logic
+
a probability landscape

I also “cheated” a little in a useful way: I mirrored the noise texture and tiled it vertically a few times. This bakes symmetry and cyclic behavior into the environment as an external forcing. The CA itself remains simple, but the probability field becomes structured — almost like a developmental field that guides where growth is allowed to happen.

Wolfram’s article also explores adaptive searches for programs that extend the lifetime of an evolution, including cyclic patterns that grow, shrink, and repeat. I liked that idea a lot, but instead of searching only for internally cyclic rules, I experimented with externally imposed cyclicity through the probability field.

The result is a family of finite, symmetric, ornament-like forms. Some look like pendants, insects, embroidered motifs, little shrines, glyphs, or pixel jewelry. I generated around 10 atlases so far, each containing 16 to 25 ornaments produced by this process.

What interests me most is not only the final image, but the mechanism:

ornamental artifacts

There is something beautiful in the tension between randomness and constraint. Pure randomness would dissolve into noise. Pure symmetry would become sterile. But structured randomness — especially near a phase transition — can produce forms that feel alive, delicate, and strangely intentional.

This is part of a broader exploration of procedural systems where composition happens before rendering: rules, probabilities, constraints, and fields are arranged in data space first, and only later become visible as images.