Generative design

There's a moment in generative design I love: I write a rule, let it run — and see something I'd never have drawn by hand. Not because the machine is smarter, but because I gave it permission to surprise me. Working generatively, for me, means not drawing the image but drawing the system that draws images.

Rules instead of brushstrokes

Here I'm not designing a single outcome but a space of possibilities. A few lines of code describe how forms grow, how colors distribute, how strokes overlap. The real craft lives in phrasing those rules — precise enough that a result holds together, open enough that every run stays its own.

Randomness is not chaos here but a measured collaborator. A little noise on a position, a chance pick from a curated palette, a variation in stroke weight. Too much randomness turns to mush; too little turns mechanical. The art lies in the span between — where a result feels alive without feeling arbitrary.

I don't design the image. I design the conditions under which beautiful images can appear.

Curating the randomness

Building a system is one half. The other is ruthless selection. I let hundreds of variants run and look at them the way a photographer reads contact sheets — most I discard, a precious few hit exactly the right nerve. The judgment stays human. The generator proposes; I decide what becomes art and what becomes scrap.

For my clients this means something very concrete: an identity that isn't a rigid logo but a living system. A pattern that looks different on every card yet always carries the same handwriting. A hero image that reinvents itself on every load — unmistakable because it's infinite, and unified because it was born from the same rules.

Generative design is, to me, the most beautiful form of collaboration with the machine. It asks me to let go without surrendering control — to translate intent into rules and then allow myself to marvel at what comes back.