design at the speed of code

lately i’ve been making a lot of prototypes. not the prototypes we used to make, by tediously connecting wires between static screens. i've been building in code, and the speed of execution is incredible.

i’m not pushing pixels anymore. i describe what i want, and the agent builds. i’ve stopped touching the work, and it feels faster than before. but different than before.

when you move fast, you lose breadth. the exploration feels thin. it feels like i’m walking a single path instead of wandering through the forest. i’m building and exploring at the same time, and i’m not sure if that’s the goal or the problem.

what i've always liked about design is how it can feel so loose early on. unserious. uncommitted. you're finding direction, following threads of imagination. everything is possible because nothing is decided.

code, however, doesn’t feel unserious. it feels like building. and when you’re building, you stop asking if you should be building something else. you fall into premature convergence, polishing a single direction instead of branching out.

i think this is the price to pay for moving the exploration closer to the real material. in Figma, you have permission to be wrong in twenty interesting ways, before you’re right. you have the infinite canvas. a boundless world to create every variation of every idea.

but Figma is a picture of the product. a static, stateless vector representation of an interface. you’re making decisions on a fictional material. and you pay for it later, on the other side. when design meets code, you get a translation error.

code is the material. no translation necessary. what you design is what you build. but the material is rigid. less malleable. and it ends up compressing your imagination.

you get speed. but speed isn't velocity. velocity needs direction. and you find direction by wandering, not by committing.

so can you really design at the speed of code? or does code move too fast for design to happen?