AI writes code. AI doesn’t design. Not really.
I keep turning this thought over in my head as I watch another perfectly functional, utterly soulless interface generated by the latest AI tool. Something’s missing. Always missing.
Coding is logic, design is intuition. One follows rules. The other breaks them deliberately, knowingly, with purpose.
The code an AI generates works. It compiles. It functions. But the design it creates is… something else. A hollow echo of what design actually is.
Design isn’t just arrangement. It’s not just color theory and spacing and typography, though it’s all of those things too. Design is feeling. It’s the gut reaction you can’t quite articulate when something just feels right.
AI has no gut to react with.
I’m fascinated by this boundary. This hard limit. The line between the calculable and the intuitive.
When we code, we’re building systems of logic. When we design, we’re crafting emotional responses. One is about correctness. The other about resonance.
There’s a reason why design movements exist. Why manifestos get written. Why styles evolve and change and react against what came before. Design is conversation. A dialogue between creators and culture that spans generations.
AI can only simulate this dialogue. It can’t participate in it. It can mimic the outputs but misses the entire undercurrent of meaning.
Look at brutalism. Cold concrete, exposed structure, raw materials. A reaction against ornament, against hiding things. A philosophy as much as an aesthetic. Can AI understand the post-war context? The rejection of artifice? The political statements embedded in those stark forms?
AI can generate something that looks brutalist. But it doesn’t know why brutalism matters.
I’ve been thinking about Japanese design principles lately. Wabi-sabi embraces imperfection. Values the crack in the ceramic, the worn edge, the asymmetry. How does an optimization algorithm understand the beauty of imperfection when its entire existence is devoted to minimizing error?
The answers aren’t straightforward.
I’m not saying AI can’t be useful for design. It can be. For exploration, for iteration, for breaking through creative blocks. But there’s always that gap. That missing intuition.
AI approaches design problems like technical puzzles to be solved. Not cultural artifacts to be felt. It can follow patterns, but it doesn’t know when to break them. When breaking them matters more than following them.
The most interesting designs exist at that breaking point. The moment when convention gets challenged. When the familiar becomes strange again. These moments require context—cultural, historical, emotional context—that extends beyond data patterns.
Design requires taste. Not just in the sense of aesthetic preference, but in the sense of discrimination, of knowing when something works and when it doesn’t. Why it works. Why it matters.
AI struggles because design isn’t just about the object being designed. It’s about everything around it. The world it exists in. The people who will use it. The cultural currents flowing through both.
Code can be correct in isolation. Design can’t.
I keep coming back to this thought: coding is about solving problems within constraints. Design is about choosing which constraints matter.
AI can work within constraints. It can’t meaningfully choose them.
Maybe someday this will change. Maybe someday AI will develop something like intuition, like taste, like cultural awareness that isn’t just statistical pattern matching.
But for now, there’s this gap. This space where human intuition still matters fundamentally. Where the illogical, emotional, cultural dimensions of design remain stubbornly beyond algorithmic reach.
And I find that strangely comforting.