alt Hacker NewsThis is true for designing, too. I see a lot of people saying Figma is dead because you can prototype directly; for novel UX problems the act of designing — of putting ideas together, assembling a component, even getting good data into a design — is where a lot of my thinking and insights develop. The brain work plus the craft work, the tight loop between thinking and creating is so critical for me and I haven’t found nearly the same feeling when prompting.
In writing and designing I think we often have a thought-feeling first: something inchoate that we can’t quite articulate yet. Getting directly to the medium itself — text or frames and layouts — is how we give shape to that feeling and turn it into a concept, insight or design.
Replies
I actually appreciate working with LLMs on design quite a lot, although I'm not a designer by trade so my situation may differ. I enjoy getting an image-generated mockup (that is usually top class), then letting an LLM implement it, then seeing where it doesn't feel quite right. Then trying to understand why it doesn't feel right. When I do the CSS myself, I iterate one component at a time and usually not get the 'find the 100 differences' feel. Animations are even more fun, you feel they suck, but you have to slow them down to understand what's really making them suck. Not claiming I can do good designs, but it helped me identify faster why some designs were bad.
I see what you are saying and for very specific problems/explorations I do still reach for Figma. But the days of crafting static screens and multi-artboard flows are gone. It's better to prototype the entire flow/interaction directly than trying to reason about it 'statically'. LLMs have also opened up an entire array of new possibilities for e.g. combining meaningful 3D and taking a 'leap of faith' into visualizations/interactions that would simply be too costly to properly explore before