Comment by nacozarina
1 day ago
More guidance is needed on how to efficiently prompt an LLM to generate a correct UI. Describing the appearance, placement and behavior of UI elements for this is quite time-consuming. The contrast with server-side code generation is stark.
I'm curious when the music will stop and people realize that programming languages are more optimal at describing certain behaviors than english.
UI design (not implementation) is an area where “describe it with a programming language” has historically not worked. So much of what makes a UI good or bad is determined reactively after seeing the initial prototype. Those refinements are rarely based on an encoded set of principles that found their way into a training set, and the code used to implement such refinements was varied enough that models and humans both struggle to infer its intentions after the fact.
This started as soon as they devised the MCP. “we have this new thing called ai you can talk to it naturally. Here’s a protocol to make it work better”