I’ve complained LLMs are bad at SwiftUI, and that’s not entirely accurate.
It has bad default approaches, because it’s all trained stale StackOverflow answers. It can be steered in the right direction if you know what you’re doing, but that’s kind of the problem.
I’ve said it before and I’ll say it again: Apple needs to massively improve its documentation and sample code to overcome this. Thousands of sample projects, actively maintained.
(Hi WWDR! Hire me to do this!)
