I think it can be easier to give a napkin sketch / whiteboard image / wireframe to get to the right UI faster, than having to verbally give very specific directions for every little element.

I was wondering the submit pic option on GPT-4 actually would be this well and perhaps out HTML/CSS to do it - maybe not possible today, haven't tried it yet.

Or who knows - maybe it will be be a like a senior product design who will just figure it all out, but that would require a long convo / submitting a design brief / customer profiles, etc.