I have a client, yep actual work, and they have blogs with lots of photos because they hold events, but they don't really write about it.

How can we solve that?

Invert blogging, normally we add photos and describe them, but as you know AI can describe what is going on in a photo.

What if each photo you uploaded photo creates a fragment.

two young and old people stand by a 1900 green tractor

I suppose It's a little like alt-text idea but supercharged.

The sum of all the photos should be able to describe the blog post, then its a matter of giving a little prompted context and boom we have the full prompt, finally a proof read as the "editor" of the article and we are done here.

The UI Concept

Sounds like a fun experiment?