It would be super cool to have. like, maybe a gallery of three to 5 pictures in the codex for reference art. So maybe you have three images of the location or three images of a character, etc. It would be cool if the AI could interpret these photos, but if not whatever.