This assumes that models won't improve and you'll always be able to tell that it's "AI slop" ... that seems like a bad bet. Five years ago you'd be laughed out of the room for suggesting that a computer could produce images from a natural language prompt and that it'd be accessible to everyone -- not just corporations with deep pockets.

Yeah if/when it becomes indistinguishable I think most people won’t care. That being said I do think someone finding out something is AI generated will be met with poor response. Does that ultimately matter? Probably not in a business world.