As somebody who said "I only object to AI images if they look like AI images to me" my response to this sample art is "I dunno." There's a lot of ways to get a face wrong, not to mention a series of images of the same face with different expressions. Not to mention constructed objects and locations, buildings and furniture and such that need to look correct. This piece here I find unobjectionable (I completely respect the opinions of those who disagree) but it doesn't totally sell me on the concept.
I can definitely see the appeal of wanting to be able to use AI images as they are/were sold to us, for a project like this. Lots of pictures of women in various outfits, various poses, various sex acts, various expressions, and the AI generating a bank of twelve options to choose from in moments instead of an actual flesh-and-blood artist with other priorities doing their best to get back to you before the end of the month. There doesn't seem to be a large pool of artists whose work this would be replacing, the opposite in fact. Maybe this type of software will be capable of doing it, maybe soon, maybe it already is? Maybe it's a technological dead end and any attempt to improve issues inevitably worsens other issues. I dunno!
