honestly it's fun to mess around with to see what madness comes out (my dad has Stable Diffusion and we once used it and plan to have fun with it again when we have time, however we mostly just generate cursed look cars and planes, maybe some planets and cityscapes) but aside from that, yeah no
we all know the blursed hands thing and whatnot, and others have pointed out the AI "ArTiStS" thing already, however i still think it's very easy to detect AI art: most notably, lighting will look off, idk why but almost all AI generated pictures, even the "photorealistic" ones, seem to have excessive white lighting shine everywhere and kind of a TADC-ish look (idk how to describe it well) tbh, second most notable is the background, chances are it will deformed or just barely revognizable people and figures, the background will sometimes be excessively blurry, etc.
for now it's still easy to detect AI generated imagery but it is rapidly advancing (from what i've heard and seen the hand thing is actually mostly fixed now, still has issues with text a lot of the time though) so it won't be very long until the programmers figure out what training data to use and what configurations to fiddle around with and most of its problems will be solved and then really only the truly keen-eyed will be able to spot the differences between real and generated
