It doesn’t matter how much you do yourself. You could draw the whole thing by hand – but once you start using Stable Diffusion or whatever, the program overwhelms the actual art. You invariably end up with generic popslop, and I frankly don’t know why you bothered to do any original work at all.
God, art historians are going to end up hating this decade.
Moe Lane
PS: My books’ cover art is AI-free. I won’t use an artist who uses it.
#commissionearned
Art historians are vehement that a urinal is high art, and that Thomas Kinkade was a hack.
Their judgement is beyond questionable.
Yeah, I’ve noticed that the AI-generated art has a very “samey” look to it. As for your prognostication of this fad lasting a decade, I’ll take the under on that. Unless the AI gets a lot better, it might not last 3 years.
On the plus side, ChatGPT seems to now know how many r’s there are in strawberry, so it’s got that going for it.
A random observation I have made:
You could take all the effort of feeding in prompts to the AI, correcting the image and fine tuning it and apply that instead to drawing lessons and arguably get a better product.