Because video generation is in its infancy and lacks object permanence. Especially pre-Sora models, but the Sora videos also struggle a lot with it. Unlike the current video generation models, human artists know how to turn individual frames into a coherent sequence. Actually, human animation also can’t reproduce the previous frame perfectly either, but except for oil painting animation, these inconsistencies are very hard to see. Anyway, the result is that all the “AI” videos are psychedelic.
It looks useful for work, but that’s about it, and people keep telling me that it’s “plagiarism” or that it’s bad and nobody should, I guess, use it all?
I see it as a means of the mass proletarianization and unemployment of artists which in its current state will also lead to a decline in the quality of art. I’m a communist - my solution is guaranteed public funding to art collectives.
Why does all AI art look the same?
You said it.
Because video generation is in its infancy and lacks object permanence. Especially pre-Sora models, but the Sora videos also struggle a lot with it. Unlike the current video generation models, human artists know how to turn individual frames into a coherent sequence. Actually, human animation also can’t reproduce the previous frame perfectly either, but except for oil painting animation, these inconsistencies are very hard to see. Anyway, the result is that all the “AI” videos are psychedelic.
It looks like throwup.
No offense or anything.
I agree that human artists intentionally making something that looks psychedelic almost always looks better.
Yeah
I’m just not sure how I feel about AI in general.
It looks useful for work, but that’s about it, and people keep telling me that it’s “plagiarism” or that it’s bad and nobody should, I guess, use it all?
Idk.
I see it as a means of the mass proletarianization and unemployment of artists which in its current state will also lead to a decline in the quality of art. I’m a communist - my solution is guaranteed public funding to art collectives.