alt Hacker NewsIt is because it is an image on a screen.
Water Lilies are enormous paintings. They are breathtaking in person because of their scale. Monet wouldn't be Monet if he had only produced images on a screen.
Art is good or it is shit, based on personal taste. Just like food, no one can tell me what food tastes good or tastes bad.
AI Art seems to produce strong emotions in people who don't go to art galleries. I love modern art, I am a huge art snob but if you want to see slop, go to any modern art gallery. Personally, I would say for my taste, at least 70% of all art at any gallery is basically shit.
Like food also, the presentation matters. To believe there is no possible way to print out a 6 foot tall by 10 foot long AI generated image that would look awesome hanging in a gallery is stupid.
Agreed except with
> To believe there is no possible way to print out a 6 foot tall by 10 foot long AI generated image that would look awesome hanging in a gallery is stupid.
Maybe if you have a really impressive pipeline where the AI does sketches, 3D models (assuming a 3D theme), keeps track of the lighting, brainstorms and reasons about ideas, keeps track of the world building (like for a cyberpunk painting),...
Otherwise AI will just do AI things. You zoom in and see there are no strokes, no ideas behind the shapes, the shapes don't even connect well or wobble in weird ways, weird distortions that wouldn't happen if a human put down strokes to communicate the idea of an object. You zoom out and the composition is weirdly centered, almost like a logo, there is too much shading dedicated to make the objects pop, everything is weirdly inoffensive, smooth, and too consistent, not enough material/texture distinction on a per object basis.
Not sure how else to explain it - you're the art snob anyway. Mostly focusing on 3D with my criticisms, but that's most of the AI art people use. Maybe it can do abstract better. Impressionism is somewhere inbetween, again, the Monet thing makes some sense to me. Then again, you could write a kick-ass procedural art generator running on a single core CPU, which would also be capable of couch or even gallery worthy outputs. So I don't want to argue the point too hard. Gallery art and product website images/thumbnails are too different in category - arguments for/against one don't translate to the other.