I mean for sure. I’m not going to make any predictions about the future when the field is so young. It’s only related to the current generation. Honestly though, art isn’t the field to look at for seeing when the jump is coming. It’s in AI actually being able to recognize relationships between things. Basic stuff like looking at enough pictures of horses and recognizing what the leg is and how many legs a normal horse will have. Even stable diffusion which has some of the best generation will still give me 5 legs or two legs coming out of the same side. These kinds of images are a boon to artists who will need to do all the final corrections.
Relationships between things is complicated. Someone resting their face on a fence is going to have a huge number of effects on the deformations of the face, especially the eyes and hair depending on how they are resting. It’s not enough for AI to have seen enough pictures of faces on fences to be able to apply that in an image. It needs to understand what pressure and gravity is doing to the underlying structures. That’s how human artists study at least. It’s why they can take that lesson and apply it to learnings about how skin behaves depending on the age of a person. They aren’t copying. They are solving problems by thinking about muscles underneath and how they change depending on any number of factors.
None of this even touches on lighting and colours.
If there’s one prediction of the future I’m willing to make, it’s that until research progresses on teaching computers to apply actual knowledge, AI within the creative space will remain assistive instead of replacing.