As I noted here:
Another article on yet another experiment to mimic human creativity in paint. Say it ain't so, bunky! robot painter Meantime, I visited a friend's home and viewed his collection of paintings. He asked if I recognized the artist in one, a red tree done simply and symmetrically. While I made a...
creativespark.art
human art has actually always involved a melding of the human mind with some or other form of technology. At least for the moment, the same is true of AI art: having played around with the tech a bit I have concluded that its uses are rather limited, and consists mainly of giving one ideas rather than actually creating completed works of art.
E.g. suppose I wanted to create a picture of a hybrid of a cat and Yoda. Perchance can do it:
But note how awkward the thing is: the paws look all wrong, where exactly does the tail begin or end, and what the heck is going on in the background?
But the face is not bad. One could potentially use this picture as a reference.
AI cannot make magic; I had to generate LOTS of cat-Yoda hybrid before I got this one; the others were all useless. I suspect this is how a lot of those "scary good" AI pictures are done: we see only the rare good ones, and the bad ones just get discarded. In the thread I mention above, go check out how hopeless AI was in generating Gary Larson-like cartoons (and Larson is somewhat notorious for bad draftsmanship!)
Now I hasten to add that I used the free online version of this software. Apparently the paid ware is more sophisticated, and of course, the tech will improve in time. But at least for the moment, we have little to fear from AI, and can perhaps even make good use of it.
In any even, as I also mention in that thread, the genie is out of the bottle. AI art is going to become more and more of a thing, whether we like it or not.