I’ve not talked to a lot of other artists in a while, but my impression is that most of them are either behind the curve, still in the Denial/Anger stage, while others are jumping in with both feet. -Some of the guys who have YouTube channels employ Midjourney for thumbnail art; through that application alone, you know they’ve been on-boarded with the Ai wagon at least to a degree, are open to playing with it. It’s a useful tool. (Though, Midjourney art all looks like Midjourney art, to my eye. I instinctively turn my nose up at it, “Bah. That’s just Ai crap.”)
It’s the Instagram social justice crowd who hatehatehate Ai art. Partly, I think, because they were all soulless surface-shiny artists to begin with, (where Ai excels), making them more obsolete than most…, and partly because Instagram is one of the spawning vats where Junior High cancel pig gossip foments, and this is just the latest, convenient psychopathic crusade.
I’ve done a lot of experimenting with Stable Diffusion, (even bought a 24Gb RTX graphics card off eBay to play with), and it’s seriously cool technology, but not nearly as Get-Rich-Quick as people think. I’m coming to the conclusion that it really is just another productivity tool, albeit a super-charged one.
Though…, there’s another element to it…
I strongly suspect it was also partly created for the purpose of sucking creative soul juice out of people. It takes a HUGE amount of attention energy to get it to do what you want. It makes great stuff, but it can’t read your thoughts and getting it close to what you’re after is almost impossible. You either abandon your vision and go with what it kicks out, (which is essentially the death of the artist), or you fight until you’re blue and eventually have to draw a bunch of it yourself. Woe unto those young kids who feel defeated from the start and never bother learning how to draw as a result. I think that’s a real possibility.
Right now, it turns the otherwise effective artist into a frustrated client of a severely retarded, but technically skilled graphic designer.
But the science is super interesting; the idea is that there are an infinite number of images possible in a field of noise, including all the images you want. And through a sort of backwards infinite probability math, you collapse the wave function into a particle reality. It’s the double slit experiment on steroids. -Plus, where lasers and photon detectors end, with generative Ai, you get a sort of control dial. -That is, you start and stop the process of wave collapse, doing it in stages, nudging it along the way to get the dart in flight to land close-ish to where you want it to go.
If somebody described it to me in a science fiction novel, I’d think, “Ha! That’s such an elegant, fun idea, great for a story, but totally unrealistic. -Like Jules Verne’s gravity blocking plates, (which propelled his character’s trip to the moon).” Except, gosh darn if it doesn’t actually work!
Its most useful application right now is if you draw a rough image with most of what you want done, and present that as the starting point, (the ‘Latent’ image), it’ll clean it up all nice and shiny for you. The more work you put in, the better the result.
The professional artist with years of practice is going to have an edge over the unskilled user. But it’s still work, and still frustrating AF. It’s still in the novelty toy box for me. My real drawing work is still all done by hand. Computers are waaaay too dumb to draw anything even close to what I do in even the simplest comics panel. -But I’m coding a kind of information into my images which is quite beyond the ken of our emerging silicone overlords, and I suspect, always will be.