I'll throw in my two cents. I see AI generated art (any kind: pictures, music, narrative text) as a problem because it breaks one of the fundamental rules of artistic creation: not everyone can do it. It devalues skill and effort. Art has value because it is not a necessity, it's vanity (of the good kind!), and it's not quantitative, it's qualitative. Also, it's born out of context and environment. When you strip all these factors, all you obtain a homogeneous blob.
That, my friend, is called "gate keeping". It is something the artistic community has done for a long time.
Indeed, before, "not everyone could do it". Artistic pursuits historically have been a luxury that required time, privilege, and patronage to progress.
Artists, if they show any aptitude for it, are told from the time they are children that they are "special". I've long said that artists are the children who never stopped drawing, because all children draw. Most stop. But not artists. But regardless, artists are used to being told they have a special talent, to hearing the awed phrase from people - "That's amazing! I could NEVER do that!"
Artists have always enjoyed being part of a special "gated group of specials" and they've even historically tried to prevent new people from entering that group. Look at the history of salons and their rejection of innovation and creativity. Or the modern art scene, where it is more WHO you know, and not what you do.
They are upset that AI technology has essentially torn down the wall around their garden and is allowing the stinking masses to enjoy the roses. "I had to PROVE myself to get in here!"
But don't worry, artists will come up with new walls and new gatekeeping methods. They always do.
(Source: I'm an artist - an actual, "oh, my god, he paid for art school" artist with a degree in the subject and years spent drawing, painting, and sculpting in a collective group with other artists. I just never drank the "we're special snowflakes" kool-aid.)
Also, I know that current AI algorithms are "not good enough" yet. But there is enough incentive to improve them, and I'm sure they will improve. And that's even worse, because being able to generate an infinite number of masterpieces (for an arbitrary definition of one) in a few minutes deprives them of all their uniqueness and value.
Yep. Economics.
But I'll say that if AI produces "masterpieces" that can't be distinguished from human generated "masterpieces", it's fair. Artists will just have to adapt, like they did with the invention of the camera. Some artists freaked out, others recognized it as a tool to increase their productivity and improve their work flow. There even developed a new class of artists - photographers, who understood and excelled at pushing the new medium.
I also know that AI algorithms are simply tools to an end: if you use them to create images and you keep them for yourself, good for you. If an artist uses them to improve himself, good for him. But there is no way to draw a line on their use, and ensure a limit on the distribution of the content they generate, so this point is basically moot.
Some things to remember here:
It is preached CONSTANTLY that being an artist just means being a creative person who is compelled to create convey some sort of message. That's why weirdos who nail bananas to walls, strew trash on the floor, or masturbate on plexiglass in front of a crowd (all real art museum pieces that earned the artists tens of thousands of dollars), all get to claim the title.
"Normal" people are going to use AI art a few times as a novel gag, and move on, except maybe when they want a picture for their Sunday church bulletin or annual BBQ invite. But there will be a group of these "normal" people, who will use the AI and BE artists. Because creating will consume them. They'll constantly try for better results. They'll sit at work and ponder what image to generate when they get home. They'll work out how to use the AI to get closer and closer to that thing in their head.
They'll BECOME artists. Because AI art still takes intentionality. Everyone with real curiosity on this and not an instant knee-jerk reaction, should look at the Stable Diffusion sub-reddit. Look for posts with workflow included, or look up Youtube videos of people showing their AI workflow. Because you'll find, for ARTISTS (whether they had that title before AI came along or not), it is indeed a WORKflow.
There are tools now for them to create poses, define facial expressions, which way the head is turned. Composition - what the AI generates and where it generates it (segmentation) - put a chair here, a picture here, a lamp here, a patch of grass here, etc. They have to determine style by the model they select, the LORAs they add on, the prompt words they use and in what order. They don't just accept what the AI gives them. They "inpaint" it. They point at little areas and say "AI, redo this area, but with these instructions". They may do that MANY times. Then many will take that picture into an image editor and hand edit it - take out this and that, move that, fix this and that. Then they'll upscale it, fix it some more, and then post it.
And you know what? Even if all some of them are doing is making pictures of beautiful naked women . . . so what? There have been thousands of traditional artists who do nothing but paint or draw pictures of beautiful naked women over and over again, and no one accuses them of not being artists due to their "lack of creativity". But there are AI artists who are doing much more creative things - a series of selfies with historical figures, each image taking them a whole day to produce.
And traditional artists' arguments against AI will continue to fall away. Plagiarism? New AI models are being trained entirely on opensource art and pictures, or from images the company owns all the rights to. See Adobe's Firefly, which will be in Photoshop and used by Google's Bard AI for image generation. No copyright infringement there.
Though I still believe that what the AI does is STYLE COPYING, which is NOT copyright-able. Artists that have videos of them collecting "reference" using PureRef for their digital drawings, and will draw Spider-Man in the style of The Simpsons, but will unironically accuse AI artists of "stealing" from "real" artists by having the AI generate an image of a Disney Princess in the style of Norman Rockwell.
If artists succeed in some way (they won't) of getting copyright law changed to protect STYLES, they will be fucking themselves.
To tie all this back around - artists are people who care about the images they create, and are creating them with a message. That includes AI artists.
Does a dev that posts an adult game here that uses terrible AI art deserve some derision? Sure. But no more than a dev that posts an adult game here that uses terrible default DAZ models with terrible lighting and renders made by screenshotting the application.
A good image is going to be a good image - whether it was made with a camera, a pencil, a brush, or an AI. We haven't even wrapped around to a full year since the introduction of this tech to the public. And the difference in the quality of the AI images in just 8 months is ASTOUNDING.
So for anyone being derisive of the tech, I want you to think about what that means for where it will be in 5 years or a decade. I will tell you want will happen - there will be a new generation of artists that grow up with this tech, who are using it now as kids, who will use it make images expressing their angsts and desires when they are teens, and who will be producing outstanding art with it when they become young adults, all without the hang-ups towards the tech the previous generation had. It's what happened with digital art.
And in 5 years (or less) I predict this forum won't give a shit about AI art being in games. The forum will only care if the art is good. "Oh, shit. Finally a game with good HoneySelect/DAZ/AI art!"
(And finally, I find it humorous that so many on a supposed piracy website give a shit about AI art possibly infringing on copyright by copying the "style" of an artist.)