Well I’m sick of being effectively silent on this issue.
For starters, I call it diffusion, not AI. Calling it AI mystifies the situation. Makes it seem uncountable by implying the machine “thinks” on its own. Machines don’t screw with the artistic eco system, the people who use/make them do. Secondly, I don’t want to pretend that we don’t live in a post-Adobe-and-its-free-alternatives world. Even considering that we don’t live in a post-scarcity utopia, I still think the technology didn’t have to be made and implemented the way it has been.
I’ve always been irked that in the face of artists being worried about companies replacing their jobs with diffusion technology, there has been an attitude of “yeah, you deserve to not be compensated.” I think it was more blatantly antagonistic early on, as for where we are now, well, the infamous Corridor Crew video. The part that seemed to push it over the edge for the “maybe we were too harsh crowd” was the “we have finally democratized the animation process” speech. How so? There is already tons of free animation software. People have been releasing animation independently on their own for years. Ebsynth has been around for quite a while. As for them listing skill as one of the gaps this tech bridges, but what was stopping y’all from learning to draw their own keyframes?...
That last question might sound really reductive to “this tech is completely fine” people (‘if we did everything the ‘right’ way, it would cost this this and this and take this amount of weeks/months/years/etc’.) and counter productive to my fellow “this tech is screwing me over” people (‘one more closed job opportunity’). But that’s ultimately the difference between someone who can draw something and someone who can’t. Don’t get me wrong, there is doing too much of something by yourself (https://forums.tapas.io/t/when-you-cant-afford-to-pay-for-help-how-do-you-manage-alone/75242/16). But I can’t help but notice it’s very common for people who aren’t very experienced in an artistic field to jump to “Diffusion technology/having someone else do x part for free is the only way for me to do this”. These people are mystifying whatever it is they don’t know how to do. They don’t consider that people they want to replace/freeload off of all had to start somewhere. Those of us who got something done had to scale something back. Many of them don’t have that experience, they over estimate what the bare minimum looks like.
As for the future of the technology... one of the first defenses I’ve seen is that the images generated by diffusion aren’t copyright infringement. Whether or not that’s true, that defense doesn’t seem to account the library of images that were already fed into the machines. The information is still there. As for how copyright law works, what some people think is fair use is actually a case of “the copyright holder didn’t think it was worth going to court over.” My attempt at explaining what that means.
Anyone familiar with the Five Nights at Freddy’s fan game scene? The way copyright law’s written, hypothetically, 30 years after its creator death (assuming he isn’t immortal), whoever inherits his estate snaps their fingers and says “all unauthorized FNAF art, games, gameplay footage, animations, etc, it’s all gotta go.” Legally speaking, if it didn’t file off the serial numbers correctly (or it’s a review, or it’s educational), it’s gotta go, and (assuming Disney doesn’t do it’s thing,) it’ll have to stay gone for another 40 years. (Also, said inheritor can’t just can’t go around republishing said work without the fan artist’s consent. If they still made it, they still own part of it unless otherwise signed away.)
Considering a lot of diffusion stuff threw cautions to the wind in this regard, and the fact that they’re being taken to court in the first place, I don’t think it’ll be an easy victory for them. (Again, they didn’t have to do it this way. The public domain exist.)
(I also low key just think diffusion technology will never “just work” the way people want it to. Aside from it not being actual AI, I’m just unimpressed with most diffusion results. It’s not a ironclad argument, it’s just a honest opinion that colors my view on the subject.)
TLDR: Most pro-diffusion people are just kinda ignorant about the process of actually making the art that was ‘once only accessible to the elite: ‘talented’ people.’
Also, people seem to forget this video exists:
Which, the TLDW is: people have tried to outright replace animators before with technology. The technology needed them for it to actually be effective.
Now history has come full circle yet again, expecting a different result.
(Speaking of the ‘process’, I spent 3 hours writing this... oof.)