is because the work big AI models steal feels disproportionately perpetrated against "small" artists?In other words, is it the case that Big AI (tm) is too wary to train directly on the Disney/Pixar vault, but is just as happy to go after your fanart, your microposts, etc. because they know you as an individual do not possess nearly the same legal teeth as the Walt Disney Corporation?
I frame this as a question because, I genuinely don't know if it's true! I'm not an expert in the AI space. Maybe AI is getting trained on the Disney vault without legal permission and we just don't really know it. But either way, I think it is true at least that many people feel like small artists are being targeted— so when they they make the copyright argument, they feel like they're standing up for the little guy. And they are! But largely I concur with the folks saying lately that the copyright framing is counterproductive (in the comments of this post and I believe jan misali had a post about it recently too). They may be standing up for the little guy, but the big corps are reaping all the benefits. Still, it might be useful to consider if this is where the impulse is coming from, and if so, might that impulse be redirected in ways that are more productive for the small creators out there?
In general when it comes to evaluating any kind of technology, public policy, cultural facet, etc., I'm a fan of the "you will know it by its fruits" strategy. Is the AI in question producing materially worse conditions for people, or better? If both, who are the worse conditions falling on— are they hurting the people who can stand to lose a little, or are they hurting the folks who are already on the margins? If "the purpose of a system is what it does," examining that output— not the theory, not the possibility, not the intent— often gives much better leverage for pulling the issue apart. (And to be clear I think the post I linked above is mostly doing that, and I am only mildly responding to it here!)