I've tried not to talk about AI art stuff much here but uhh, the actual legal arguments put forward by the plaintiffs in that stable diffusion lawsuit seem bad. Like, really bad. Factually incorrect claims aside, I'm talking "most expansive interpretation of copyright and IP ownership I've ever seen seriously put forward, by an enormous margin" bad. Whatever your feelings about AI art or even any particular model, we REALLY don't want a precedent for people claiming ownership over anything at all influenced by their work.
And just so I'm not misinterpreted, when I say "pieces influenced by their work" I am NOT talking about AI generations. See, when something like stable diffusion "learns" an artists style, it mostly doesn't do it by looking at THEIR art. At least, not for anyone remotely recognisable or influential. It does it by looking at vast swaths of art that references or is inspired or influenced by that artist. It looks at countless things drawn by countless humans and posted online that could be described as "influenced by such-and-such artist".
Now, in terms of how this influences your opinion on these models and AI art in general, you might think "that sounds even worse!" right? But the thing is, those numerous small artists, they're not the plaintiffs here. Instead what is happening is that, in claiming ownership over the training data and the results thereof, the original Big Artist with a recognizable name and style is claiming ownership over the works of all of those other people. There is a legal argument being made here that has nothing to do with machine learning and which is incredibly, almost unprecedentedly disgusting.
edit: i would like to add, this is a problem with the legal arguments -currently being put forward-, this is not something i'd view as an inherent or irreconcilable problem with trying to sue stability diffusion or openai or anyone else in that field