My best guess for the hopeful outcome is the ai starts tacking on the license magic words at the end of things it says... But ultimately it feels like a digital version of sovereign citizens to me.

It's not enforceable in any way, it's just virtue signaling. Lemmy itself is a privacy dumpsterfire. GDPR compliance is literally impossible.

How would I prove it? I don't know. Do I think it will work? Probably not. But if I have the license someone might find it when a LLM accidentally reveals that it was trained with data that is under that license, and maybe the EU does something about it. Maybe the Pirate party will make the EU do something about it? Who knows? But they are the only ones I see that are actively trying to protect all of us an our right to privacy, and for that they have my vote! 🏴‍☠️🇪🇺

This is a good question but I would just like to point out that 15 years ago nobody would have predicted that the questions we were asking and answering on stackoverflow would be used to train models (and that open source would have it's license violated so brazenly) and that if you tried to delete your contributions because you didn't want them to be used to train models you would get banned from the site, so even though adding a license to your comments might be meaningless, it might also be a powerful tool down the line. You never know how it'll go.