Ha ha, no. It wouldn’t even work. Nevermind the people saying big-iron companies will just pay to own the AI market - this technology is already halfway democratized, and this move would guarantee all that shit moves to data-hoarders with fat GPUs.

You don’t need to pay money for an English corpus when you have Wikipedia, Archive.org, and The Pirate Bay.

I have twenty-five years of images saved because I felt betrayed by a site going down, back in the AOL days.

None of us have five zillion compute-hours sitting beneath our desks, but a lot of research is about avoiding the need for that, and it’d be a lot more if all the cool stuff was happening with individual randos.

Ditto any effort to identify specific sources, or even spot AI content automatically. You wanna make this an adversarial ordeal? Cool, that’ll push everything toward being totally inscrutable. A fine goal.