Copyright is pretty terrible, and I don’t think we should expand the power of copyright in order to get a handle on AI.
The good news is, we have other tools.
- Gen AI firms are deliberately using creators’ own content to compete in the same space; that’s anti-competitive, hit em with antitrust
- Nobody should be forced to accept an AI training provision in their contract; hit em with unions
- If you identify a private detail of a person through large-scale statistics, that’s an invasion of privacy
- Deepfaking someone is a violation of likeness rights and privacy; for sexual content, it should be a sexual offense as well
- AI slop is polluting our communication channels while the devs say they bear no responsibility for it; we should treat info pollution as seriously as we (should) treat environmental pollution
Content provenance is also a good idea, though I don’t like the centralized nature of C2PA.
We have plenty of options for regulating here. They want you to think copyright is the only path, because they win either way there.
Square root of negative one