it’s practically impossible to comply with this law.
It’s practically impossible when you’re scraping the whole damn internet with absolutely no regard to any copyrights whatsoever, yes.
The whole point of this law is to make AI companies stop doing that. To make them stop stealing every copyrighted work they can get their hands on to feed their plagiarism machines.
(It is possible to train an AI while still respecting copyright. Adobe’s image-generating AI features were trained only on images that are public domain or images Adobe actually owns the rights to.)
Adobe did something ethically?
BWAHAHAHAHAHAHA
Sorry
You were saying that you actually believe that?
Adobe’s image-generating AI features were trained only on images that are public domain or images Adobe actually owns the rights to.
Absolutely not true. They trained their models on all the work people put up on Adobe Stock, changed the terms of use to be opt-out, and then disseminated the lie they’re they’re doing “ethical AI”. It’s complete bullshit. There’s no way to train generative AI and respect copyright.
No, they didn’t do it ethically. But they did to it legally. Which is more than you can say about most AI companies.
A future international loophole still exists. A California-based software developer complies, passes copyright audits, and then creates his proprietary software. But he will just export it to his home country, renames it, and the world still gets its virtual Taylor Swift or Lucy Lu bots.
A lot of student ai developers and ai projects are only in California on a temporary basis.
