The Unbelievable Scale of AI’s Pirated-Books Problem
The Unbelievable Scale of AI’s Pirated-Books Problem
Meta pirated millions of books to train its AI. Search through them here.
I like Molly White’s recent take, that it might be more productive to treat this as a labor issue instead of a copyright issue (at least in principle). Even if the AI corporations aren’t technically re-selling copyrighted works, they’re still profiting from the authors’ unpaid labor.
That's an evergreen. Tried and trusty.
Copyright law needs to be fixed, and not in favor of these corporations, but in favor of artists.
Wanting copyright law to be fixed does not mean wanting it go away entirely for the sake of bullshit like LLMs.
Check out the research of Rufus Pollock who did a bunch of complex math to show ideal copyright length should be 15 years.
https://rufuspollock.com/papers/optimal_copyright_term.pdf
If the admins of the Pirate Bay got put in prison for far less piracy and far less profit from piracy... the same ought to happen to Sam Altman et. al.
Good stuff. Too many people lately are all "no copyright would be an improvement." Yeah, maybe for the corps who could freely use your output as they wish.
What a brilliant Idea. Hover up all copyrighted works then regurgitate it in different forms without having to pay the copyright owners. Sounds like a great tech bro idea.
The finance bros tried that one too. Mortgage-backed security was the magic word. Cut up all the little mortgages, repackage them, and sell for profit. Then it all crashed down in 2008.
There's no problem. Those books should all be in the public domain anyway.