Re:Extracting books from production language models (2026)
by isodev on Saturday January 10, 2026 @08:08PM
(#46571732)
But that would only hide the problem, doesn’t resolve the fact that models, in fact, violate copyright
[ Parent
| on HN ]
Re:Extracting books from production language models (2026)
by jgalt212 on Saturday January 10, 2026 @08:42PM
(#46571943)
True, but it would certainly reduce litigation risk in so much as copypasta is ipso facto proof of copyright violation.
Re:Extracting books from production language models (2026)
by protocolture on Saturday January 10, 2026 @09:32PM
(#46572172)
That hasnt been established. Theres no concrete basis to assert that training violates copyright.
Re:Extracting books from production language models (2026)
by isodev on Sunday January 11, 2026 @03:21AM
(#46573680)
Everyone knows that what models do to obtain training data is not legal. We just need a very old system about copyright to catch up already so we can ban the practice.
Re:Extracting books from production language models (2026)
by protocolture on Sunday January 11, 2026 @08:04PM
(#46582436)
"What they do to obtain" ah nice goal post shift.
When legally obtained, training is fine. Training doesnt violate copyright. Unauthorised copying and distribution does. Which is why OpenAI should have just paid for physical copies of all those books and scanned them.
>We just need a very old system about copyright to catch up already so we can ban the practice.
No we really don't need copyright to get worse. Its pretty damn harmful as it is.