Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

So why do they pay to license so much content? And you can’t really be suggesting that public domain should exclude AI training?


That's a good question, why do they choose to pay for certain content while also continuing to insist that all publicly available content should be fair game? It's an incoherent stance, and the only plausible explanation I can come up with is that those "licensing deals" are out-of-court legal settlements in a trenchcoat.


It's understandable that they don't have the luxury of consistency. They make deals to reduce risks. (It's usually not a great business decision to sue someone who is paying you a lot of money, even if in theory the court could order them to pay more. But how long? Also there are already lawsuits ongoing. Maybe they want to get those dropped?)


To sell the illusion of not using copyrighted works they make it publicly aware that they license some works.


You think the public in general is highly concerned with protecting copyrighted works? Like, they’re all saying “I would love to pay OpenAI but they are infringing on my beloved Reuters copyrights”?


"The public"? No. But it makes for nice investigation defense to be able to say you've made an attempt to avoid training on copyrighted works.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: