News for hackers. Stuff that matters.

A comment on OpenAI's justification for why training data is fair use, not infringement [pdf]. Parent comment

Re:OpenAI's justification for why training data is fair use,... by Pet_Ant on Thursday October 05, 2023 @02:44PM (#37782401)
You aren't licensing it. You are reading it. A machine reading it is no different than a human. I mean it really shouldn't be. Should a visually impaired person using a screen reader have to "license" the book?
[ Parent | on HN ]
Re:OpenAI's justification for why training data is fair use,... by SirMaster on Thursday October 05, 2023 @03:09PM (#37782846)
But not all the works that have been consumed by LLMs are available to read without a license to do so.
[ Parent | on HN ]
Re:OpenAI's justification for why training data is fair use,... by pclmulqdq on Thursday October 05, 2023 @05:12PM (#37784368)
That particular transformation of the book does well on a fair use analysis. If the blind person were to use the screen reader to put the book on Spotify, it would almost certainly fail a fair use analysis.
[ Parent | on HN ]
Re:OpenAI's justification for why training data is fair use,... by hashtag-til on Thursday October 05, 2023 @05:52PM (#37784747)
People come up with the most bizarre corner cases to justify freeloading on other’s intellectual property.

Content is obviously being relicensed as the models are not fully open.

On top of that, it is also being exposed as pay-for-use subscription models without paying anything to original authors. How is that fair use?

[ Parent | on HN ]