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 verdverm on Thursday October 05, 2023 @12:26PM (#37780622)
With Chain-of-Thought and Mixture of Experts, it may be that, in the long-term, most LLM output is primarily consumed by other LLMs [1]

[1] (meme, the future of communication) https://www.reddit.com/r/ChatGPT/comments/123tyge/the_future...

[ Parent | on HN ]

There are no comments on this comment yet.