• Phoenixz@lemmy.ca
    link
    fedilink
    arrow-up
    33
    arrow-down
    1
    ·
    9 days ago

    The court, which stressed that it’s not opining on fair use, said that a jury could undoubtedly find that the output infringes on Martin’s works.

    Well of course it does. Llms can only talk about what they know, they can’t talk about these books if the books are part of its knowledge, similarly like how I can’t talk about the books if I haven’t read them. The entire point is that the LLM got the data from the books (and of course without paying) and with that, open AI and others are profiting “massively” from it without giving anything back to the author.

    Llms, as they currently are, should just be prohibited, or we should abolish copyright entirely

    • Ludicrous0251@piefed.zip
      link
      fedilink
      English
      arrow-up
      1
      ·
      6 days ago

      Yes and no - there’s so much GoT fan content out there that they could theoretically pick up on the key names and places without the books

      That said, I’d be willing to bet these shitty AI companies went the extra mile to pirate these books to train their LLMs, because copyright law is only for the poors.

      I generally despise LLMs, but I’d be intrigued to see LLMs be the thing that reigns in copyright law a little bit because it’s gotten too out of hand.