• JoeKrogan@lemmy.world
    link
    fedilink
    English
    arrow-up
    13
    ·
    11 months ago

    Thats actually not a bad idea, train a model with all the data in scihub a then release the model to the public

    • Liz
      link
      fedilink
      English
      arrow-up
      7
      ·
      11 months ago

      It would still be useless, the thing would just produce bullshit papers.

      • Chocrates@lemmy.world
        link
        fedilink
        English
        arrow-up
        1
        ·
        11 months ago

        That would likely be explicitly illegal if the NYT case succeeds (and it isn’t fraud as OpenAI alleges)

    • ඞmir@lemmy.ml
      link
      fedilink
      English
      arrow-up
      1
      arrow-down
      1
      ·
      11 months ago

      This is just the most inefficient zip file ever created