• QuadratureSurfer@lemmy.world
      link
      fedilink
      English
      arrow-up
      7
      arrow-down
      3
      ·
      edit-2
      2 months ago

      Yeah, a decision to modify copyright so that it affects training data as well would devastate open source models and set us back a bit.

      There are many that want to push LLMs back, especially journalists, so seeing articles like this are to be expected.

      edit: a word.

      • CaptainEffort@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        3
        ·
        2 months ago

        Exactly this. If you want ai to exclusively be controlled by massive companies like Meta and Google, this is how you do it. They’ll be the only ones that can afford to pay for public copywritten content.

    • tabular@lemmy.world
      link
      fedilink
      English
      arrow-up
      7
      arrow-down
      3
      ·
      2 months ago

      Copyright is the legal method to limit redistribution of easily copied material, not as if there’s anything else people could appeal to.

      I ain’t a fan of copyright but make it last 10 years instead of X + infinity and maybe it’s not so bad. I can’t argue against copyright fully as I think copyleft is essential for software.

      • OsrsNeedsF2P@lemmy.ml
        link
        fedilink
        English
        arrow-up
        4
        ·
        2 months ago

        But those aren’t the options on the table right now. The options are “nullify copyright” or “keep infinite copyright”