• balsoft@lemmy.ml
    link
    fedilink
    English
    arrow-up
    1
    ·
    2 days ago

    GPL absolutely should cover shit like this. Training an LLM on your code makes it definitionally a derivative work, therefore it must be licensed under GPL too (with limited fair use exceptions which shouldn’t apply here). The problem is that the US government is not willing to enforce this at all, because it is owned by the same billionaires as the AI companies.

    • merc@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      1
      ·
      2 days ago

      Training an LLM on your code makes it definitionally a derivative work

      If so, then every painter who studied Picasso is making Picasso-derived works. That’s not how copyright works.

    • The_Decryptor@aussie.zone
      link
      fedilink
      English
      arrow-up
      1
      ·
      2 days ago

      The problem is that the US government is not willing to enforce this at all, because it is owned by the same billionaires as the AI companies.

      That’d be an uphill battle, even prominent OSS projects would fight against that unfortunately.

      If the output of an LLM would found to be derivative of the input, that’d cause lots of problems for (e.g.) Linux, they love claude and have been funneling its output into the kernel for a while now, they’d rather not think about the licensing situation there.