OpenAI now tries to hide that ChatGPT was trained on copyrighted books, including J.K. Rowling’s Harry Potter series::A new research paper laid out ways in which AI developers should try and avoid showing LLMs have been trained on copyrighted material.

  • uis@lemmy.world
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    2
    ·
    1 year ago

    Huh. Doesn’t this means technically AI cannot do copyright infringement.

    • LifeInMultipleChoice
      link
      fedilink
      English
      arrow-up
      2
      ·
      1 year ago

      Nah, it would mean that you cannot copyright a work created by an AI, such as a piece of art.

      E.g. if you tell it to draw you a donkey carting avocados, the picture can be used by anyone from what I understand.

      • uis@lemmy.world
        link
        fedilink
        English
        arrow-up
        1
        ·
        1 year ago

        you cannot copyright a work created by an AI, such as a piece of art.

        That’s what I said. Copyright infringement is when there is another copyrightable object that is copy of first object. AI is not witin copyright area. You can’t copyright it, but also you can’t be sued for copyright infringement too.

        if you tell it to draw you a donkey carting avocados, the picture can be used by anyone from what I understand.

        Yes. Same for Public Domain, but PD is another status. PD applies only to copyrightable work.