• Sorgan71@lemmy.world
    link
    fedilink
    arrow-up
    17
    arrow-down
    17
    ·
    6 months ago

    But its not plagiarism. Its just training. In terms of information its no different from a human just seeing an art piece.

    • Kachilde@lemmy.world
      link
      fedilink
      arrow-up
      13
      arrow-down
      3
      ·
      6 months ago

      But they don’t learn when you train them. They mashup what they have been fed, and shit out an approximation of your request each time. When you type “in the style of xyz”, the system doesn’t remember who xyz is. It does a search for data on xyz, and copies it.

      If these models were learning, you wouldn’t still need paragraph long prompts every time you wanted to plop out another plastic-skinned anime chick with fucked-up flipper hands and huge bazoongas. It would have learned how to make that shit by now.

      • Hackworth@lemmy.world
        link
        fedilink
        English
        arrow-up
        3
        arrow-down
        6
        ·
        6 months ago

        Honestly, I’m having a ton of fun with generative A.I. as a marketing producer. It’s allowing me to play in new ways and realize ideas that I would have never had the time or resources to execute otherwise. It’s given me the confidence to explore entirely new mediums and workflows that have bled over into my personal art.

      • herrvogel@lemmy.world
        link
        fedilink
        arrow-up
        7
        arrow-down
        10
        ·
        6 months ago

        the system doesn’t remember who xyz is. It does a search for data on xyz, and copies it

        This sentence is so not how those AI models work, that it leads me to believe you don’t actually know enough about them to be having this discussion.

    • kshade@lemmy.world
      link
      fedilink
      arrow-up
      2
      ·
      6 months ago

      It’s like a cover band trying to replace the real deal by writing “original” songs that are suspiciously derivative.