• eleijeep@piefed.social
    link
    fedilink
    English
    arrow-up
    23
    ·
    2 天前

    I’ve been waiting for something like this to appear. Not just a “plagiarism detector” but something that actually identifies the data in the training pool that most closely represent a particular AI model output. You could do the same for text and images too, and I’m surprised this is the first one that I’ve heard of.

    I’m not a fan of the MAFIAA but if this type of reverse-search tech can hold AI companies to account then it’s a step towards reining them in.

    • oni ᓚᘏᗢ@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      15 小时前

      Cool would be that AI products have their own data as sufix in the content that they generate, something like ID3, and we get new AI file extensions, like .aijpg, .aipng, .aimp3, etc.

    • XLE@piefed.social
      link
      fedilink
      English
      arrow-up
      4
      ·
      1 天前

      When it comes to stuff like copyright lawsuits against AI companies, the only way you can fight big money (at least in the US) is with more money.

    • General_Effort@lemmy.world
      link
      fedilink
      English
      arrow-up
      2
      ·
      1 天前

      I remember a guy about 3 years ago trying that grift with images. Went nowhere because the images it flagged as the “source” looked nothing like the generated images. In music, it might be more successful. Marvin Gaye’s estate showed the way.