About 5% of the total

  • ChowJeeBai@lemmy.world
    link
    fedilink
    English
    arrow-up
    6
    arrow-down
    1
    ·
    8 months ago

    The way things are going, with everyone training on everyone else’s data, we’re witnessing the death of creativity and the rise of a grey blandness.

    Except when it comes to how ‘creative’ the prompt engineers are when it comes to coaxing an output of these AIs.

    • streetfestival@lemmy.ca
      link
      fedilink
      English
      arrow-up
      3
      ·
      edit-2
      8 months ago

      Your comment reminded me of how the quality of web search results has really deteriorated from say 5 years ago. Putting it all together, I wonder if we’re also witnessing the entshittification of the open internet, paving the way for the big tech companies to create paywalled platforms of more useful internet (edit: I guess we are already seeing that with shitter, facebook, netflix, spotify etc., although at the moment they’re not all paywalled in the traditional sense)

  • gregorum@lemm.ee
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    3
    ·
    8 months ago

    well… ok, so does that level of abstraction give them cover on not having used any person’s/artist’s actual copyrighted material for training data? that may be a gray area for some, but i’m kinda satisfied.

    • TheOneCurly@lemm.ee
      link
      fedilink
      English
      arrow-up
      3
      arrow-down
      2
      ·
      8 months ago

      So if I buy a load of stolen apples and bake them into pies I’m in the clear?

      • gregorum@lemm.ee
        link
        fedilink
        English
        arrow-up
        3
        arrow-down
        4
        ·
        edit-2
        8 months ago

        This isn’t the same thing. This is a matter of copyright infringement, not theft. Yours is the same flawed idea used by the film and music industry to claim that piracy is “theft” when it’s not.

        I reject your question due to its fallacious premise

        • jacksilver@lemmy.world
          link
          fedilink
          English
          arrow-up
          6
          ·
          8 months ago

          The phrasing was poor, but what Adobe did is effectively the money laundering version of copyright theft. Just churn it a couple of times and then the original creators are now “AI models” and not the peoe whose work was used to create the “AI models”

          • gregorum@lemm.ee
            link
            fedilink
            English
            arrow-up
            1
            arrow-down
            2
            ·
            8 months ago

            I agree that what Adobe did was… underhanded, but if anyone infringed copyright, it was Midjourney. I don’t really disagree that Adobe’s grubby for what they did, but my ethical evaluation here is whether they violated their presentation/promise that their Firefly AI wasn’t sourced from human-produced content (ill-gotten or otherwise). It wasn’t.

            Noooow, it would have been naive to think they wouldn’t find some shitty way around that, and shame on all of us for not being as imaginative as Adobe and their lawyers were, but here we are.

            And, FWIW, I was a little sly myself when I said I was only ‘kinda ok” with it. In my equivocation, I didn’t really commit, and I’m still evaluating my position. I don’t really see this as a black-and-white issue. Not to say that I’m a moral relativist, but I understand that there’s a lot of nuance here, and i understand how something’s are worse than others.