• @givesomefucks
    link
    English
    4922 hours ago

    I’m curious if this was some kind of DDOS, or if someone was trying to use it to train AI and was pulling everything to make a local copy

    About the only way to ensure you’re not training AI on bots, is to use “old internet” when bots were more obvious

    • Aatube
      link
      fedilink
      4122 hours ago

      Someone already announced on X that they took the archive down because “NATO bad”

      • @givesomefucks
        link
        English
        6722 hours ago

        Yeah, but that’s the same as someone writing on a bathroom stall at this point.

    • TimeSquirrel
      link
      fedilink
      3521 hours ago

      use “old internet” when bots were more obvious

      This is going to become a valued commodity like pre-atomic low background steel, isn’t it?

      • @givesomefucks
        link
        English
        2121 hours ago

        And it’s mostly all in one place…

        I never realized how valuable that data was

      • @[email protected]
        link
        fedilink
        English
        1021 hours ago

        It already is as far as I know. I’ve heard before that ChatGPT is strictly trained on data from before, like 2018 or so for this reason.

      • Echo Dot
        link
        fedilink
        English
        315 hours ago

        Right but they’ve achieved absolutely nothing other than being mildly annoying. Doesn’t really seem like it would be worth funding.

        • @[email protected]
          link
          fedilink
          English
          313 hours ago

          That’s not entirely correct: One achievement was me donating money to the IA after this nonsense 😁