• capital
    link
    fedilink
    English
    614 days ago

    We get it, y’all hate LLMs and the companies who make them.

    This comparison is disingenuous and I have to think you’re smart enough to know that, making this disinformation.

    If/when an LLM like ChatGPT spits out a full copy of training text, that’s considered a bug and is remediated fairly quickly. It’s not a feature.

    What IA was doing was sharing the full text as a feature.

    As far as I know, there are some court cases pending regarding determining if companies like Open AI are guilty of copyright infringement but I haven’t seen any convictions yet (happy to be corrected here).

    All that said, I love IA and have a Warrior container scheduled to run nightly to help contribute.

    • @[email protected]
      link
      fedilink
      English
      414 days ago

      Hmm, true. IA wouldn’t be as supported if we couldn’t get the full text of the source.

      Can you tell me more about the “warrior container”?

    • @[email protected]
      link
      fedilink
      English
      1
      edit-2
      14 days ago

      have a Warrior container

      This is an ArchiveTeam project, which is a totally separate effort to the Internet Archive. As far as I know, they’re not related other than the fact that ArchiveTeam use The Internet Archive for storage.

      • capital
        link
        fedilink
        English
        114 days ago

        Ahh my mistake.

        Might be time to financially contribute to IA.