• jeffhykin@lemm.ee
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    1
    ·
    10 个月前

    It’s not fair game for for-profit bussinesses training LLM’s. That’s part of why Reddit made the move; so that companies would need to pay Reddit for access to the data for legally training models

    • AustralianSimon@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      10 个月前

      They changed the terms and made the API pay to use for large volumes of use. People using it to train models have already pillaged what they need and you can get the data prior to APIgeddon elsewhere.

      • jeffhykin@lemm.ee
        link
        fedilink
        English
        arrow-up
        1
        arrow-down
        1
        ·
        edit-2
        10 个月前

        Sure, but it’s still true that there are legal protections we can add that make it not fair game for Lemmy. At best it would be unfair-game (illegal scraping of Lemmy)