• LainTrain@lemmy.dbzer0.com
        link
        fedilink
        English
        arrow-up
        0
        arrow-down
        1
        ·
        9 months ago

        Open source doesn’t really count, especially with Lemmy where you just can host it yourself and there’s no way for anyone to harvest your data really

  • FaceDeer@kbin.social
    link
    fedilink
    arrow-up
    0
    ·
    9 months ago

    Those adorable baby photos you posted for your friends and family? Sorry, they’re Meta’s now.

    You literally gave them to Meta. You uploaded them to Meta after agreeing to Meta’s terms of service, which essentially say “we can do what we want with the stuff you give us.” Big Bad Meta isn’t sneaking into your houses and stealing photos off of your nightstand, they’re not “seizing” anything.

    Also, Meta’s AI “pipe dream” has so far been very successful. Their open-source libraries and LLaMA models have become industry standards. Their future work is likely to be productive.

    • LainTrain@lemmy.dbzer0.com
      link
      fedilink
      English
      arrow-up
      0
      arrow-down
      1
      ·
      9 months ago

      It’s very impressive to me that 7B LLaMA actually says anything coherent if it was trained with FB data instead of suggesting that the best way to slice strings is with a bleach enema and some essential oils…