• pebbles@sh.itjust.works
    cake
    link
    fedilink
    English
    arrow-up
    9
    ·
    1 day ago

    The days of just scraping the internet are over. Data quality is a huge part of LLM training, I don’t think insta getting sloppified will save us unfortunately.

    • Jiral@lemmy.world
      link
      fedilink
      arrow-up
      1
      ·
      23 hours ago

      Yeah, but you need huge amounts of data to train, where to take it if not from the internet? I mean the existing books, records etc before the slop age are soon all fed as well.

      • pebbles@sh.itjust.works
        cake
        link
        fedilink
        English
        arrow-up
        1
        ·
        7 hours ago

        From my understanding a lot of books are still getting digitized, so there is a constant trickle of old literature and news papers and such. But really the answer is synthetic data. Data made by AI and other means for the purpose of training. Which that is very different than just insta slop.

        • Jiral@lemmy.world
          link
          fedilink
          arrow-up
          1
          ·
          4 hours ago

          The digitisation is done at such verociuous speed that this will soon slow down to a trickle because anythimg left is hard to come by.

          “Synthetic data” created with contribution from LLMs is poisining the models just as well. It is leading to reinforcement of hallucinations and artefacts.

          • pebbles@sh.itjust.works
            cake
            link
            fedilink
            English
            arrow-up
            1
            ·
            2 hours ago

            I would reccomend you look into sythetic data. It is a real thing and is used intentionally in basically every LLM of the past year or more. Last I heard of an LLM without sythetic data was Dots LLM1 and that wasn’t competitive when it came out.

      • luciferofastora@feddit.org
        link
        fedilink
        arrow-up
        2
        ·
        16 hours ago

        What they meant is that you could no longer just scrape the internet, but would need to filter the data for quality before mixing it into the slop substrate.