I assume they all crib from the same training sets, but surely one of the billion dollar companies behind them can make their own?

  • NaibofTabr@infosec.pubEnglish
    43·
    2 days ago

    Trained on a corpus of messages written primarily by the people who spend the most time using the Internet to talk to their friends… teenagers.

    Imagine dumping the entire content of Snap, Instagram, Kik, Facebook messenger, etc, into a blender and attempting to derive a style of speech from it. The most impressive thing about these LLMs is that they’re (marginally) coherent.

    • otp@sh.itjust.works
      6·
      2 days ago

      Given that they’re using stack overflow, I’m guessing they also have “professionals who have more time than work to do at work”

    • HammyMcBurgers@sh.itjust.worksOP
      46·
      2 days ago

      Lemmy try to not be ageist challenge (impossible).

      LLMs talk more like middle aged aunts and uncles than teeny-boppers anyway.