• knfrmity@lemmygrad.ml
    link
    fedilink
    arrow-up
    12
    ·
    2 years ago

    It was also trained by underpaid humans to be liberal.

    Ultimately, no matter how an AI “learns,” it will reflect the worldview of its creator. In this case that’s a handful of billionaires.

    • TarkovSurvivor@lemmygrad.ml
      link
      fedilink
      arrow-up
      5
      ·
      2 years ago

      if you ask chat gpt, it will admit that it’s trained on an English language corpus and that the English speaking world is largely liberal and that chat likely reflects the biases of the corpus on which it is trained.