• a_beautiful_rhindB
    link
    fedilink
    English
    arrow-up
    1
    ·
    1 年前

    CAI/LaMDA and PI are trained more on convos than facts and QA. So they appear more “real” and personable.

    I don’t think we have an open model like that yet. Trained, not finetuned. Hence no new Blake Lemoines and a distinct feeling of “lack” when interacting.

    That’s my crackpot theory.

    • FPhamOPB
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 年前

      So you saying, basically a lame google version of pygmalion…