• JohnEdwa
    link
    fedilink
    English
    arrow-up
    5
    ·
    edit-2
    2 days ago

    Making stuff up is the entire function of an LLM, they are predictive text generators that string words together based on how the algorithm predicts a human would given the same input to produce a plausible answer. Not the correct answer necessarily, only one that feels like it could have been written by a human.

    Censoring or controlling them in a way people would want to or expect companies to do is basically impossible because it would require them to actually be able to understand what they are talking about in the first place.