This is the humanless future, hurray!

  • TwilightVulpine@kbin.social
    link
    fedilink
    arrow-up
    1
    ·
    1 year ago

    Could a language model actually independently discern if a source is trustworthy? Seems that’s something difficult to determine when it comes to possible leaks. The kinds of AIs that we have today can’t really conceptualize a world outside the texts they process, they can only check based on other texts and user input.