• jungle@lemmy.world
    link
    fedilink
    arrow-up
    7
    arrow-down
    1
    ·
    edit-2
    1 year ago

    That’s not AI works. It’s exactly as biased as the humans who produced the content on which it is trained.

    That said, I also don’t believe these models have been trained exclusively on white straight men’s conversations, that would take some effort to achieve.

    More likely, it’s been trained on internet forums, so similar to what it’s being asked to moderate. And as long as there’s a human at the other end of an appeal, it should be fine.