• amemorablename@lemmygrad.ml
    link
    fedilink
    arrow-up
    3
    ·
    2 days ago

    Fair point worth considering, especially as time and energy is concerned in dealing with names that are spewing BS. I’d expect bots to be more so the case on platforms like Twitter or Reddit (and I remember some of them being very obvious after Musk took over). On here, it’s a bit more gatekept and obscure, and even with LLMs being more capable of impersonating a person than previous technology, they could easily go off the rails without monitoring or use AI-ism tells. So if someone were to try a thing like that in a context this confined, I’d still expect them to be doing some reading and using the bot more as an assistant to BS. Whereas like on Reddit, the sheer quantity of input on bigger subs makes it easier to flood shit with BS, while making it a painful endeavor to go through and look at each account to see what weird shit they may be getting up to.