Government fears AI could be used to spread fake news in run up to elections::undefined

  • UnderpantsWeevil@lemmy.world
    link
    fedilink
    English
    arrow-up
    3
    arrow-down
    2
    ·
    edit-2
    1 year ago

    there is the amplification of the message that is occuring and guiding conversations that is the work on troll farms

    Absolutely. But far more often than not, the calls are coming from inside the house.

    Peter Thiel and The Mercer Family are producing orders of magnitude more right-wing social media flak than any foreign government. Hell, I’d wager that groups like the Oathkeepers and the Proud Boys are producing more flak than any foreign government.

    This more natural “grass roots” conversation is what I’m likening to their concerns about AI.

    We’ve had call-and-response bots for decades now. The threshold by which automated systems can engage with a human participant and goad that person into some kind of activity is disturbingly low. You don’t need some kind of sophisticated AI technology to string people along. Chaturbate bots have been doing this successfully since the 2010s, for instance.

    Displacing all this shit onto “Russia” really blinds people to the native (and very often for-profit) nature of these operations. The vast majority of these spam bots and engagement farming tools are just MLMs wearing political insignia. And you don’t even need to confine yourself to the Republican Party to find them. Online presences like Joshua4Congress and The Lincoln Project were more than happy to farm liberals and leftists for cash by pushing these comically overly-optimistic campaign messages in order to draw in donations that they simply paid out to themselves as consulting fee kickbacks.

    AIs might be better at churning this crap out faster and distributing it farther. But the impact will be marginal relative to what people have been up to manually for some time.