← Back to context

Comment by direwolf20

6 hours ago

How do we know whether they're people or bots?

Well in my case I was referring to actual vocal conversations I've had with humans, either in person or on MS Teams.

I suppose that there could be an extremely elaborate LLM to control humanoid robots to try and fool me, but I do not believe that's the case.

  • Yet. But another year or two of progress on AI deepfakes and you will be talking to a bot and be none the wiser.

    • I find it odd that Americans aren't angrier about Russia running disinfo bot networks for over a decade now, building up to almost everything in the current cultural moment.

      Also odd that the tech behind this isn't more talked about. I lost the source but there was something about bot networks that could argue both sides of a topic to feign the illusion of a real debate - and this predated ChatGPT by many years.

      Big platforms like Google or X have only mildly experimented with heavenbanning and discourse manipulation at scale. These Russian networks have had at least a decades' worth of experience with it.

      Somehow, in reducing all political opponents to bots, the discourse does seem to forget that there's often someone behind the bots, a tangible nation-state of a target.

      2 replies →

I mean, tens of millions of people voted for this. So even if social media sentiment is mostly bot-driven, it's provably backed up or supported by what real people deeply believe and want and will continue to vote for in mid-terms.