Raw LLM Responses

Is it just me or it almost feel like a misplaced paternalism when he says "You can send them messages that they find very convincing not to vote for someone, for example" I mean, the propaganda and the "art" of rhetoric has been around for centuries, and it is also common sense (since the interviewed himself appeals to it) that convincing someone is NOT an easy task, at all! Specially if such matter is engulfed in controversy, dissident and debate. On the other hand, when we have CONSENSUS, there's where the danger lies! There are extensive research and published experiments where we managed to "convince" people to give wrong answers if their peers are also giving the wrong answers and being flagged as correct. I'd love more meat on his argument of "corrupting elections" based on AI. I am sure there are ways to do that (I mean, humans have been corrupting elections even before we had electronic machines...) but I'd counter that people, in general, are growing ever more skeptic of the information they are fed, and less prone to trust things on a knee-jerk basis (unlike the interviewed who admitted he would immediately click on a link saying "Trump did this crazy thing"). All in all, goes to show he is biased against a certain political side, which does not make him look good, or rational, on his arguments.