Comment by bhouston

My point is that robots will not cause a genocide. The fact that we are talking about this instead of anything salient to hackers or news is evidence that we're sufficiently outraged by it. Our businesses have protocols for handling robots that present harms to their workers, our governments have contingency plans to prevent serious adversaries from attaining real power. What, if anything, is a serious depiction of how LLMs would pose a societal threat? This article certainly isn't.

If you look at pre-Nazi Germany, they collapsed for exactly the reason I described in my previous comment. Nazism didn't just manifest overnight; the German government was weakened, the economy was ruined, and the people didn't feel like their government represented them anymore. The government itself was hobbled by the Treaty of Versailles, which left them with a fully industrialized nation that was 60% unemployed, deeply in debt, and arguably unable to defend itself. The tragedy of Nazism isn't that the German people "fell for it", but instead that they perceived no other choice. It wasn't any special Nazi technology invented to destroy the government - simply MEFO bills and some under-the-table industrial influence. If you want to talk about scary stories to make bureaucrats cry, issuing two currencies at once is scarier than 10 Terminators.

Make me afraid of a MechaHitler, seriously. Make me afraid of a purpose-built Stalin, Reagan or Pol Pot AI, if you think that the simple existence of an edgy LLM is enough to endanger society. I fully intuit the danger that the Third Reich imposed on the rest of the world, and I'm asking you to explain how a functional government would allow humanoid or non-humanoid robots to possess enough sapience to interact with humans in public.