3 ms·
Tbh, I can't follow any of the reasoning that AI is a danger to anyone, except through the traditional form of enabling the worst parts of bureaucracy. I just t
by PittleyDunkin 2y ago
Tbh, I can't follow any of the reasoning that AI is a danger to anyone, except through the traditional form of enabling the worst parts of bureaucracy. I just think it's further evidence of a tremendously wasteful and suicidal society.
- ben_w 2y agoThat's one possibility for the harms it may cause. I think there's a reasonable chance that someone who thinks the idea is laughable will, at some point, ask an AI to destroy the world specifically to demonstrate that it can't… actually, that part has already happened*, the questions are: (1) how competent does it have to be at following that instruction to actually do so? (2) What resources would someone have to allow it to access (deliberately or accidentally) to pull that off? * someone has asked an AI to destroy the world, but I don't know their motivation. Fortunately for everyone, this one was kinda dumb: https://community.openai.com/t/chaosgpt-an-ai-that-seeks-to-destroy-humanity/160028 https://community.openai.com/t/chaosgpt-an-ai-that-seeks-to-...