@felurian8 haha I do trust the Safety ppl truly mean ainotkillusall - and that's the scenario, if real, that I agree justifies extreme measures. Other possible disruptions and dystopic scenarios are separate conversations about how bad is the outcome and how can we mitigate it.
More notes on nearby ideas
Reply · Jun 3, 2022@deepfates My biggest concern, personally, is that someone takes AI safety rhetoric very literally and decides to commit an act of terrorism that…Reply · Nov 24, 2023@QiaochuYuan yeah that's what makes them scary- killing most ppl would seem like a good trade in their eyesReply · Aug 27, 2023@ESYudkowsky this is why I chat w gpt about scenarios, tho I am still not convinced of *extinction* risk as opposed to bad scary stuff that kills lots…Reply · Mar 29, 2023@TetraspaceWest be very insistent!!! my p(doom) is tiny, but it's sooo annoying when goal posts shift mid convo- I take MIRI at their word that they…