4 ms·
Are nuclear weapons and their effects only hypothesized to exist? You could still create cults around them, for example, by claiming nuclear war is imminent or
by RandomLensman 3y ago
Are nuclear weapons and their effects only hypothesized to exist? You could still create cults around them, for example, by claiming nuclear war is imminent or needed or some other end-of-times view.
- ethanbond 3y agoThere were people who were concerned about global annihilation from pretty much the moment the atom was first split. Those people were correct in their concerns and they were correct to act on those concerns.
- RandomLensman 3y agoIn a way, they were not correct so far (in part also because of their actions, but not only due to their actions).
- ethanbond 3y agoPretty much exclusively due to their actions and some particularities of the specific technology which don’t seem to apply to AI.
- RandomLensman 3y agoIf you put, for example, US presidents in the concerned group, i.e., actual decision makers then fair enough. But it wasn't just concerned scientists and the public.
- ethanbond 3y agoUhh correct. Unsurprisingly though, many of the people with the deepest insight and farthest foresight were the people closest to the science. Many more were philosophers and political theorists, or “ivory tower know-nothings.”
- RandomLensman 3y agoMaybe. There were also those scientists working actively on various issues of deterrence, including on how to prevail and fight if things were to happen - and there were quite a few different schools of thought during the cold war (the political science of deterrence was quite different from physical science of weapons, too). But the difference to AI is that nuclear weapons were then shown to exist. If the lowest critical mass had turned out to be a trillion tons, the initial worries would have been unfounded.
- ethanbond 3y agoNone of whom were people saying "there's no risk here just upside baby!"
- RandomLensman 3y agoPeople were on totally opposing sides on how to deal with the risk, not dissimilar to now (with difference that the existential risk was/is actual, not hypothetical).
- ethanbond 3y agoSure, there are also some (allegedly credible) people opening their AI-optimist diatribes with statements of positive confidence like: “Fortunately, I am here to bring the good news: AI will not destroy the world” My issue is not with people who say “yes this is a serious question and we should navigate it thoughtfully.” My issue is with people who simply assert that we will get to a good outcome as an article of faith.
- RandomLensman 3y agoI just don't see the point in wasting too much effort on a hypothetical risk when there are actual risks (incl. those from AI). Granted, the hypothetical existential risk is far easier to discuss etc. than to deal with actual existential risks. There is an endless list of hypothetical existential risks one could think of, so that is a direction to nowhere.