reform AI safety people you might not have -time- to go though Yudkowsky’s entire character arc yourself just skip to the end #nomoa
meh my main objection is to “trying to get a broad swath of the public on board with one’s preferred AI policy is something close to a deontological imperative” which like NO, that is a FACTUAL QUESTION about WHETHER IT WORKS, you are trying to lock in a POSSIBLY NON-WORKING STRATEGY, and if that is the case you will EXPEND VERY PRECIOUS RESOURCES. I don’t think it’s a deontological imperative to ignore public outreach! Extend me the same courtesy!
@TetraspaceGrouping That line of reasoning doesn't look too faulty. AI will likely have enormous influence on the world, so wanting an AI policy that a broad swath of the public endorses is pretty important so that important opinions aren't ignored.
I do agree preventing everyone from dying is more important than everyone having a say in how they're not gonna get mass murdered.
@TetraspaceGrouping Hopefully we'll still have enough agency after the AI takeover to tweak the system to take into account people (and animals) who are ignored today.
@SelonNerias True, for the bigger True Alignment Problem of 🐛Handing Over the Lightcone* I am much more in favour of spending more on everyone knowing what’s happening, and I think if the dangerous time we’re currently in is over we’ll have more slack to achieve that.
*🐛 for uncertain phrasing because this sounds like literally making a big The AGI and letting it rip but people might not do that