4 ms·
There's a group of people who have reinvented religion because they're afraid of an AI torturing them for eternity if they don't work on AI hard enough. It's ve
by sumeno 1y ago
There's a group of people who have reinvented religion because they're afraid of an AI torturing them for eternity if they don't work on AI hard enough. It's very silly but there are many people who actually believe this is a risk: https://en.wikipedia.org/wiki/Roko's_basilisk https://en.wikipedia.org/wiki/Roko's_basilisk
- hollerith 1y agoYou are cherry picking the single most absurd event in a history of over 20 years of public discussion of AI catastrophic risk. Only about .0003 of all public discussion of AI catastrophic risk over those 20 years has invoked or referred to Roko's basilisk in any way. I don't know of anyone worried about AI who is worried mainly because of the basilisk. Next you'll mention Pascal's Mugging, which likewise is the main worry of exactly zero of the sane people worried about AI -- and (despite the representations of at least one past comment on this site) was never even a component of any argument for the dangerousness of continued "progress" in AI.
- sumeno 1y agoSo you agree that there is more than "a single person worr[ied] about AIs taking over humanity." I was specifically pointing out how absurd the most ridiculous people in that category are
- hollerith 1y ago[flagged]
- sumeno 1y agoBecause it is still relevant today: https://en.wikipedia.org/wiki/Zizians https://en.wikipedia.org/wiki/Zizians
- hollerith 1y ago[flagged]
- davidgerard 1y agothat's a truly amazing example that just happened to spring to the forefront of your mind to throw into the discussion
- hollerith 1y agoI merely repeated one of the distortions of events from ancient history that was used more than once on this site in recent months to try to smear the Berkeley rationalists. I think it illustrates the lazy ad hominem tactics of those who are made uncomfortable by any talk of catastrophic risks from continued "progress" in this decade's shiniest technology. I should add by way of explanation that there is a close connection between those who see continued "progress" in AI as very risky and the Berkeley rationalists: in fact, the writings that form the core of the rationalist movement (very analogous to how Paul Graham's writings form the core of HN) were written to try to enable more people to appreciate the riskiness of AI "progress" -- and those with a strong emotional or ideological attachment to (or financial interest in) continued AI "progress" know about this connection, which is why they like to try to smear the rationalists.
- sumeno 1y agoYou're projecting an absolutely wild amount of stuff here
- BeetleB 1y agoI was referring to people I personally know. The existence of crazy/anxious people in the world is well established, and not in dispute.
- joegibbs 1y agoNot even them, simulations of them. They're not even real! Why the hell should I care if some far-future AI tortures simulations of me? Go ahead, spin up a hundred, hell spin up a trillion!
- mmmore 1y agoI really don't think that very many people are concerned about AI because of Roko's Basilisk; that's more of a meme.
- kelseyfrog 1y agoThe AI alignment folks seem much more anxious about an AI version of Pascal's Wager they cooked up.
- mmmore 1y agoSure some of them maybe, but given that many concerned people think the chance of extinction is 10% or higher, it's not really low probability enough to be considered a Pascal's Wager.
- kelseyfrog 1y agoI didn't know that 10% was a common threshold. Thank you for the insight.