Is AI Safety a Pascal's Mugging?

Channel:Robert Miles AI Safety · Length:13:41 · Published:2019-05-16 · Views:388587

Watch on YouTube

Summary

Robert Miles examines whether the case for AI safety is a Pascal's mugging, a tiny probability of an enormous payoff. He argues it is not, because AI risk does not depend on vanishingly small probabilities.

Why HN recommends it

HN points to it as a reasoned rebuttal to the claim that AI risk is just Pascal's wager.

the main point of the video is that kind of situation happens when the probability involved is vanishingly small — HN, 2026

AI safety, AGI & impact · Probability & game theory

Subtitles (excerpt)

Hi. Today we're going to talk about Pascal's wager and Pascal's mugging. This is necessarily going to touch on religious topics, pretty heavily actually, and I'm just going to say at the beginning of the video that, personally, I don't believe that any god or gods have ever existed, and I'm not going to pretend otherwise in this video, so be forewarned if that's likely to bother you. So, as you may know, Pascal was a 17th century philosopher who was interested in, amongst other things, the question of the existence of the Christian god. Various philosophers at the time were arguing that god didn't exist, and there was a lot of discussion going on about the various kinds of evidence for and against god in the world. But there's this thing that's quite common when people think about religious questions, where it feels sort of unsatisfying to talk about worldly evidence, as if you were considering some everyday question. There's a feeling that these supernatural concepts are very grand and mysterious, they're special, and so just straightforwardly considering the evidence for and against god is not the right way to do things. This is, of course, encouraged by religious thinking. The idea that some hypotheses aren't

On Hacker News

Here’s a video called “Is AI safety a Pascal’s Mugging?”: [VIDEO] From what I remember the main point of the video is that kind of situation happens when the probability involved is vanishingly small, and all the events you listed don’t have a vanishingly small probability, so they are not Pascal’s wager situations, just a normal rational safety concerns with particularly high consequences — circuit10, 2026-02
The difference to Pascal’s wager is that with that the probability is vanishingly small, but given current progress the probability of AI causing an existential threat is actually quite high. You can’t just call anything that has very bad negative consequences Pascal’s wager Video about this: “Is AI Safety a Pascal’s Mugging?” [VIDEO] — circuit10, 2023-08
All of this discussion really makes me think of Robert Miles "Is ai safety a Pascal's mugging?" from 4 years(!) ago[0]. All of this discussion has been had by Ai safety researchers for years in my layman understanding... Maybe we can look to them for insight in to these questions? [0] [VIDEO] — mptest, 2023-07
Pascal's wager is actually logically flawed. If there are two separate AGI invented, and you could only have chosen to invest in one, then where's the infinite capture of value? See this video on a reasonable argument about this: [VIDEO] — chii, 2019-05

Related videos