Hi Greg! I basically agree with all of this. But one natural worry about (e.g.) x-risk reduction is not that the undesirable event itself has negligible/Pascalian probability, but rather that the probability of making a difference with respect to that event is negligible/Pascalian. So I don’t think it fully settles things to observe that the risks people are working on are sufficiently probable to worry about, if one doesn’t think there’s any way to sufficiently reduce that probability. (For what it’s worth, I myself don’t think it’s reasonable to ignore the tiny reductions in probability that are available to us—I just think this slightly different worry may be more plausible than the one you are explicitly addressing.)
Hi Greg! I basically agree with all of this. But one natural worry about (e.g.) x-risk reduction is not that the undesirable event itself has negligible/Pascalian probability, but rather that the probability of making a difference with respect to that event is negligible/Pascalian. So I don’t think it fully settles things to observe that the risks people are working on are sufficiently probable to worry about, if one doesn’t think there’s any way to sufficiently reduce that probability. (For what it’s worth, I myself don’t think it’s reasonable to ignore the tiny reductions in probability that are available to us—I just think this slightly different worry may be more plausible than the one you are explicitly addressing.)