If you discount small probabilities to zero, should you still be a longtermist? Yes! Or so I argue in an interview I gave to Future Matters-newsletter.
I might be misunderstanding, but I still think probability discounting is plausible. For instance, if I hadn't adopted probability discounting when I was a child, I would probably have expended a lot of effort into not moving because of my belief that accidentally breaking an atom would cause a nuclear explosion. To me, it seems humans are quite bad at assigning probabilities, and at a sufficiently low probability it's a good heuristic to simply discount it entirely.
I might be misunderstanding, but I still think probability discounting is plausible. For instance, if I hadn't adopted probability discounting when I was a child, I would probably have expended a lot of effort into not moving because of my belief that accidentally breaking an atom would cause a nuclear explosion. To me, it seems humans are quite bad at assigning probabilities, and at a sufficiently low probability it's a good heuristic to simply discount it entirely.