It does surprise me -- HN is a really skeptical audience, or at least has a highly skeptical subset. Think about how bearish HN is on autonomous vehicles: sure, there's a group who worship Elon, and believe that autonomous vehicles are perpetually on the brink of mass adoption. But most autonomous vehicle discussions come with a HUGE grain of salt, with a lot of folks pointing out edge cases that will prevent mass adoption for the forseeable future.

I think nuclear is just so complex it's too hard/easy to make viable arguments in either direction, and things just boil down to "danger" and "cost". My skeptical bone points out that implementation is always messy compared to theory, so we should expect building flaws/design compromises/flawed processes to result in failures at some small percent of nuclear facilities. And given how catastrophic those failures can be, it's hard to justify it. But the climate side of the argument is equally scary: if we don't switch away from carbon-based power sources, massive areas of our world are going to become uninhabitable, underwater, or dramatically change in ways that destroy life as we know it. So maybe that small fraction of failures is acceptable? And in all fairness, the major nuclear failures I can think of -- Fukushima, Chernobyl -- came with comparatively few deaths, and a moderate cost of now-uninhabitable land. Perhaps less than climate change will come with over the coming decades.