Think about fat tails and asymmetric risk. If you read in the newspaper next year: "Hundred thousand people dead/displaced due to power plant accident" - more likely to be coal or nuclear?
Coal kills slowly but is the power source that's killed most people over time. Nuclear has killed the fewest by far.
> Think about fat tails and asymmetric risk. If you read in the newspaper next year: "Hundred thousand people dead/displaced due to power plant accident" - more likely to be coal or nuclear?
What's your evaluation of the probability of this event ? Without putting figures in the equation, I could also ask you in the same way to think about this event : if you read in the newspaper next year: "Billions of people dead/displaced due to one in a thousands year drought, leading to global famine" - more likely to be coal or nuclear?
Nuclear. Naive probability matters relatively little here. The risk of coal plants due to CO2 is a thin-tailed long term risk. No single coal plant will ever cause a headline (or an event) like this. Thin-tailed long-term risks can be mitigated and planned for and the result is always a sum of many smaller choices (albeit global warming is an existential risk for sure!).
Nuclear is different. 99,999% of time it's cheaper and cleaner than coal. Doesn't matter. The tail-risk is what matters and that's also, why (conventional) nuclear can't ever be economically competitive against any other power source without direct or indirect subsidies (there is no insurance that will insure a nuclear power plant).
Highly recommend Taleb`s paper on fat-tailedness [1].
https://en.wikipedia.org/wiki/Harspr%C3%A5nget_hydroelectric...
No large chemical industry handlng dangerous chemicals, in developed countries or otherwise, has an insurance large enough to pay for huge expensive spills. That's why the company usually defaults when it happens and leaves the cleanup for the government later on. This has happened on numerous occations in the US, Europe, and over here in Sweden.