Positive feedback loop isn't the only risk of nuclear power. Fukushima had a negative void coefficient too, right? Rather than pretending there's negligible risk, I'd rather say it's there but the alternatives are worse.