I used to feel this way: that large nukes are the answer because they provide megacity-scale energy for an increasingly urban world. But as I've studied the nuc

I used to feel this way: that large nukes are the answer because they provide megacity-scale energy for an increasingly urban world. But as I've studied the nuclear industry a few things have eroded away at this belief. First, iteration. You can iterate on the design and supply chain of a small niche nuclear reactor much faster than a large baseload one. And you can do so without bankrupting the likes of Westinghouse or Areva. Once you figure out a new design at small scale, you can scale up. We did with with the current reactor designs and we should do it again for the newer fancy ones. I think this is what Oklo and similar reactor companies are planning on. Second, and more novel, is that you can imagine large-scale nuclear farms of small reactors. Build small reactors that are sized to handle their own decay heat without fancy safety systems and manufacture them in a factory. Put them out there on shared security infrastructure and siting. There's more energy produced from internal combustion engines than all baseload plants in China, for instance. "You want scale? I'll give you scale!" And finally if you still need to go big, there's also the potential of large modular reactors. If you build huge power plants in shipyards on floating platforms you can get economies of scale AND economies of mass production, while improving safety by being intimately coupled to the ultimate heat sink: the ocean. Wild idea politically but technically very intriguing. Would require much more remote operation than current plants because a crew of nuclear rough necks would have high salaries. But totally doable, and could probably decarbonize the world very rapidly.