[Ericson2314]

The ideal case is that SMRs are not the end goal, but a way to rebuild the supply chain. As soon as we have SMRs in prod, rather than building more of them, we should attempt to increase the size of deployments with minimal falling back on in-situ construction.

SMRs take the supply chain metaphor a bit too literally: we do need practice but assembling prefabbed parts at a larger scale is fine too. There is a spectrum of options and we just need to avoid "special snowflake boondoggles".

Given the US's fucked NIMBY culture, it well may be that SMRs are the best route despite these inefficiencies. Just don't expect the "solar model" where we just shit out lots of lousy product and that's it.