It seems to me like he argues that scaling nuclear is mostly known (but expensive) problem.

And that scaling renewables is largely unexplored, and very likely similarly expensive problem.