Discussion about this post

User's avatar
Raymond Coxe's avatar

Bob - thanks for a thoughtful and "grown up" piece. As a former nuclear engineer (all the way through the doctoral process), it's been interesting to see another resurgence in interest of 'advanced' nuclear designs that were 20 years old in the mid to late 1980s.

Your article hits the bullseye on an often-underappreciated fact about new energy technologies (and new commodity technologies in general) - it's an incredibly rare technology that emerges from the laboratory that is immediately and widely embraced and deployed, even if the technology is clear superior to what existed before it.

Commercial electricity is a good example - fully supplanting gas lamps, water mills, and all manner of human labor with electricity driven tools took multiple years for a single region of the US, and multiple decades for the entire US. Crucially, that transition required new commercial and regulatory innovations (e.g., a public utility regulatory body that was chartered with regulating electricity companies, sorting out how land could compulsively acquired if needed for a power plant or transmission line, etc.).

Later, the development of technical feasible renewables that needed some policy support elicited the same need for commercial and regulatory innovations - RPSs, regulatory-approved resource plans that included renewable mandates, state and federal tax and production incentives, etc. All of these non-technical innovations contributed to building the order book for renewables, which in turn drove down their costs.

Advanced nuclear will be no different. Commercial and regulatory innovations (e.g., finally developing an implementable long-term waste disposal process, or supporting supply chains for high assay, low enriched uranium) will be necessary if advanced (or conventional) nuclear is to meaningfully contribute to a clean energy grid worldwide.

You correctly point out that designing those commercial and regulatory innovations will take time and effort, and should only be pursued if a clear-eyed view of *everything* that is required to get nuclear to "work for us" concludes that the time, money and political capital to be spent is worthwhile.

All too often, new technologies clamor for policy support too early, and with optimistic (perhaps wildly so) projections of technical advances and cost reductions. For example, my personal opinion is that data center demand for AI is fueled by an optimism that may be justified ultimately, but seems a bit premature (except perhaps in the area of writing code). Just as we don't need dozens (if not hundreds) of data centers that fall silent in 10 years, we should match the pace of investments in any nascent energy technology (green cement and steel, advanced nuclear, etc.) to the pace of realistic understandings of "how much impact, how fast".

Thanks for kicking off the conversation.

Alan Nogee's avatar

Nice piece, Bob. One under-appreciated ironic constraint: nuclear will be competing heavily with the data data centers they are being proposed to serve for very scarce skilled electricians and more, driving up the costs of both, as well as of “electrifying everything.”

“A dire electrician shortage is a ‘life or death’ threat to the AI data center boom…,” https://fortune.com/2026/03/02/ai-data-centers-electrician-shortage-gen-z-training-careers/

“All the nuclear workers are building data centers now,” https://heatmap.news/energy/data-centers-labor

1 more comment...

No posts

Ready for more?