By Amory B. Lovins
Image courtesy of Klaus-Uwe Gerhardt / Pixabay
In her March 13 Bulletin article (“The war on Iran will speed the transition away from fossil fuels and toward nuclear energy, creating strategic challenges for the United States”), the distinguished Rachel Bronson’s equivocal but emphatic ode to nuclear power disappoints by omitting fundamentals.
A kilowatt of nuclear power capacity produces several times the annual output of a kilowatt of solar or wind capacity, but at many times higher cost per kilowatt-hour. Capital markets therefore shun nuclear investments but invest one or two orders of magnitude more in solar and windpower. Those renewables therefore add two orders of magnitude more net capacity per year than nuclear, which remains a less-than-one-percent contributor to global electricity growth—a trivial term but a huge distraction. Capturing most major countries’ policy and media through a skillful PR campaign conjuring a “nuclear renaissance” is no substitute for marketplace success.
China met its 2030 solar-and-windpower goal six years early, India five years early. The world added more renewable capacity in the first half of 2025 than it had added total nuclear capacity over decades. Batteries just got so cheap that three-fourths of India’s firm capacity additions are in the form of solar plus batteries.
In most of the world, solar or wind plus backup is the cheapest source of bulk power. That backup can be in the form of efficient and timely use, or seven other carbon-free grid-balancing alternatives including batteries, or fossil fuels (insecure, climate-harming, with volatile price), or nuclear power (proliferative, potentially dangerous, and the costliest and slowest option—so it worsens climate change by producing less electricity and displacing less fossil fuel per dollar or per year). So renewables have swept over 90 percent of the global market for new generating capacity. Nuclear gets only the sliver compulsorily funded by taxpayers.
Bronson also decries variable renewables’ so-called “storage challenge.” In fact, their grid integration is a solved problem: Many countries are mainly renewably-powered with superior economics and reliability, while nuclear power was lately the most intermittent generator in France. She claims AI can’t run on renewables; yet Apple’s data centers have done so for a dozen years, and last year in the United States a 100 percent-solar microgrid powering modular data centers was built in four months, at lower cost and with greater reliability than grid power.
She also overlooks the developing world’s extraordinary solar leapfrogs, from Pakistan to Vietnam to sub-Saharan Africa. But her biggest omissions are negawatts (which over several decades can use global electricity five times more efficiently) and flexiwatts (timely use, which can save 30 to 50-plus percent of peak demand). Competing or comparing those profitable demand-side resources with new supply can speed renewables, cut emissions and capital costs, avoid building pre-stranded supply assets, and improve equity, reliability, resilience, and security.
The war in Iran should indeed refocus us all on energy security. Buying fossil fuels from a different vendor only renames the problem. But do claims that the world should and will switch to nuclear power treat observed economics, actual competitive landscapes, and empirical installations and output? If not, they’re advocacy, not analysis—and rhetoric, not evidence.
Physicist Amory Lovins teaches energy efficiency at Stanford University, co-founded and chaired RMI (formerly known as the Rocky Mountain Institute), and advises major firms and governments worldwide.
This article was published on March 25, 2026 at the Bulletin of the Atomic Scientists.