The media is currently obsessed with a "nuclear revival" narrative. They point to the 40-year anniversary of Chernobyl as a turning point where the world finally regained its nerve. They talk about Small Modular Reactors (SMRs) like they’re iPhones coming off an assembly line. They cite the sudden hunger of Big Tech data centers as the ultimate catalyst.
It’s a fantasy.
The industry isn't having a revival; it’s having a PR moment funded by venture capital and desperation. If you look at the cold, hard physics of project finance and the actual output of the global grid, the "renaissance" looks less like a rebirth and more like a high-stakes bailout for an aging giant that still doesn't know how to build anything on time.
The Myth of the Small Modular Reactor
The hottest trend in energy circles is the Small Modular Reactor (SMR). The logic sounds seductive: instead of massive, bespoke billion-dollar concrete monoliths, we’ll build tiny reactors in factories and ship them to site. Economies of scale will take over. Costs will plummet.
Except they won’t.
Nuclear energy is governed by the Square-Cube Law. This isn't a suggestion; it’s geometry. As you shrink a reactor, the surface area (where heat escapes and where you need shielding) doesn't shrink as fast as the volume (where the power is generated). This means SMRs inherently lose the efficiency of scale that made traditional nuclear even remotely competitive in the 20th century.
I’ve sat in rooms with energy analysts who treat SMRs as a "plug-and-play" solution. They ignore the reality of NuScale Power—the poster child for SMRs—which saw its flagship project in Utah collapse in 2023 because the target price of power jumped from $58 per megawatt-hour to $89. That's a 53% increase before they even broke ground.
We are trying to solve a manufacturing problem with a technology that requires the most stringent, slow-moving regulatory oversight on the planet. You cannot "move fast and break things" when you’re dealing with fissile material. The "modular" part of SMR is a marketing term; the "reactor" part is a regulatory nightmare that kills the margin every single time.
Big Tech Won’t Save the Grid
The latest branch of the revival narrative involves Microsoft, Amazon, and Google. Microsoft’s deal to restart Three Mile Island Unit 1 (rebranded as the Crane Clean Energy Center) is being hailed as the dawn of a new era.
It’s actually a symptom of a broken market.
Microsoft isn't buying nuclear because it’s the most efficient way to power a data center. They are buying it because they have "Net Zero" commitments that are physically impossible to meet with wind and solar alone due to intermittency. They are willing to pay a massive premium—essentially a "green tax"—to claim 24/7 carbon-free energy.
But here is what the headlines miss: Three Mile Island Unit 1 was a working reactor that only shut down in 2019 because it couldn't compete with cheap natural gas. Restarting an existing plant is a far cry from building a new one. The "revival" isn't creating new capacity; it’s cannibalizing existing assets at inflated prices.
When you look at the Levelized Cost of Energy (LCOE), the gap is staggering. According to Lazard’s 2024 analysis, the unsubsidized LCOE for new utility-scale solar is roughly $29–$92/MWh. For new nuclear? You’re looking at $141–$221/MWh.
If you think the "industry insider" secret is that nuclear is cheap, you’ve been reading the wrong reports. The secret is that nuclear is an insurance policy for billionaires who can’t afford a 10-minute blackout. For the average ratepayer, it’s a predatory price hike.
The China Distraction
Proponents love to point at China. "China is building 22 reactors right now!" they scream. "Why can’t we?"
Because we don't live in a command economy where the state can socialize every risk and ignore the cost of capital. In the West, nuclear projects are crushed by interest rates.
Imagine a scenario where you want to build a $10 billion reactor. It takes 15 years to build (the current average for Western projects like Vogtle 3 and 4 in Georgia). For those 15 years, you are paying interest on $10 billion without earning a single cent in revenue. By the time the plant turns on, the debt service alone makes the electricity twice as expensive as the market rate.
$$Cost_{Total} = \sum_{t=0}^{n} \frac{C_t + O_t}{(1 + r)^t}$$
In the equation for the Present Value of Costs, that $r$ (the discount rate) is the nuclear killer. China "wins" at nuclear because their $r$ is effectively zero or subsidized by the state. In a Western capital market, nuclear is a suicide mission for a CFO.
The Safety Obsession is the Real Problem
Here is a take that will upset both sides: Nuclear is too safe.
We have regulated nuclear energy into a corner where the "As Low As Reasonably Achievable" (ALARA) principle has become a mandate for infinite spending. We spend millions to prevent a statistical anomaly of a radiation leak while ignoring the fact that coal power kills more people every single day through particulate matter than the entire history of nuclear power combined.
But you can’t "deregulate" nuclear in a democracy. The public won’t allow it. Therefore, the cost will never go down. We are stuck in a cycle where every safety improvement makes the technology more expensive, which makes it less viable, which leads to fewer plants being built, which means we never get the "learning curve" benefits seen in the solar industry.
Solar gets cheaper because we make billions of panels. Nuclear stays expensive because we make one reactor every decade and treat it like a unique piece of jewelry.
Stop Asking if Nuclear is Green
The question "is nuclear green?" is a distraction. Of course it’s green. It’s the densest, most reliable form of clean energy we have.
The real question is: "Is nuclear investable?"
The answer, for almost everyone not named Bill Gates or the US Department of Energy, is no. The current "revival" is a collection of government subsidies and tech-bro vanity projects. It is not a sustainable market movement.
We are seeing a desperate attempt to extend the life of 40-year-old plants because we’ve realized we can’t build new ones. Calling that a "revival" is like calling a heart transplant a "youth movement." It’s an emergency measure, not a strategy.
The Actionable Truth
If you are an investor or a policymaker, stop waiting for the nuclear miracle.
- Bet on Geothermal: If you want 24/7 baseload power, enhanced geothermal uses the same drilling tech we perfected in the fracking boom but without the 15-year regulatory lead time of a reactor.
- Prioritize Long-Duration Storage: The money being funneled into SMR startups would be better spent on iron-air batteries or pumped hydro. Solving the storage problem makes the "intermittency" argument against solar irrelevant.
- Accept the Decline: We will keep a few legacy plants running to keep the lights on in Silicon Valley, but the dream of a nuclear-powered world died in the 1970s under a mountain of compound interest and concrete.
The hype cycle will continue. The press releases will keep flowing. But the meter doesn't lie. Nuclear is the most expensive way to boil water ever devised, and no amount of "renaissance" branding will change the physics of the balance sheet.
Get out of the way of the sun.
