In short: I'm against it, but with a big "but." Today, nuclear power is used as an argument the same way "what about the troops?" is used — a joker card to elegantly exit any discussion about pollution from energy production. Nuclear doesn't emit CO2, the new reactors eat all the waste, and everything seems solved. But it's a shallow argument that ignores practically everything.
Nuclear isn't zero-emission
Nuclear doesn't emit CO2 during energy production. But everything else does. Building a nuclear plant emits CO2 at torrential rates: these are multi-billion-euro structures, complex, requiring special materials. And uranium and plutonium don't show up enriched at your doorstep — there's an entire supply chain to consider. Which, by some estimates, is on balance more ecological than oil and coal. Fair enough.
Fourth-generation reactors don't exist
The supposedly brilliant reactors everyone talks about are fourth-generation ones. They don't exist. Not one has been built, and they won't appear before 2030 at the earliest. They should produce less waste and be safer — all beautiful but entirely theoretical.
We don't even know the economics of these reactors. What if we spend a decade building plants, invest a fortune, and then discover unaccounted-for variables before recouping the investment? Meanwhile, we still don't know how to dispose of current waste. In Germany, water is seeping into containment chambers. In Italy, nobody knows where to put it, with regions fighting endlessly about locations.
Money is always scarce
If we allocate money to nuclear, we must take it from renewable energy investments — which are already cheaper than coal today. Renewables are the future, and investment must remain high. Countries like Costa Rica already run entirely on sustainable energy.
As someone who loves finance and works as a project manager accustomed to managing budgets, this point is extremely important to me. Money is always tight, and it makes a real difference where you direct it. Starving renewables to fund nuclear reactors made entirely of unknowns is simply wrong from a rational decision-making perspective.
The black swan: the argument nobody understands
This is my favorite argument, because it's the one never understood by 90% of engineers, politicians, and technicians. People keep thinking Fukushima won't happen again, that it was a rare case like Chernobyl, like Three Mile Island. According to standard statistics, these catastrophes should occur once every million years. Yet they've occurred roughly once every decade.
The explanation lies in safety assessment models: it's not that statistics are wrong — it's that the wrong statistical model is being used. Disasters fall into the category of "unknown unknowns" — what we don't know we don't know. Humans not only don't know many things, they don't know they don't know them. And safety models, built by humans, reflect these epistemological limits.
The difference from other accidents is scale. A car crash kills a few people. A bridge collapse, a few dozen. A nuclear plant explosion? Apocalyptic damage. The realistic estimate of Chernobyl's toll is around 500,000 people with cancer, with effects persisting to this day.
My "but"
I'd be in favor of fourth-generation nuclear plants given: perfectly described and plausible economics, rapid construction timelines, small plants to limit damage in case of a black swan event, and a development plan that still prioritizes renewables. And since China and Europe are experimenting with nuclear fusion — the clean energy of the stars — if I must invest in atomic energy, wouldn't it make more sense to bet directly on fusion?
