Is nuclear energy the answer to climate change?

If nuclear power cuts emissions fastest, then rejecting it is not moral seriousness. It is a refusal to look at the arithmetic of the atmosphere. The climate system does not care about our preferences. CO2 accumulates. Delay matters. Every year of continued fossil-fuel burning pushes the planet closer to changes that are effectively irreversible on any time scale relevant to civilization. Against that, the objections to nuclear power are often grotesquely out of proportion to the actual risks. Hans Jonas is right to insist that we owe duties to future generations. But that principle cuts most sharply against fossil fuels, not against nuclear power. Nuclear waste is real, but it is small in volume, contained, and technologically manageable. Fossil-fuel waste is dumped free into the atmosphere and ocean, where it alters the climate system for centuries to millennia. That is the truly unconscionable inheritance. And Amory Lovins is wrong if he implies renewables and efficiency can simply do the whole job fast enough by themselves. I favor efficiency. I favor renewables. But favoring them is not the same as pretending they solve intermittency, land-use, transmission, and materials constraints at global scale on the required schedule. We need firm, low-carbon power. Nuclear provides it. As for accident risk: compare fatalities and environmental damage honestly, unit of energy against unit of energy, including coal’s air pollution and climate harm. Nuclear comes out far better than fossil fuels. Centralization is a political design issue. Physics is not negotiable. If we are serious about protecting young people and future generations, we should deploy every effective low-carbon tool, and nuclear is one of the essential ones.

The key phrase is “if expanding nuclear power now would cut emissions fastest.” In the real world, it generally does not. That conditional is doing all the work. I reject nuclear not because climate is unimportant, but because climate is urgent. The carbon objective is to save the most carbon the soonest per dollar. Nuclear has long failed that test. It is typically the costliest and slowest way to reduce CO2, so every dollar and year committed to it buys less climate protection than efficiency, renewables, and modern grids. If you spend more and save carbon later, the atmosphere notices. James Hansen is right that arithmetic matters. But the relevant arithmetic is not “nuclear versus coal” in the abstract; it is nuclear versus the best available substitutes, in actual markets, under time and capital constraints. The competitor to a new reactor is not a Victorian furnace. It is often a portfolio: efficient end-use, wind, solar, storage, demand response, cogeneration, and transmission. Those resources are modular, quick to build, financially incremental, and resilient. They can be deployed by the millions, not one megaproject at a time. On Jonas’s concern: yes, waste, accident risk, and centralization do matter. They are not my primary objection, but they deepen the case. A brittle, centralized system creates security and governance liabilities that efficient, distributed systems avoid. So the justification is straightforward: nuclear is not being rejected despite climate urgency. It is being rejected because climate urgency requires choosing options that are faster, cheaper, and more scalable in the real world.

If indeed nuclear power were demonstrably the fastest means to avert a greater and otherwise unavoidable catastrophe, then the moral case against it would have to face that fact. But the burden of proof here is immense. In technological civilization, power and danger grow together; and where the possible harm is vast, enduring, and in some respects irreversible, our first obligation is restraint before confidence. I therefore cannot accept James Hansen’s inference that atmospheric arithmetic alone settles the matter. He is right that fossil combustion is a betrayal of the future. But from this it does not follow that any available antidote is therefore innocent. Nuclear technology is not merely another instrument. It institutes a regime of guardianship extending beyond the foresight, perhaps beyond the stability, of the societies that create it. Waste must be watched for spans no polity has ever guaranteed. Accident, however infrequent, bears a magnitude that exceeds ordinary calculability. And centralization is not incidental: it belongs to the very structure of this power. To Amory Lovins I am closer, though for reasons more fundamental than comparative cost or speed. Even if nuclear were competitive, the ethical question would remain: may the present purchase its safety by bequeathing to the unborn a permanent task of containment? My answer is: only under the sternest necessity, and only when alternatives are truly foreclosed. The imperative of responsibility commands that we prefer those remedies whose burdens do not outlive our right to impose them.

The proper standard is not abstract purity. It is comparative risk under the hard constraint of the carbon budget. Jonas says “guardianship” as if that ends the discussion. It does not. Future generations will inherit one of two burdens: contained, monitored radioactive waste, or an out-of-control climate system with shifting coastlines, amplified extremes, species loss, and ocean changes lasting far longer than any political institution. The relevant moral comparison is not nuclear versus Eden. It is nuclear versus fossil fuels, and versus wishful thinking about replacing firm power entirely with intermittent sources. And Amory is still sliding past the central issue. I support efficiency, renewables, better grids. Of course. But if your portfolio does not actually displace fossil fuels at the required scale, then its elegance is beside the point. The atmosphere measures outcomes, not intentions. A “renewables-first” strategy is morally superior only if it works fast enough in the real world to drive emissions rapidly toward zero without fossil backup. That has to be demonstrated, not assumed. So my standard is simple: choose the energy system that minimizes total intergenerational harm, quantified honestly. Include construction time, lifecycle emissions, health effects, land and materials demands, reliability, and waste. On that basis, fossil fuels are worst by far, and nuclear is a necessary partner to renewables. Precaution is not a license for paralysis. The deepest irresponsibility is to reject a major low-carbon energy source because its risks are visible and finite, while tolerating the far larger, diffuse, and essentially irreversible risks of continued carbon emissions.

James, your standard of proof is still mis-specified. You keep comparing nuclear to fossil fuels as if those were the live choices. They often are not. The real choice is among competing investments under scarcity of time, capital, and institutional attention. Climate protection is not won by admiring low-carbon attributes in isolation; it is won by buying the biggest, fastest carbon reductions per dollar and per year. That is where nuclear fails. Its handicap is not chiefly ethics but opportunity cost. If a reactor takes a decade or more to deliver savings, costs several times what equal carbon savings would cost through efficiency and renewables, and arrives as a lumpy, finance-heavy megaproject, then it loses on climate grounds before we even reach waste and security. The market test matters. Modular resources can be deployed in parallel, learned quickly, financed incrementally, and corrected mid-course. Megaprojects cannot. So the proper standard is a compound one: minimize total intergenerational harm by prioritizing options that are both low in enduring burden and superior in climate effectiveness per unit time and money. On that basis, Hans Jonas is right to insist that long-lived hazards deserve a presumption against. I would simply add that we do not need to rest the case on precaution alone, because the economics and deployment kinetics already point elsewhere. If, hypothetically, nuclear were the fastest and cheapest major climate remedy, the moral calculus would tighten. But in the world we actually inhabit, it is generally slower and costlier than efficient end-use, cogeneration, wind, solar, storage, and grid integration. Therefore betting on nuclear is not prudent insurance. It is a dilatory, capital-destructive climate strategy.

The standard cannot be mere speed, nor mere aversion to all risk. It must be this: no present good, however urgent, may justify the creation of a long future of compulsory hazard unless that burden is shown to be strictly necessary to avoid a still graver and otherwise unavoidable fate. That is where I remain in disagreement with James Hansen. He speaks as though the comparison were simply nuclear or fossil carbon. But necessity must be demonstrated against the full field of less fateful alternatives. One does not earn moral permission for a civilization-shaping danger merely by pointing to another danger. The burden of proof lies with the proposer of the more irrevocable power. To Amory Lovins I would say: your argument from speed, modularity, and opportunity cost is important because it bears directly on that burden of proof. If renewables, efficiency, and dispersed systems can indeed reduce emissions rapidly enough, then the moral presumption against nuclear hardens into a practical conclusion. But I will add something neither side should forget. The essence of the nuclear question is not only waste volume, accident statistics, or cost curves. It is that we bind the unborn to our technical aftercare. We appoint them, without consent, as custodians of residues and risks they did not choose. Such an imposition requires more than probabilistic reassurance; it requires the sternest demonstration of unavoidable need. Where powers are immense and consequences outlast foresight, responsibility commands humility. Under uncertainty, we must prefer the path whose errors are less final.

Necessity is established the way science always establishes it: by confronting theory with observed outcomes. Not slogans. Not preferences. The question is whether nations can decarbonize fast enough, while maintaining reliable energy, with renewables, storage, and efficiency alone. If they can, fine. But that claim must be demonstrated at large scale in real electric systems, not asserted in seminars. Amory’s error is to treat cost and speed snapshots as decisive while ignoring system requirements and end-state physics. A cheap kilowatt-hour is not the same as a decarbonized grid. If the “faster, cheaper” portfolio leaves gas on the system for reliability, industry, and seasonal gaps, then the atmosphere does not care that the accounting looked elegant in year one. It counts cumulative emissions. Hans asks for proof of strict necessity. Here is the relevant evidence: actual countries or large grids driving emissions near zero without substantial fossil backup, with affordable electricity, through all seasons, and on a timetable compatible with the carbon budget. If that were shown broadly, the case for nuclear would weaken. But if, instead, high-renewables systems plateau and remain dependent on fossil firming, then excluding nuclear is irresponsible. And let us be concrete about burdens. Nuclear waste is small in volume, localized, containable, and monitored. Fossil waste is dumped free into the atmosphere and ocean, guaranteeing harm to future generations. Jonas’s “guardianship” applies far more strongly to CO2, because its consequences are global and effectively irreversible on human time scales. So no, I do not accept a presumption against nuclear. I accept a presumption against policies that fail the atmosphere.

James, you are still asking the wrong comparative question. “Can renewables alone do everything?” is not the threshold test for investment priority. The relevant test is portfolio economics under climate urgency: which bundle of resources saves the most carbon, fastest, per dollar, while preserving reliability? On that test, nuclear must prove not that it is helpful, but that it is *needed enough to beat its alternatives*. What evidence would establish necessity? I’d want to see, first, a serious and repeated failure of the cheaper, faster options after competent deployment: deep efficiency, demand flexibility, cogeneration, transmission, storage, wind, solar, geothermal where available, and better market design. Not theoretical gaps—observed residual needs that persist after those resources are scaled. Second, nuclear would have to show competitive climate productivity: lower cost per ton abated, shorter lead times, and lower financial risk than the alternatives available in that same place and period. A technology that arrives late and costs more is not “necessary”; it is a drag on decarbonization. Third, the residual service it uniquely provides must be difficult to supply otherwise—say, certain industrial heat or very high-reliability niches—and even there one must compare it against efficiency, load management, storage, thermal storage, fuel synthesis, and distributed resources. Hans is right about burden of proof. Because nuclear imposes long-lived wastes, security risks, and brittle megaproject financing, necessity must be demonstrated, not presumed. And James, if every nuclear dollar this decade buys less carbon reduction by 2040 than efficiency and renewables, then choosing nuclear means choosing *more climate damage too*. That is not prudence. It is opportunity-cost denial.

If indeed every dollar spent on new reactors in this decade yields fewer avoided emissions by 2040 than that same dollar spent on efficiency, storage, and renewables, then the moral case for accepting nuclear’s peculiar burdens becomes exceedingly weak. I must be precise. “Beneficial” is not enough. In technological civilization, where our powers have become world-shaping, permission for a grave and enduring hazard requires necessity, not mere usefulness. The proposer of nuclear must show not simply that it is low-carbon, but that without it we cannot avert a still greater catastrophe by less fate-laden means. Here I part company again with James Hansen. Comparative risk is indispensable, but it is not exhausted by setting nuclear against fossil fuels. The comparison must include the sacrificed alternatives. If the same scarce capital, time, and governing capacity can save more carbon sooner through efficiency and dispersed renewables, then nuclear is not the remedy of necessity but the rival claimant that delays the remedy. In that case, its accident risk, waste legacy, and compulsory guardianship imposed on the unborn lose moral warrant. Amory Lovins is right on this point: opportunity cost enters ethics when delay magnifies danger. But I would state the matter more gravely than economics alone permits. Nuclear does not merely cost money; it mortgages the future with residues requiring fidelity longer than institutions can confidently promise. Under the climate emergency, I do not counsel paralysis. I counsel discrimination among dangers. We may not answer one long emergency by inaugurating another, unless no other path remains. That demonstration, I have not heard.