AMA: Geoengineering, nuclear, power prices, and more

The geoengineering answer turns on a paradox Kann cannot resolve. Injecting aerosols into the stratosphere to cool the planet is extraordinarily cheap per degree of cooling, orders of magnitude cheaper than carbon removal on a dollars per degree Celsius basis, and that is exactly what makes it a poor venture business: if a couple of billion dollars buys a meaningful fraction of a degree, the market is small even in the world where it works. His conclusion is that if we pass some point of no return the world probably should do this, and that the doing belongs to governments or a wealthy individual rather than a startup. Pierpoint agrees, adds the unresolved question of what other environmental effects any given method would have, and raises the second obstacle, public perception: a startup releasing sea salt particles in Alameda was effectively shut down by the city council, and she argues the community engagement burden in this category is astronomical. Kann’s two reference points for how a stigmatized category behaves are direct air capture, where the pioneers spent a decade in the wilderness before the market came to them, and Make Sunsets, which he describes as doing it in the open in defiance of the rules and calls clearly the wrong approach, while admitting he is not sure what the right one is.

Electricity prices get the most developed answer, and it is a causal chain rather than an opinion. Transmission and distribution costs have been rising for years, but cheap natural gas largely offset them. That offset is now failing from several directions at once: new generation of every kind costs more, gas turbines conspicuously; tariffs and general inflation push the same way; incentives have just been removed from solar and wind, the cheapest sources per kilowatt-hour, making them more expensive; and load growth forces transmission and distribution investment that gets rate-based and passed through to customers. Kann treats California, where he thinks prices are high partly for local reasons such as wildfire costs, as a harbinger rather than an exception. The consequence he calls unshakeable is that electrification is a trade, more capital upfront for lower running costs, so its value proposition erodes as electricity gets expensive relative to gas. Policy can soften that with rebates, and products can be made flexible enough to run off-peak, but neither reverses the arithmetic once buyers move past the early adopters. Pierpoint’s own all-electric heating and cooling conversion roughly doubled what she had been paying on gas, and she notes most people are price motivated and not like her. Asked what would bring prices down, Kann nominates energy efficiency, the unloved workhorse he thinks may come roaring back precisely because bills are about to hurt. Pierpoint adds demand flexibility and anything that lowers the cost of building infrastructure, and reframes the question away from ranking clean energy sectors toward finding leverage points, her example being that cheaper wire-laying would unlock vehicle charging.

Asked what he has been wrong about, Kann gives one error in each direction, and the pair is more interesting than either alone. He was too optimistic about the Inflation Reduction Act surviving, having reasoned that a rollback required Republicans to take the presidency, Senate and House and then choose to spend capital attacking tax credits, which the US has historically let expire rather than repeal. That is what happened, and what he says he underappreciated was how much a small number of hardliners could dictate terms once it was folded into a larger budget bill, though he notes the result was uneven, some sectors fine and others hit harder. He was too pessimistic about nuclear for a structurally similar reason: he had counted the preconditions, regulatory change at the Nuclear Regulatory Commission, public support, companies clearing approvals, financing appearing, and treated them as independent long shots. They turned out to be correlated, so quiet bipartisan momentum plus the AI load growth wave moved them together. He is now optimistic, with the caveat that he is more realistic about the timeframe than many. Two gates remain open: the cost and process of regulatory approval, and who absorbs cost overruns on the first large projects, which he says the industry is deferring rather than answering. It could be the federal loan programs office or it could be private capital, but somebody has to hold it, and even a 500 megawatt small modular reactor is a lot of money. Power purchase agreements at prices he hopes are sufficient to make project economics work are now available, which he says was not true five years ago; necessary, not sufficient.

On where climate tech sits in its cycle, the answer is US-specific and blunt: the bust, and a steeper, faster fall than last time because this one was triggered by a political change rather than by accumulated failures. What makes it different is that a separate energy supercycle arrived simultaneously. Energy became, in his framing, about the most important sector in the world next to chipmaking, which cuts both ways since it also builds a lot of gas generation, but it means real customers with real willingness to pay and to try new things. So parts of the sector are better off for having the bubble burst quickly just as a tailwind arrived, and parts get no tailwind at all. The durable lesson he draws is that being clean is not a value proposition: “the same thing but clean,” and especially “the same thing but clean and more expensive,” lacks durable customer demand outside subsectors with policy support, and he thinks that fallacy explains most of the companies that had the rug pulled out from under them.