The News
AI and data center companies can push down the costs of climate tech by leveraging their power as massive buyers of clean energy and industrial materials — and they should, Bill Gates told Semafor.
Gates, arguably the world’s top individual investor in climate tech, broke ground this month on the first next-generation nuclear power plant in the US. He has since pledged to invest “billions more” in plants like it, which he sees as the most promising way to supply the upcoming boom in power demand for data centers and EVs.
But there’s a key hurdle that the new plant, in Wyoming — and others to come — needs to overcome: It has to produce cost-competitive electricity, a prerequisite for widespread adoption. Gates is confident that it will. But much of his broader climate tech portfolio has yet to pass that test. The next generation of climate tech is critical for emissions that remain unaddressed. What’s holding it back, he said, is that there still aren’t enough early buyers willing to stomach the “green premium” until it can gain enough scale.
In this article:
Tim’s view
There are some problems in the energy transition even pockets as deep as Gates’ can’t solve.
The lesson he takes away from a decade investing in climate tech is that the “green premium” — how much more expensive the low-carbon option is compared to conventional power — is the only data point that really matters. Few consumers and businesses are willing to pay a green premium no matter how much they might care about the environment. If that premium can be eliminated in key technologies, Gates believes, it will be a tipping point on the way to solving climate change; if we don’t, climate will be much more difficult to solve.
The crucial first-line climate technologies — wind, solar, EVs — have either already erased their green premium or are poised to. But for the next, harder round of emissions from industrial manufacturing, transportation, and aviation, the current landscape of government incentives and private startup investment doesn’t get us there fast enough, Gates said. To reach the scale at which they can be cost-competitive, they need to be able to draw investment not just from the niche climate crowd but from mainstream Wall Street, what Gates calls “normal money.” For that, they need to be viable as businesses, and that means they need customers, not just investors. It’s a chicken-and-egg problem that tech companies, which are putting unprecedented demands on the grid and have the resources to pay more now for green power and construction materials, can be “very constructive” in solving, Gates said.
Since Gates launched his climate tech investment vehicle, Breakthrough Energy, at the landmark 2015 Paris climate summit, it has invested $2.2 billion to build one of the most impressive stables of companies in climate tech, including green steel producer Boston Metal, electrolyzer manufacturer Electric Hydrogen, geothermal power producer Fervo Energy, battery recycler Redwood Materials, and the carbon removal startup Heirloom. The group is bagging eye-popping fundraising rounds, including $1 billion last year for Redwood, building its first commercial-scale facilities, and locking in early sales contracts, thanks largely to Breakthrough’s patience, willingness to write big checks for risky first-of-a-kind projects, and long-term support for companies from inception to maturity. Breakthrough has about another $2.3 billion standing by for future investments.
“When we started, green investing was really weak and kind of out of fashion. The venture model really didn’t fit, given the capital intensity and timeframes involved,” Gates said. “But since then it has gone better than I could have expected.”
Most of the Breakthrough companies are surviving having a green premium, for now, by relying on various forms of taxpayer support and a small group of early buyers. But for Gates, who has spent most of his philanthropic career focused on public health crises in developing countries, that’s a major limitation on how much climate tech can get adopted outside the US and Europe. “Middle-income countries are 60% of emissions, and they will not adopt high-premium stuff,” he said.
Tech companies can’t solve the green premium issue alone. Another part of the solution — popular among the conservative-leaning climate lobbying groups that have received funding from the Breakthrough Energy Foundation — is a price on carbon. But that will always be a tough sell in the US. So Gates’ key question — “who’s going to do the ‘bootstrap’?” — still doesn’t have a decisive answer.
Q&A
This conversation has been lightly edited for clarity and length.
Tim McDonnell: What’s the most important thing you’ve learned in the last decade of running Breakthrough Energy?
Bill Gates: We definitely have a political atmosphere where you can’t go to voters and say, ‘Pay more for this green product.’ The Breakthrough theory of change, that middle-income countries and even rich countries aren’t willing to bear green premiums, is, sadly, being borne out. So you hear [policymakers say], ‘you must buy a heat pump,’ or, ‘here’s the deadline for internal combustion cars.’ But it’s a collective action problem. To get these technologies to what I call ‘zero green premium,’ requires either tax credits or project financing or both. These companies are doing well, but our current trajectory doesn’t get you anywhere near climate goals, not even close. You really have to see things like steel or cement or airplanes, which are very hard, come out of the lab, scale up, and get the cost down. But who is going to be a customer for the original green steel or cement, when the product coming out of the lab [is more expensive]?
TM: If avoiding green premiums is so important, what do you think about the Biden administration’s tariffs on clean tech imports from China?
BG: There’s no doubt that when you put a tariff on, say, solar panels from China, you’re creating somewhat of an impediment to climate progress. But ideally, we get the [other costs of solar project development] down so much that even if costs are slightly higher for the panel, it’s not a very big deal. The eventual view is that panel costs will be super low. But with China, whether it’s on EVs or batteries or panels, the politicians are making a choice, and you can often slow down your climate progress by prioritizing industrial goals.
TM: Closer to home, now that you’ve broken ground on the TerraPower plant, can you say more about where you see advanced nuclear going in the near future?
BG: We love wind and solar. But people underestimate and are sometimes naive about how hard it is to maintain reliability. So you need a baseload form of green energy. But we have a lot of work to do over the next six years to get fully certified and prove that our cost of electricity is super low. We’re not asking ratepayers to take that risk, which in the past was not a good experience for them. The risk is entirely held by private investors and the federal government.
Fission doesn’t have any science problems, it’s just got cost and engineering problems. So ideally we’ll have five or six of these reactors built by 2030 in the US, and then we would go to other pro-nuclear venues like the UK, France, Japan, or South Korea. I did TerraPower not because it’s easy to make money in nuclear fission. I did it because of the potential climate benefits. And I hope we eventually get competition between fission and nuclear fusion. At times people see both as a dead end, and that’s not what I see.
TM: Having led a tech company, what’s the responsibility of tech companies to mitigate, from their side, the potential climate consequences of the AI and data center boom?
BG: The tech companies need to be legitimate buyers of green electricity, making sure that they’re not just taking green electricity and making other people buy more dirty electricity. You have to be careful to get that right. The US electricity system is not great at dealing with large increases in demand. And so we are just going to be behind. But [the tech companies] don’t want to do it in an illegitimate way, and I think they will be very constructive as buyers for clean energy and other technologies. When they’re building buildings, are they willing to buy green steel and green cement?
TM: When you think about the decade ahead, what’s something you think will be very different about our lives in 10 years that most people aren’t expecting now?
BG: The application of AI to everything is the biggest change agent. People say, these digital guys overhype everything. I claim we’re not this time. AI is accelerating innovation, it’s not some stupid mania thing.
I hope a decade from now we can say that in every area of emissions we’re either at a zero green premium or we have a clear path there. But even when you get to zero green premium, you have to come up with the capital base to allow all the steel plants and cement plants to make the transition.
I hope in 10 years that fission and fusion are clear tools to solve this problem. I’d hope in fission we’d have 30 or 40 [advanced] reactors going. I hope geologic hydrogen has worked out. And from an investment point of view we need to show that [climate tech] is very attractive. You hear these gigantic [investment] numbers for what it will take [to decarbonize the economy]. That’s mostly project finance, and once you prove you can build reactor 10, then that’s low-risk, gigantic-volume kind of money. In some sectors like EVs or solar we are starting to tap into that ‘normal’ money. But in many sectors like geothermal or nuclear, they appropriately still say, ‘you’re just too risky for normal money.’