When Bill Gates is interested in something new, his organizing, capacious intelligence learns everything about it, and he imagines ways it could be better. Now the cofounder of Microsoft and the Bill and Melinda Gates Foundation is interested in energy. At his offices in Kirkland, WA, he spoke to Jason ­Pontin, Technology Review’s editor in chief. Gates called for energy “miracles” and a more rational energy policy, and he explained how being a software “fanatic” prepared him to invest in new ideas.

TR: The Gates Foundation has invested in solutions to big problems like infectious diseases in poor countries. Providing clean energy for the nine billion people the planet will hold in 2050 is a problem that’s civilizational in scale. What can philanthropy contribute to energy research?

Bill Gates: Well, basically not much. The energy market is an absolutely gigantic market, big enough that if you can come up with cheap ways of making electricity, then that should be done with typical big-firm risk taking, small-firm risk taking. On the other hand, the way capitalism works is that it systemically underfunds innovation, because the innovators can’t capture the full benefits [of their innovations]. But there’s a net benefit to society being more R&D-oriented. And that’s why in health research, governments do fund R&D.

TR: You are a member of the American Energy Innovation Council, which calls for a national energy policy that would increase U.S. investment in energy research every year from $5 billion to $16 billion. I was stunned that the U.S. government invests so little.

BG: I was stunned myself. The National Institutes of Health invest a bit more than $30 billion.

TR: So why couldn’t huge, regular, dependable investments from your foundation make a difference?

BG:We might have some involvement where it’s connected to things that wouldn’t happen for poor people otherwise. So there may be some particular biomass approaches for getting local energy out where there’s no roads and infrastructure, where there wouldn’t be a market signal for that type of innovation.

But [as an investor] I’ve put my money into Vinod Khosla’s venture fund. I’ve put money into Nathan Myhrvold [and his Intellectual Ventures fund]. Nathan has this thing that invents ideas broadly, many of which are energy-related. Some of those energy-related things will result in startups. One has so far: this amazing, wild nuclear [reactor design], TerraPower.

TR: If energy research is underfunded by $11 billion, what is a better approach to funding new energy technologies?

BG: It’s not a problem that lends itself to a Manhattan Project-type approach. It has to be low cost and usable in different circumstances. You can’t just get a bunch of smart people together and know which path they should go off and pursue. Actually, it’s amazing that that worked for the Manhattan Project.

TR: It worked because it had a very specific end: they wanted to build the biggest bomb in the world and end the war.

BG: They knew what they wanted to do. I guess in a vague sense we can say that we want energy that costs, say, a quarter of what coal electricity does and emits zero CO₂. We can write that down. But there are many paths to get there, each of which a realist would look at and say, “Wow, there’s a lot of difficult things along that path.” So I think it’s very important, both to give poor people cheap energy and to avoid hugely negative climate change, that the U.S. and other governments fund basic research. The irony is that if you actually look at the amount of money that’s been spent on feed-in tariffs and you properly account for it–tax credits, feed-in credits in Spain, solar photovoltaic stuff in Germany–the world has spent a massive amount of money which would have been far better spent on energy research.

TR: Let’s talk about policy, then. The prospects for a strong climate bill in the U.S. Congress now look dim. So do the chances for any binding international treaty. But almost everyone agrees that there needs to be a price on carbon or a tax.

BG: No, that’s not right. It’s ideal to have a carbon tax, not just a price on carbon, which is this fuzzy word that includes cap-and-trade. You’re using the tax to create a mode shift to a different form of energy generation. And then you just take all the carbon-emitting plants, you look at their lifetime, and you say on a certain date this one has to be shut down and when a new one is put in place, it has to be low-CO₂-emitting.

That’s a regulatory approach, and it’s very clear. Innovators are designing things for the power-plant buyers 10 years from now, who are looking at the regulatory and tax environment for the next 40 years. If you said to a utility company executive, which is more likely to stay in place: a cap-and-trade thing, whose price will vary all over the map, that will have some international things that will be shown to be a waste of money? Or a tax and a regulatory framework for plant replacement over the next 50 years? We should have a carbon tax. What we owe the developing world is this: we’re willing to pay high prices for energy plants above coal and drive prices down the curve so by the time they need to buy them, they don’t have to pay the high price.

TR: That sounds politically unlikely.

BG: Which is more likely: a carbon tax with all sorts of markets and options and uncertainties about prices, and traders in the middle, and confusion about who initially gets the most advantage? Or a regulatory thing and a 2 percent tax to fund the R&D so that utilities know they can buy a plant that’s emitting hardly any CO₂? Raising energy prices by 2 percent and sending it to R&D activities seems easier in a weak economy than raising them 20 percent. Now, 0 percent is the easiest option of them all, but unfortunately, that doesn’t get us the solution to this problem.

TR: You’re saying that meeting our energy needs will be both highly complicated and fraught with unknown problems.

BG: It is disappointing that some people have painted this problem as easy to solve. It’s not easy, and it’s bad for society if we think it is, because then funding for R&D doesn’t happen.

TR: You’ve talked about the need for “energy miracles.” But we’ve been waiting for such breakthroughs for decades. TerraPower is a traveling-wave reactor, a design that dates back to the 1950s.

BG: Well, no, we haven’t been working on those things. The nuclear industry was effectively shut down in the late ’70s. And so evolutionary improvements on those so-called Gen 3 designs really didn’t happen, and more radical things didn’t happen.

But let me get back to the main thrust of your question. The CO₂ problem is simple. Any amount you emit causes warming, because there’s about a 20 percent fraction that stays for over 10,000 years. So the problem is to get essentially to zero CO₂ emissions. And that’s a very hard problem, because you have sources like agriculture, rice, cows, and small sources out with the poorest people. So you better get the big sources: you better get rich-world transportation, rich-world electricity, and so on to get anywhere near your goal. If X or Y or Z gets you a 20 percent reduction in CO₂, then you’ve just got the planet, what, another three years? Congratulations! I mean, is that what we have in mind: to delay Armageddon for three years? Is that really it?

The U.S. uses, per person, over twice as much energy as most other rich countries. And so it’s easy to say we should cut energy use through better buildings and higher MPG and all sorts of things. But even in the most optimistic case, if the U.S. is cutting its energy intensity by a factor of two, to get to European or Japanese levels, the amount of increased energy needed by poor people during that time frame will mean that there’s never going to be a year where the world uses less energy. The only hope is less CO₂ per unit of energy. And no: there is no existing technology that at anywhere near economic levels gives us electricity with zero CO₂.

TR: Then what kinds of energy miracles do we need?

BG: Almost everything called renewable energy is intermittent. I have another term for it: “energy farming.” In fact, you need not just a storage miracle, you need a transmission miracle, because intermittent sources are not available in an efficient form in all locations. Now, energy factories, which are hydrocarbon and nuclear energy–those things are nice. You can put a roof on them if you get bad weather. But energy farming? Good luck to you! Unfortunately, conventional energy factories emit CO₂ and that is a very tough problem to solve, and there’s a huge disincentive to do research on it.

TR: You’ve said that nuclear energy has the best chance of being an energy miracle.

BG: Well, it’s the one I’ve gotten involved in. I spend time at TerraPower. I don’t claim to be the person who’s surveyed all the possibilities. I think solar thermal has a lot of promise. Solar chemical: some people see the possibilities at the research level. Algae: I’ve actually got some money in some of those [ventures]. Then there are crazy things like these high-wind kite guys. You really don’t want to rule anything out.

TR: Will TerraPower really build a traveling-wave reactor? And if so, where?

BG: We’re in discussions with basically everybody. TerraPower itself will not raise the money to build the reactor. We will partner with some mix of sovereign and private actors to get TP1, which is what we call our first reactor, and our dream is to build that by 2020. It’s more likely to be built in Asia than in North America or Europe. China’s the obvious one.

TR: TerraPower is far out.

BG: It’s very far out. It definitely needs to be categorized as a high-risk, wild thing, but the world only needs a few wild things to succeed. But you’ve got to get the pilot plant built, which is hard. You’ve got to have all the science and economics work the way they work on paper.

TR: How has being a philanthropist broadened you in a way that your career as a software entrepreneur did not?

BG: Believe me, when somebody’s in their entrepreneurial mode–being fanatical, inventing new things–the value they’re adding to the world is phenomenal. If they invent new technologies, that is an amazing thing. And they don’t even have to know how it’s going to help people. But it will: in education, medical research, you name it. So I was one of those fanatics in my 20s where I didn’t know about poor people. I worked night and day on software. I thought a lot about software. That’s a great mode to be in, but in my 30s I got exposure to management, although I was still writing some of the code. Then in my 40s, the majority of what I was doing was large-organization management and picking strategies, but I didn’t write any code that shipped in products. Now, in my 50s, I’m in a role that’s kind of like that.

The full transcript of the interview can be read here.