Bill Gates on why the ‘miracles’ of solar and wind energy won’t save us from climate change—and the breakthroughs that just might

A conversation with the Microsoft cofounder and famed philanthropist about the genius of cement, the thrill of the Impossible Whopper, and the book it took him 10 years to write.

This article is part of Fortune’Blueprint for a climate breakthrough package, guest edited by Bill Gates.

Subscribe to unlock this article and get full access to

This article is part of Fortune’Blueprint for a climate breakthrough package, guest edited by Bill Gates.

More than a decade in the works, Bill Gates’ new book, How to Avoid a Climate Disaster, hits shelves today. The “how to” part is anything but easy. But the clarity of Gates’s plan—and the reason for absolute urgency—may well turn millions of readers into overnight activists. (It should.) At Fortune, we wanted to dig deeper, exploring the challenges that Gates raises and a bunch of others as well. To that end, we’ve asked the famous philanthropist to serve as “guest editor” of Fortune for the day. Please see our lineup of stories and interviews here, which includes an exclusive essay from Gates you’ll want to read.

Fortune worked with Bill Gates and his colleagues at Breakthrough Energy and Gates Ventures to put together this extensive package of stories on the signal crisis of our age. That crisis is not the COVID-19 pandemic—it’s climate change. Click the image above to read the rest of the stories.

In advance of the book’s publication, Fortune editor-in-chief Clifton Leaf also sat down with the Microsoft cofounder and high-stakes venture investor for a sprawling conversation about what needs to be done now to stop climate change in its tracks, the energy “miracles” that are sending the wrong message, and where he’s placing his multimillion-dollar bets right now. That our discussion took us to his fascination with the Impossible Burger (and a smattering of “dish” on some other famous investors) is just gravy.

The conversation has been edited for length and clarity.

Bill Gates
Photo Illustration by Fortune; Background by Getty Images; Portrait by Spencer Lowell

Bill, you’ve just written a playbook for avoiding a “climate disaster” when much of the planet is still focused on another disaster—the coronavirus. Did you have any concern that readers might not be ready to engage with two existential threats at the same time?

So just a little bit, first, about why I wrote the book. It was a few years before I left Microsoft in 2008 that some friends who worked there said, “You know, Bill, you should get involved in the issue of climate change.”

I had already developed a fascination about the physical economy—things like steel, cement, electricity—by reading Vaclav Smil. He got me thinking about the fact that we take the ease of these things for granted. Electricity, for example, is so reliable and so cheap, and Smil talks about all the amazing work that went into that. But these are things that contribute to greenhouse gases and driving climate change, of course. So I began thinking about the idea that this was all going to have to change. And I thought, “Is that really gonna happen?”

Of course, the foundation was doing all this work in poor countries, where buildings are often built out of scrap metal. There are no transmission wires. For water, in some places, they have little tanks up on their roofs because the distribution system doesn’t exist or if it does, it’s super unreliable.

So anyway, my friends introduced me to a couple professors—Ken Caldeira [at the Carnegie Institution for Science] and David Keith [who’s now at Harvard]. And I think we started meeting six times a year. They would bring other experts in, and we would take a topic like storing energy or electric cars or making steel, and we’d have a half-day discussion with a lot of reading in advance. I was fascinated with the topic.

And I had enough of a framework that I gave a TED Talk in 2010 [called “Innovating to Zero”]; I have three given three of those talks. I have one about how government budgets are a problem—and I promise you, someday that will be viewed as an advanced warning. I have the 2015 one on the pandemic, which is now probably the most viewed thing I’ve ever done. [Editor’s note: That prescient TED talk, “The next outbreak? We’re not ready,” has now been viewed 39 million times.] But the one on climate—it’s not very long. I think, what is it, 15 minutes?—you might look at that.

The point of that talk was to say, “Look, this is superhard and requires innovation—a lot of innovation.” And even though I know a lot more today than I did a decade ago, that’s still my basic framework. I have to admit, that’s my basic framework for anything. That’s the framework for the book.

But just as it is now, the world was also a little distracted by another global crisis a decade ago when you gave your TED Talk on climate change.

Unfortunately, in that time period—circa 2010, after the financial crisis—the energy [for addressing] climate change goes way down. And then, as the economy starts to recover, interest goes up a little bit. So the next big milestone for me was the year in advance of the Paris climate talks in 2015 when I was saying to everybody, “Hey, how come they have these meetings and the idea of an R&D budget and spurring innovation is not in the meeting?”

The framework is just, “Hey, let’s come in as countries and talk about our near-term progress.” And there has been near-term progress using wind and solar for electricity generation and using battery electric-powered cars. And I’m not saying those things are easy, but those are the easiest sources of emissions to lower. And so whenever you’re just saying, “Okay, what can you do five years from now, or 10 years from now?” nobody comes to the meeting and says, “Okay, I’m making my steel a whole new way.” Because making steel is not something one country does differently than other countries. Steel is a globally competitive industry. And so the various sources that cause over 70% of our greenhouse-gas emissions never come up in those, “Hey, let’s reduce short-term emissions” discussions. Weirdly, the hard stuff—the 70% that steel and cement or aviation are great examples of—almost doesn’t come up at all.

But then, in 2015, at the COP21 [Sustainable Innovation Forum—held alongside the United Nations Climate Change Conference in Paris that year], it does come up. The French organizers wanted to do something a little different—and they wanted [Indian Prime Minister Narendra] Modi to come. And Modi didn’t want to come for the, “Hey, what’s your short-term reduction?” game—because India still needs five times as much electricity as they have today in order for much of the population to live even a basic lifestyle. And so he wasn’t going to come to that. But this side thing at COP21, which was called Mission Innovation, was not only focused on upping energy R&D budgets, but also on making sure there were private sector investors who are willing to take the risk—to turn these promising ideas into companies—if R&D labs turn the ideas out.

At that time, the bulk of the green investing was being done by people like John Doerr at Kleiner Perkins, or by Vinod Khosla’s venture funds—and, you know, they hadn’t had a very good track record. Kleiner Perkins had bet on Fisker [Automotive] instead of Tesla. There were a lot of solar panel [bets] that didn’t go well.

So I made a commitment to raise money to invest in the space. And that became Breakthrough Energy Ventures—which is part of Breakthrough Energy, the umbrella for everything I do on climate. Breakthrough Energy has a few different areas to it, including an arm that focuses on policy solutions. But the venture investments are the biggest aspect at this time. (You can read our inside look at Breakthrough Energy here.)

So we raised that fund—with just over a billion dollars. And actually now, we’ve invested in 50 companies. That’s gone super well. We’ve gotten other people to invest. And by the time this book comes out [today, Feb. 16], we’ll have the billion dollars for what we’re calling BEV2, which will be the next 50 companies—you know, most of which will fail. And even the ones that succeed will be harder to get there than your typical software company due to the amount of capital involved and the kind of industrial partnerships you need to make them work. So it is quite a different kind of investing. My aim is to help scale that innovation.

But as you’re scaling up these investments and rallying other investors and policymakers to focus on these radical energy-related innovations, the pandemic hits. You yourself seemed to concentrate your attention—and your philanthropy—on finding therapies and vaccines for the coronavirus, which you spoke to Fortune about in September.

So the current generation, you gotta give ’em credit: Even through we’re in the midst of a pandemic, people truly care about the climate crisis, too. It’s not like it was during the financial crisis [of 2007–08] when people were like, “Hey, things are tough now, and that climate stuff, that’s way out there.” Even by 2010, if you polled the public, you’d find that interest in the climate had gone way down. It began to build up gradually over the next decade, but as we hit the pandemic, I thought, “Okay, what’s gonna happen?” But it’s actually gone up somewhat during the pandemic, which is kind of weird.

Likewise, look at the experts President Biden is picking—he’s putting climate people throughout the administration. And he says he’s focusing on several overlapping crises—you know, the pandemic and climate change are there at the same level, which is pretty impressive. And, of course, during both the presidential primaries and the election, the number of climate questions asked to the candidates was greater than we’ve ever seen. Or look at the European recovery plan, its capital investment is very climate-oriented too. Over a third of those dollars are connected to that.

The last thing you’d want is to have people think, ‘Oh, we’ll just make sure there’s a few more solar panels.’ And then, 10 years later, emissions haven’t changed. And so unless there’s a plan, you’ll just end up generating a lot of cynicism and disappointment.

So I feel lucky, in a way, that young people are making this topic important. And I didn’t do that when I was that age. You know, I wasn’t blocking traffic.

But my view is, “Hey, since you guys care about this so deeply—and you have the idealism to say, ‘Let’s do what it takes to get to zero by 2050’—you really deserve to have a plan where you work backwards and say, ‘Okay, what has to happen to steel and cement and aviation?’” The last thing you’d want is to have people think, “Oh, we’ll just make sure there’s a few more solar panels.” And then, 10 years later, emissions haven’t changed. And so unless there’s a plan, you’ll just end up generating a lot of cynicism and disappointment.

And so you decide to lay out that plan in a book.

Yeah, so my book is about the plan. It doesn’t explain why climate change is bad. Okay, I have one chapter on that. And I have a chapter on adaptation. But most of the book is this: “Hey, there are 51 billion tons of greenhouse-gas emissions. Here are the sectors that are responsible for those emissions.” And then we can look at each sector and say, “Okay, how much more expensive will it be to produce the same results with zero emissions?”

And that amount is the metric I called the Green Premium. [See Gates’ commentary on published today, “How ‘Green Premiums’ can help us solve climate change.”] So you can ask: “Okay, what’s the Green Premium today for cement? Is there a company out there whose technology can cut that in half? Is there any chance it can get to zero?”

Getting the Green Premium to zero is magic, which will happen for electric cars over the next decade. That is, you won’t have to government-subsidize the innovation or subsidize it from your individual budget. Once you’ve gotten that volume, and things have gotten cheaper, and you have the charging stations, etc., then the politicians can have some involvement. They can then say they’re going to ban gasoline cars by, you know, 2035, 2040, and the public doesn’t go, “What?!?”

If that Green Premium gets down close to zero, such a policy becomes possible. But we need to do the same across all the areas where we have lots of carbon emissions. And some of the areas, like in steel production or in making cement, are just superhard. So, anyway, that’s the basic thing.

One of the things I really like about the book is that the reader gets to experience your own learner’s journey into climate science. Every time you marvel about a fact—like that gasoline is cheaper than soda, for example—or explain why certain gases either absorb or reflect the sun’s radiation depending on their atomic makeup, it feels as though you are, somewhat joyfully, sharing that discovery with us. How conscious were you about sharing that personal journey of discovery in writing the book?

For me, all this stuff is so interesting. But when you read about it somewhere, it can often seem really opaque—like you’ll read an article that’ll say, “Oh, this reduction is equivalent to 10,000 houses or 40,000 cars.” There comes a point where you’ve really got to decide, “Am I going to take the effort to understand, numerically, what the total is or not—and then understand what these pieces of the total are?” I wanted to do the fairly simple math, so that when someone talked about the carbon reductions from those 10,000 houses or 40,000 cars, I could answer the question, “Okay, what percentage of the total carbon burden is that?”

When most people read those articles, they’re like, “It’s just gibberish.” And at first, when you try and get it, the number of energy measures are so large that, you know, it’s very confusing. For me, it was appealing to just map it to one thing.

I have to admit, my book owes a lot to Ken Caldeira and David Keith, who educated me and brought all those people in. And then there’s people like [former Microsoft CTO] Nathan Myhrvold and [prolific inventor] Lowell Wood, who, whenever I get confused about physics or chemistry, I just send them an email. And, of course, this guy Vaclav Smil, who’s written so many books—his fascination with what the industrial economy was like in 1800 and the great breakthroughs that followed has been an inspiration. I love this stuff. And it is actually simple. As with most areas of domain, there are only a few concepts, but you really have to get those concepts.

There’s the rate of energy, the amount of energy, and how long you can store it. For power, these all add up to “Can you use it when you want to use it?” And so Smil does these things where he shows, “Okay, how much does a city use, or how much does an individual household use?” And so you can get those numbers and that kind of common sense in your head.

As part of this, I actually took my son to a coal plant. And then to a cement plant and a paper mill just to see these things. Because we software guys are—I mean, I was terrible in the chemistry lab. So I was like, “Let’s just do the equations of chemistry, not, you know, do the stuff in those test tubes. I might blow something up.”

Growing up, I didn’t build little trains or planes or anything like that. So I’ve always felt, “God, that physical stuff, that’s the real thing.” I mean, somebody built those machines and those factories.

Anyway, yes, it’s been fun learning all this stuff. If I say to people how many hours I spent with those guys or how many books I had to read, I have to be careful because that sounds like bragging or something. But my greatest benefit is that I love being a student. And I love the idea that if I learn enough, it’ll actually be simple. I have enough confidence in that—and in smart friends who will help me get there—that, yes, I can say, “Okay, I’m going to learn about these things and eventually get to the point where it all fits together.”

Whenever I see something where the investment, if it succeeds—in terms of the societal impact per dollar—is like a thousand times return, I feel like the venture capitalist putting early money into Google or something.

At one point in the book, you seem to have gazed into the future. In the first galley that I read way back in the fall you mention “the U.S. withdrawal from the 2015 Paris Agreement” and say that it’s “a step that U.S. President Joe Biden later reversed.” It’s amazing that you knew that had happened long before the election.

Most of the work on this book was actually done about 10 months ago. I thought a lot about whether I should put the book out then. At first, the question was, if Trump gets reelected, then the U.S. won’t lead—and that would obviously affect our ability to do the things I call for in the book. You know, this book is about innovation, and the U.S. controls maybe greater than 50% of the innovation capacity in the world today—it’s in our universities, national labs, venture capital. And so if you don’t have the U.S. pushing its innovators, then that’s felt all around the world. Because in the U.S., we’re not just innovating for our own citizens, we’re also innovating on behalf of everyone else. If we learn to make clean [carbon-free] cement cheaply, then 30 years from now, when developing nations are housing their people, they’ll choose the clean cement.

And so, as I was writing, the last few chapters got really convoluted, having to couch things in, “Okay, if a Democrat gets elected…” and so forth. And then the pandemic hit, and I thought, “Oh, people might be a little distracted.”

So I just decided to do a postelection revision of the book and maybe add some technical advances. And so I did a whole run-through of the book after the election—and I was able to put a number of things in. On chapters one through nine, there were very few edits of any significance. On chapters 10 through 12, where I talk about government policy and the like, I actually did make a fair number of changes. Even with Democrats controlling the White House and Congress, with the Senate nearly evenly split—and with the kind of deficit we’ve gotten into—we’re going to have to come up with a plan that doesn’t require a huge level of resources. But that’s very doable. I’m actually quite optimistic about that.

The politics of this are so complex. In your road map in Chapter 11, you suggest both incentives and disincentives or penalties, in terms of policy initiatives—both carrots and sticks. But in the past, we’ve seen that the sticks are much harder to put in place. In the case of the Affordable Care Act, for example, where there was a Democratic majority in the executive branch and in both houses of Congress, the mere concept of an individual mandate and an associated penalty ended up causing a legislative civil war—one that’s still going on to this day. So when I came to the part in your book about setting a price on carbon, effectively, a “carbon tax,” big enough to offset the Green Premium, I wondered how confident you were that something like that could actually happen in any near-term timeline? As you write: “Putting a price on emissions is one of the most important things we can do to eliminate Green Premiums.”

Well, without innovation, even if carbon capture can come down to, say, $100 a ton, we’d still have an enormous challenge because we’re emitting 51 billion tons of carbon a year. And so somewhere in the world you’d have to find $5 trillion that you would spend per year. That’s over 5% of the world economy. And there’s just no way, not a chance, that that happens.

Of course, any kind of carbon tax is politically very difficult. When France tried to raise their price for diesel, people shouted, “Hey, you can’t do that!” And inevitably such things are always cut because they fall disproportionately on people. You can try and offset that. But in that French case, the people who lived outside the cities and had to drive longer distances felt the elites in the cities were ignoring them—which is, you know, the general political phenomenon in all rich countries now. In the end, they got that diesel tax repealed.

So there’s not much willingness of the public to pay now to avoid these negatives—most of which are far out in the future. I mean, yes, there are some weather things—forest fires, very hot days—that are happening now. And I’m glad people are paying attention to those. But if we don’t act now, these will be nothing compared to what we’ll get in 2080 and 2100, and that will be even more so for anyone living near the equator. So going outdoors in India in 2100 during their summer months is gonna be unbearable. Outdoor work will be impossible. The body can only sweat so much.

Steel is so reliable. Just as cement is. They have a formula for exactly what goes into it. They don’t test its characteristics; they just test that you followed the formula. And so there’s no niche in the marketplace that inherently wants to pay more for clean steel or clean cement.

Near the end of your book, as you’re summarizing your plan for getting to zero emissions, you talk about “accelerating the demand for innovation.” That notion is really critical, it seems, to success—and it also seems different from much of what you’ve done in your philanthropic work. In the foundation you run with Melinda, you have often funded the supply of innovation—investing in lots of these startups. But creating markets in some places where they don’t exist—building the demand side—is a different sort of challenge.

In the drug industry, for instance, Big Pharma companies provide a market for innovation through legions of biotech startups, competing to in-license experimental drug candidates that might be promising in one indication or another. Is there a similar model on the climate front, with big energy companies providing a ready market for zero-carbon discoveries from energy startups? Or as with defense industry innovation, will the “market” have to be provided by governments?

You know, energy’s a lot harder. Software, by comparison, is the easiest in this regard. Medicine is kind of in the middle. And these climate-related things are the hardest. In software, there’s always some customer who doesn’t like the current software—and who thinks that yours is either simpler, cheaper, or has more functionality in some area. So you carve off that piece from whoever’s dominant, and you can go after a certain set of customers because you’ve matched up to their needs better. That provides a ready market for software innovators.

Likewise for medicine. You look at the various diseases, and maybe you make something that’s oral instead of injected or that reduces some side effect. Or hopefully, every once in a while, you find a completely novel way to attack a disease that there is no medicine for. There you also have all the regulatory issues and inspection issues and side effects, etc.

With climate, the thing that’s so difficult is that when you make steel—clean steel, there is no added use or benefit beyond what it does for the planet. The mere fact that you’ve made it slightly differently than the historically accepted process is likely to provoke questions among buyers: “Does it last, does it get brittle, does it rust?” People will look askance at any change. That’s because steel is so reliable. Just as cement is. They have a formula for exactly what goes into it. They don’t test its characteristics; they just test that you followed the formula. And so there’s no niche in the marketplace that inherently wants to pay more for clean steel or clean cement.

Now, what happened with solar panels was that there were satellites where you needed a source of energy, and the solar panels were the only way to do it. And you got Germany and Japan to buy them even when it was non-economic. Not a huge volume—but enough volume that you started this learning curve. And then other countries, including the U.S., put their tax credits in, and you got to expand this learning curve. And we’ve seen the learning-curve pattern play out with wind and lithium-ion batteries. Those are three miracles that have happened.

Now, sadly, most people think, “Okay, just because you make batteries good enough for cars, we can make them good enough for the entire electrical grid—and we’ll have another miracle there.” Well, they don’t realize that many attempts at miracles like, for instance, fuel cells, or hydrogen-powered cars, have yet to pan out.

There are other complications with innovating, for example, in the case of fission-powered nuclear reactors. Here, the cost and the fear of safety issues are such that, unless we rebuild the reactor with a whole new design that’s utterly different in terms of economics and safety, that will be a dead end. And so that’s what I’m trying to do with [nuclear energy company] TerraPower [where Gates serves as chairman of the board].

And so if you project the successes we’ve had onto all these areas, you can end up thinking this thing is far easier than it is. And your question’s a super good one because that’s one piece in the overall strategy described in the book, having a buyer for that clean steel.

To that end, say the Green Premium is $100 a ton today. If someone were to come along and start a clean-steel company that has a mere $50-per-ton premium, you’d go to that person and say, “Hey, hallelujah, good job, buddy!” But in fact, that’s a zero-sized market, because $50 more per ton just means it costs more money.

Now, what we’re going do is to try and get richer companies, including the big tech companies, to agree that whenever they build a new building, they’re going to use 20% clean steel. And their reward is they get to say, “Hey, we used 20% clean steel.”

So we have to create the demand side. Your question is a super good one. I, of course, love the supply side of innovation. You know, to go find the professor, go find the guy with the crazy idea. And so Breakthrough Energy Ventures has lots of those. Among the 50 companies in the fund is one where a guy has figured out how to store energy long-term by pumping water into the ground and storing it between underground layers of rock. It takes energy to pump it down, but that can be done when electricity is abundant. Then, when he wants energy back, he opens the well and the pressurized water rushes out, powering a turbine.

And that shouldn’t work. The company’s called Quidnet. I think there’s some reference to Harry Potter, or something. I don’t know why. I should know that. But anyway, funding that guy is an example of what we like to do. That’s the supply side of innovation.

But you need the demand side, too. Take these artificial meat companies. Well, amazingly, their market is driven by consumer choice. When Burger King began offering the Impossible Whopper, the demand was super high, which is good.

But in steel and cement, because the green-tech is so obscure, it’s much harder to create that demand. In the case of aviation fuel, the extra cost of a clean fuel is going to mean your plane ticket is, like, 25% higher. You know, people might say at a cocktail party that they’re glad to pay that 25%. But actually, I doubt that when push comes to shove, that such an effort can be funded by consumers paying the full cost of that Green Premium.

So it is much harder than creating a marketplace for innovation in the software or even medical industries. And that’s probably why Breakthrough Energy Ventures has a 20-year life span and a different set of incentives. It has investors who know that most of the companies in there are likely not to succeed.

Have you have ever thought of opening up the fund, or one like it, to public investors and letting people buy shares in order to help create that demand?

Yeah. We haven’t yet formulated this, but I do think being able to say to companies—as they figure out what their carbon emissions are and the cost of that, estimated at some price per ton—that they should be willing to write a check out to mitigate that. And those checks, in principle, could finance a fund that does an auction to say, “Okay, who can give us green steel at the lowest price? Who can give us green cement?”

And so I like that idea of getting, both individuals who feel guilty about their carbon footprint, and companies—particularly quite profitable ones—who could underwrite a buying fund that does what the space industry did for solar panels: that is, create the demand side. That is something I want to organize. And, you know, people will say, “Okay, I’m a member of this Breakthrough Energy buying effort” at some level and be proud of it.

You could do as Warren Buffett does at Berkshire Hathaway and create “A” shares for rich companies and “B” shares for the rest of us.

Yeah, platinum and gold.

I would have said to you, at the outset, that ‘fixing meat’ was as hard as creating zero-carbon steel and cement. And now, though it’s not yet solved, there’s a pathway.

If I might draw a strategy from watching your long philanthropic career it is that you seem to have structured your efforts around what one might call first movers. If you can fix X, then you can reduce all the bad outcomes that cascade from X.

So when you and Melinda began the Bill & Melinda Gates Foundation, a logical first step was to invest in global health, working on ending scourges like malaria, diarrheal disease, neglected tropical diseases, polio, and others. Reducing the number of children who get sick and die each year was not only a good thing in itself, of course, but it would also in theory cause a cascade of good effects: reducing endemic poverty, for one thing, by helping to ensure that there were more healthy, productive citizens to strengthen a developing nation’s economy.

So if you fix one thing, you can have a positive effect on something downstream. You see the same multiplier effect when you look at your foundation’s investments in education—and particularly in educating young women and girls. As Melinda says, the No. 1 indicator of under-5 mortality is the educational status of the mother. So fix that, and you can lift generation after generation out of poverty.

You can follow that same thread with your other major investments: trying to improve sanitation, for example, which will reduce cholera, typhoid, diarrheal disease, dysentery, and Lord knows what else. And so forth.

So now, you’ve turned in a big way to climate change—which, arguably, might be the ultimate “first mover” in a chain of problems, at least if we don’t stop it in its tracks: Fix climate change…or the world ends. If that reasoning is true, does your focus on climate change become even more dominant in this mix of causes that have grabbed your and Melinda’s attention over the years? And how do you begin to triage all these urgent demands for your time, attention, and money?

Yeah, in the foundation space—in the global health stuff—the impact per dollar is huge. We’re saving lives for less than $1,000 per life saved. And there’s this sort of market failure that, because malaria doesn’t exist in rich countries, there’s no market signal that innovators should come up with drugs and vaccines for malaria, or an overall strategy to get rid of it. The poor countries where it’s endemic don’t have the resources to solve malaria.

When I gave the first $30 million to fighting malaria, I became the biggest funder in that effort. You could say that’s a sad thing—or an opportunity to dramatically push for change. People in climate talk about carbon capture. Well, I’m the biggest funder of those companies, and that amounts to tens of millions of dollars. That includes companies like Climeworks, Carbon Engineering, and Global Thermostat. There’s like four or five companies in that carbon capture space. And fortunately, there will be more because there are several different approaches there.

Whenever I see something where the investment, if it succeeds—in terms of the societal impact per dollar—is like a thousand times return, I feel like the venture capitalist putting early money into Google or something. You know, I’m drawn in. And seeing those potential innovations succeed over time, that’s what I enjoy doing. I enjoy gathering the kind of people who can see those things and be willing to be patient. And maybe even, in some cases—like preventing or curing HIV—bet on multiple potential paths, knowing that a number will fail. But still, even if you add all that money to the different approaches, even if just one works, it’s very, very magical.

So it’s kind of an innovation mindset. There are things like improving U.S. education where, as yet, if you take broad indicators such as, “How good are U.S. students at math—or reading and writing?” or “How many kids drop out of college?” we haven’t had the kind of scale impact we’ve had on our global health work. And yet, we still believe we can make an impact. In fact, we doubled down on this effort. The idea that now more kids will actually have laptops and Internet connection, which got pushed very forward during the pandemic, has us really building a new generation of curriculum that’s very different.

So hope springs eternal even in the areas where, so far, you’d have to say the impact has been almost hard to detect. You don’t know in advance what’s going to work. And most efforts to fix a problem turn out to be harder than you think. A few—like when I put money into Beyond Meat and Impossible Foods—I would have said to you, at the outset, that “fixing meat” was as hard as creating zero-carbon steel and cement. And now, though it’s not yet solved, there’s a pathway. It’s almost like with electric cars where, as the cost goes down and the quality—in this case, of the various ways of making synthetic meat—improves, the Green Premium will go to zero. That one actually surprised me in how quickly it went.

You know, if Edison came back, he would look at our batteries and say, ‘Wow. These batteries are about four times better than my lead acid batteries. Good job.’ That’s a hundred years of battery innovation.

One thing that surprised me in the book was what seemed like your belief in a natural limit to energy storage. There are some who still hold out hope for the near-infinite storage battery—or at least one with hugely expanded capacity—which would make storing energy from the sun or wind or whatever far more feasible in a grid-like way. To me, that was one heartbreaking part of the book. It’s hard to come to grips with the idea that there’s no Moore’s law for energy storage.

Yeah, the fact that Moore’s law has worked as long as it has is super mind-blowing. That software and digital stuff makes people think miracles are always possible—whereas this is more like what’s happened to gas mileage over a hundred years. You know, if Edison came back, he would look at our batteries and say, “Wow. These batteries are about four times better than my lead acid batteries. Good job.” That’s a hundred years of battery innovation.

And people get confused between batteries for electric cars and what range you get—in which just expanding range by another factor of two makes that really good—and batteries for grid storage where the aim is to collect solar energy in the summer and use it in the winter. So in that example, for that whole battery, you’re getting one benefit. One time per year you get to pull that energy out.

The cost and scale of that is so difficult. You know, we’re more than a factor of 20 away from it. I’ve lost more money on battery companies than anyone. And I’m still in, like, five different battery companies—a few directly and a few through BEV [Breakthrough Energy Ventures]. There are only three ways to solve the electric grid problem: one is a miracle in storage, the second is nuclear fission, and the third is nuclear fusion. Those are the only possibilities.

Thank you, Bill.

Thank you so much. This was fun.

Explore Fortune’s Blueprint for a climate breakthrough package: