Exponentially with Azeem AzharRenowned venture capitalist Vinod Khosla tells Azeem Azhar that humanity will have to rely on new and unproven technologies to reach net-zero emissions by 2050.
To tackle climate change, we'll have to put our faith in technologies that don't yet exist. That's the premise for today's conversation. I'm Asimazar. Welcome to the Exponentially podcast. Renewables are booming. In fact, solar is the fastest growing energy technology in history. Why isn't this record breaking growth enough to bring us to net zero? It turns out that some of the cornerstones of our modern life are big ticket items when it comes to carbon pollution. More than thirty percent of emissions come from making steel, chemicals, cement, and food. So how do we keep what we need and take out the carbon This challenge requires new solutions. I've come to Silicon Valley to meet Vinode Koestler, a venture capitalist who has made himself and his investors billions of dollars. He's pouring that cash into the technologies that sound like science fiction. Synthetic meat, fermented jet fuel, and nuclear fusion, all of them huge bets to help solve the biggest problem humanity has ever faced. Now, when I first met you more than a quarter of a century ago, Venode, you were really well known as a technology investor in traditional it. You had built some microsystems, you had invested in computer networking Juniper, which returned two and a half thousand times its original investment to its investors, and one of the first Internet search engines, Excite. But in recent years you've got a reputation in a new area, which is funding entrepreneurs who are built Now, when I first met you, more than a quarter of a century ago, Venode, you were really well known as a technology investor in traditional it. You had built some microsystems, you had invested in computer networking Juniper, which returned two and a half thousand times its original investment to its investors, and one of the first Internet search engines, Excite. But in recent years you've got a reputation in a new area, which is funding entrepreneurs who are building hard technologies to tackle climate change. So it feels like there was a realization, a realization that climate change was a real problem. Can you talk me through how you came to that realization.
Well, the first time I started thinking about climate was in the late nineties seriously, and I realized we had a large problem about ten percent of the world's spopulation, seven hundred million people. So I had a rich lifestyle, ritch in education, rig in Madison, rich in housing, REGI in transportation, regiin energy, In every way seven billion people wanted it, And then I did the math. Math didn't work. Ten times more, ten times more cement, ten times more doctors, ten times.
Ten times more meat, ten times more meat.
Right, All those were intractable problems with a linear approach to life, and I decided to work on technology multipliers. Can you substitute for more stale with technology? Yes? Could you do housing differently with technology? Yes? Could you produce more doctors with technology? And I don't mean human doctors, AI doctors, yes? Could you build teachers with AI to be personal coaches for every student on the planet affordably? And all this has to be done accessibly and affordably.
Well, it sounds like there are two dimensions there. One part is this dimension of how do you deliver equal prosperity to all of humanity, not just the richest ten percent? And then the second question is how do you do it within the envelope, that is, the capabilities and the resources of the buiersphere, specifically our carbon budget. Coming to that first question, I'm curious, as someone who is an immigrant to the United states. You grew up in India, do you think that, having come from one of the poorer countries at the time in the world, that plays a part in your calculus of the rest of humanity the other ninety percent?
Well, absolutely, First, I'd say hard problems are really fun to work on, right because they're real challenged and everybody assumes they can't be done. But on climate there were technology approaches that made the world a better place and that were much more tractable for my skill set of technology based in ovation.
One of the ways I think about technology is that technology is things getting cheaper. When people think about that technic, they think it's a widget like this tablet that I've got here, or it's a computer, or it's a car. But the heart of technology is that, unlike so many of the other things we experience in the world, it gets cheaper every single year, and that means it's democratic. It's also inclusive. And the last time that I traveled to India, the thing that you note is that everyone there has a supercomputer because they all have smartphones. How should we think about that in terms of climate technologies?
I had this view that no large innovation ever comes from an institution of any sort, the institutional system where there's large companies, academics, other places. They're great at extrapolating the past, not inventing the future they want, because inventing the future you want involves much more innovation, which means much more risk and much more probability of failure. So very large innovations only come from the entrepreneurial world. Could you imagine somebody at Hyatt or Hilton doing Airbnb. Could you imagine somebody at Hurts or As doing Olber. Could you imagine somebody at Walmart or Target to Amazon? Could you imagine anybody at boeing a Lockheed or air Bus to rockets like rocket Lab or SpaceX have done. Could you imagine why didn't Fox, NBC or CBS do media? It was Twitter and YouTube and Netflix and Facebook. People they didn't know they were in the media business. So large innovation only comes from these big players, and frankly, the Internet itself. In nineteen ninety six when we started Juniper, every major telco player told me they would never use TSPIP in the public Internet.
So TCPIP being the technical standard that makes the Internet work and makes it different to the old phone lines that it served us very well for one hundred years.
So these Internet technology was taboo in the telecom world, and no telecom player wanted it. They wanted an alternative. The buzzword was ATM asynchronous transfer mode. Yes, that's right. And when we started, Jennifer, every player told us never, and Cisco, which was the prereer of Internet technology, said they would never do a router up for the public Internet. Never. That was a direct code from the cit of Cisco to me. Yeah, back in that.
It's kind of their business now though, isn't it.
It is everybody's.
Yes.
My point is you have to take these large dress and we did say we'll will build it and they will come, and that's exactly what happened. In fact, this particular episode was a very large return for Clima Perkins for an hour. Is there twenty five hundred x about seven billion in profit on three or four million dollar investment not back makes up for a lifetime. Are the losses and stupid things I've done in my life.
There's a very interesting parallel between that story of early internet technologies and I think what you have said about technology and climate change. One of the things that you've argued is that even though we've seen tremendous progress in the spread of solar power and wind power and electric vehicles. In the cold country of Norway, ninety five percent of new cars are electric. In China's enormous car market, twenty five percent of car sold today are electric. Despite all that progress, that these proven technologies will not be sufficient for us to reach net zero emissions by twenty fifty, and that we need to take bets on big, unproven technologies. Why do you think that's the case. Why can't we get there with what we already have and what is already proven to be growing exponentially.
I use a term called the Chindia price, the price at which India and China would adapt to technology, and that's the price at which these technologies are cheaper than their fossil alternatives. They may not be day one, but they have to get there where they're cheaper than cold based power plants, or natural gas based power plants, or cold natural regular cemant or regular steel. So these alternatives have to be cheaper than their fossil alternatives. In the good news is there's only a dozen things that matter. There's a dozen things that are very large emitters of carbon, and if we fix those dozen, we solve the problem. If we don't fix those dozen, we will have a problem no matter what else we do.
And those dozen are things like steel where we pretty.
Steel, semans, aviation fuel, industrial heats, the industry, heat, HVAC, air conditioning in people's homes, so it's all the bigger matters. Agriculture is a big one, so animal protein and nitrogen fertilizer. So there's a dozen of these that are very very important to solve in the bulk of their missions. If you solve this, I think the climate problem is done right, which leads me to the most exciting part of this. And my view is we only need a dozen people, which I call instigators, to instigate the change in each of these areas good example, there was no chance anybody at General Motors of Volkswagen or your favorite company was going to make electric cars happen. But Elon Musk didn't know the electric car business, so he made it happen, starting from first principles right and taking a lot of humorist a lot of risk, a lot of criticism as well, a lot of criticism. Near bankruptcy, it is hard.
I guess your argument would be that because of the ninety percent, the ninety percent poorest on the planet who expect to have a better quality of life.
We can't stop.
The investment in people's own prosperity. And so the only way you square that circle is through technology, which is things getting cheaper, and that's why we need to pick off these dozen problem areas in parallel. And the mechanism that you have seen work both in renewables and decarbonization recently, but also through the Internet, has been the instigator. Is that a fair summary.
That's a fair summary. So Elan Musk. Without to Landmarsk, we wouldn't be on the path to electrification today. Whether you like him or not, whether he succeeds or fails, he has changed the worldview. The same has happened with somebody like Pat Brown with meat.
He does the impossible burger.
When Pat started and we invested in impossible foods about a little more than ten years ago, it was definitely not an area people invested in, and people asked me, what are you doing investing in hamburgers? But it was this sustainability objective of reducing emissions from meat production. The traditional way, and he actually uses traditional plant proteins plus heame, which is a brood protein brood just like your beer, which is the blood component of it that gives it all the taste. Now, no food company would have done that.
So we've got electric vehicles and we've got burgers that can replace the traditional beef burger. What are three or four other problem areas that you're finding?
So long time ago, we invested in Lanzattag Jennifer Hommegren is trying to change aviation fuel to a sustainable aviation fuel. In fact, the goal is to produce it eventually from municipal waste, So municipal waste to aviation fuel would solve the carbon emissions from aviation. That's a really big example. Now, we invested in Commonwealth Fusion Systems. When I met Bob Momguard, he was just a senior fellow at the MIT Plasma Fusion Lab and thinking about starting this company. And we help them get going. And if we get fusion, if you're right, we will have fusion for the planet and reliable energy. Now, solar and wind are great, they've done very well, but you don't want to watch your football game only when the sun shining and or the way is blowing. Right, al Gore is just wrong on the cost of solar power. Solar power that's reliable and available when you want to watch your football game, not when the sun's shining is much more expensive today and not practical.
Well, I was going to say that that. Of course, you've also backed jag Deep Singh who makes batteries with quantum Scape, So potentially you could put jag deeep's batteries and other storage solutions with solar and wind and we can then have that consistency like.
Even we are taking alternative approaches to the same problems, right, so we may have a battery solution, and we have multiple battery solutions. And I'm very optimistic about deep geothermal, which would make thermal power one hundred times more available if you could drill deep enough. And my dream would be to drill right under coal plants and tap four hundred degree heat under the coal plant so we get heat and replace the call without replacing the call plant. So we have millimeter wave drilling just like your microwave oven heating to do deep drilling as an effort. So I'm saying just because we're doing fusion doesn't mean we aren't doing geothermal also and batteries for storage, so solo power becomes reliable.
There is so much technology in here, this almost feels like science fiction. It's pretty remarkable, and it does feel that these are unproven. So if I was a policymaker or just a concerned parents, I'd need summer suaging that we could rely on these bets, how we know they could even pay off.
Uncertainty is a critical part of what we need to deal with. I would say only the improbables are important because it's the only way to do this technology multiplication so all seven billion people can enjoy the lifestyle of the riches seven hundred million. So we have to take risks. Now, what is the solution if Commonwealth fusion doesn't work and your thermal heat doesn't work. There's ten other fusion efforts that others are doing, and I'm very very glad that others are doing it. And not only did Bob Mumtngard, who's a real instigator in fusion, He's instigated a dozen other people to start other efforts.
Right, competitors with different technical and scientific approaches.
Not only that we just invested last week in a new effort called Reality of Fusion right where Bob is helping them build the magnets for his competitor in fusion.
So let's talk about fusion power, which is really from the realms of science fiction. You have backed a founder called Bob Mungard who was an academic. He was up at the Massachusetts Institute of Technology, and his business Commonwealth Fusion Systems CFS. Fusion is a really difficult technical problem. It involves getting a plasma of rare gases up to temperatures many many thousands times hotter than the center of the sun, containing that, controlling it, and then somehow extracting useful electricity from it and doing that all safely in an industrial sized unit somewhere. So how do you even go about addressing that problem?
Well, they think logically in small chunks. So we raised the first one hundred million dollars for Commonwealth Fusion with an odd argument that if we built a twenty tesla magnet, which was the key risk in getting through real fusion, if fusion didn't work, we could use the magnet for nuclear medicine, for example in your local hospital. So there were other uses for these strong magnets, MRI machines, power lines, nuclear medicine.
But let's be clear. You're talking about a twenty tesla magnet here for fusion, whereas a high end MRI machine might have a three tesla magnet. And this is orders of magnitude right from three to four to five to twenty. It's not just six times more, it's seventeen orders of magnitude more powerful. So we are we're talking about a really significant increment.
The thing to understand about magnets, for example, and this was the key realization, the size of a fusion reactor will become scaled inversely with the fourth power of the magnet strength. So if you go from five tesla magnet to twenty tesla, your fusion reactor will be two hundred and fifty times smaller. Think about it. That not only makes it manageable, constructible, fast, changeable, fast, much lower cost modular.
And so with the modularity you can string these things together. You increase volumes. When you increase volumes, you improve learning rates, so your unit costs comes down. It's what we did with silicon chips.
Yes, not only that high temperature semiconductors, but developing and they enable these high magnets. So transition in one technology in high temperature superconductors allowed for high Tesler magnets, which allowed for fusion. But high temperature semiconductors are enabled and in fact can be do cheap if you use the solar technologies thin film solar technologies that are used to make solar cells. So those same semiconductor processes could make high temperature superconductor tape, which would make magnets stronger, inch and much more scalable in high volume. So this is an important point when you put all these together. The idea of solar technology is being used to make superconductor tape used to make magnets that are very very strong, which make fusion reactors really much much smaller. Two hundred and five fifty times smaller is a big deal. But let me go even further. There's a beautiful plan Bob and I have talked about. And everybody says, you can't build five thousand power plants the regulation permitting all grid connections. There's this much simpler entrepreneurial solution to this. We won't build fusion plants. We'll take coal plants and replace the cold boiler with the fusion boiler.
And let the plant be provided you can get the size down.
Provided we can set get the size down, which we're pretty sure will be the same size or smaller than a boiler. If I might spend one more minute on this analogy, because this is where experts are almost always wrong, and that's why they've not participated in any large innovations. Experts at big energy companies like the or Siemens, or experts at Volkswagen. Yeah, they don't think nonlinearly. I say all this to you and say, how are you going to build that many fusion reactors? Let me give you the following analogy. In the Second World War, there were five liberty ships built in the United States in the ten years before the Second World War started. Right the next five years, we've built close to five thousand liberty ships. If we can do that, Can we build five thousand fusion reactors using high volume technologies like solar manufacturing, No question, we can. I think the pundits are fundamentally wrong about these things.
Well, there's something that I really like about the way that you tell this story, because the beauty about the impossible is that sometimes we've done the impossible before if you think about the magical things that we do with semi conductors, extreme ultraviolet lasers, the vibration free factories that are the size of aircraft hangers, but with cleaner air than operating theaters. This would have all seemed like science fiction in nineteen sixty or nineteen seventy, when silicon chips were starting to become an industrial product. And of course history only ever rhymes, it doesn't repeat. But we do have some precedent of achieving these improbable outcomes.
That's why I say improbable so important. In fact, the only things that important, and we just got to take enough shots saw that some of these improbables happen in these dozen areas to solve climate change.
The climate change problem is really significant. We can't get there with our existing sets of technologies, and we need to find some unproven technologies that will be supported by these key instigators, these characters with certain qualities to make the improbable happen. But there's still this question. I suppose of scale and the billions of tons of steel and cement and so on, and twenty fifty is only less than three decades away, how do you actually get to that scale? I could see that through these exponentials, these technologies could get cheaper, very very quickly, and we could start to build them in great quantities. But that in of itself doesn't get you there by twenty fifty, right, It might get you there by twenty seventy or twenty ninety. Is it going to involve government incentives or interventions of some sort or some other magical thinking.
So those kinds of claims, yeah, are the purview of experts in gurus right, and pundits who pontificate but don't actually do things. Scaling is so much your problem. If a reduced fusion to replacing five thousand coal boilers and natural gas boilers in the United States, that's a very solvable problem. I just gave you the map did with liberty warships, that becomes a scalable problem. So do we have to demonstrate a power plant by the early thirties, Yes? Do we need to build ten or twenty in the thirties, yes. That then sets us up starting in twenty forty to really scale it in the next five years to every plant in the US and in India and China meet the Chindia price. Because these are semi conducted like technologies from solar power generation. We've seen what that cost has done, so that's scalable.
I want to come back to governments though, because we do have governments and they do play a role. The Liberty Warship example you gave out was a course of government initiative, and many of the core technologies of the Internet started as government funded initiatives. And we've seen how government interventions can transform industry. So, for example, the arrival of the Environmental Protection Agency drove down vehicle tail pipe emissions successfully over thirty or forty years. So in your picture of improbable, difficult, yet to scale technologies that are required over the next twenty or thirty years, what role should governments play.
Well, government has an important role to play. I'm talking about a lot of technologies. We had the initial costs of the first plank, the first five plants, the first ten plants is above the cost of the fossile compotors. So subsidies do play a role. Unfortunately, subsidies get locked in instead of being declining with scale or time, and they always should decline with scale and time. But it's very important to get these nascent technology started because they have a premium cost, which Bill Gates called the green premium. Yeah, that has to be covered by somebody, and governments can cover it till it gets to skip. No reason for subsidies on solar or ing today.
Absolutely they're cheap enough, so a government could step in with some time limited declining subsidies in order to drive those technologies down the cost curve so a market can form. The thing about that policy that I find so interesting is that it should appeal to those on the left of the economic spectrum because it democratizes technology, making it affordable for smaller firms and for households across every country and in every country. And it should appeal to those on the right of the economic spectrum because by having that early intervention, you create a market which entrepreneurs can then occupy and build businesses around. So it appears to be a policy that should appeal to both the left and the right. Why do we not see more of that policy cropping up in the United States and across.
The People on the right have generally not liked the idea of any government intervention in anything. Aim to be honest, this kind of approach will disrupt the traditional business in each of these areas, and so there will be losers. It's not a lot of fun if you're the disrupted party. One has to remember, somebody gets disrupted. It's traditionally the online players who are not innovating, and they're the traditional constituency of the right and have the loudest voice. People expect General Motives or General Electric or siemens to have more clout and say in these arguments. We've seen that with AT and T trying to control the Internet.
Of course, we've talked about these magical, unproven technologies from fusion to novel steels and cements and so on. What do you think are the chances that we can get to a global net zero by twenty to fifty the back of these radical and unproven technologies.
I would put it the following way. We will have the ability to get to global net zero. Whether we do it or not is a political question. Technology is absolutely necessary, but not sufficient. There's other factors like politics. It's not very fun being disrupted. This is why, from my point of view, if he can build fusion boilers and get the cold power plant industry in the natural gas power plant industry behind us, the people who own those plants so their value doesn't go to zero. There are partners we're doing that in Cemac, a plant that will come online the next few months in Reading, California, just north of here, a cement plant where an auxiliary product is the carbon captured as products. So the total capacity of the existing cement plant will actually go up and the old product will just be lower carbon because the carbon's being put into new cab nets that make more beautiful. Idea. If he can reaper person existing and planned, it becomes easier to get the incumbents they disrupted along and will accelerate the process.
And easier then to build the coalition that we need to political side of things to tackle this huge problem. Well, Vinode, thank you so much for sharing this wisdom with me today. I really have appreciated it.
It's been fun. Thank you.
Reflecting on my conversation with Vinode, I'm struck by three things. Firstly, he puts an awful lot of importance on the idea of the instigator, the individual who can have an outsize impact, and I think there's a lot of truth to that. Think about Steve Jobs or Elon Musk. The second thing he does is he can visualize the potential of the improbable or the highly unlikely. He can paint a picture of the future when these technologies get real. But then finally he brings it back down to something more prosaic. We just need need to get going, and if we get going, we might finally get there. Thanks for listening to the exponentially podcast. If you enjoy the show, please leave a review or rating. It really does help others find us. The Exponentially podcast is presented by me Azeem Azar. The sound designer is Will Horrocks. The research was led by Chloe Ippah and music composed by Emily Green and John Zarcone. The show is produced by Frederick Cassella, Maria Garrilov and me Azeem Azar. Special thanks to Sage Bauman, Jeff Grocott and Magnus Henrikson. The executive producers are Andrew Barden, Adam Kamiski and Kyle Kramer. David Ravella is the managing editor. Exponentially was created by Frederick Cassella and is an Eat the Pie I plus one limited production in association with Bloomberg LC
