Vinod Khosla at Milken Institute: Part III
This will be the conclusion of Vinod Khosla’s (VK) recent lengthy interview at the Milken Institute 2009 Global Conference. The interview was conducted by Elizabeth Corcoran (EC) of Forbes and can be viewed here.
In Part I, VK discussed the role of government money, capital intensity of renewable projects, and some of his solar investments. In Part II, VK discussed butanol, cellulosic ethanol, nuclear power, and cap and trade. Here in Part III, VK discusses his beef with electric cars, has lots to say about Black Swans, discusses his problems with nuclear in more detail, talks about green jobs, sugarcane ethanol, and weighs in on indirect land use issues for biofuels.
EC (39:00): Let’s get to those electric cars. You don’t like the Prius.
VK: Let me be clear, and I am going to sneak in my Black Swan. I do drive a hybrid, but not a Prius. I drive a Lexus hybrid. Hybrids are an uneconomic way to reduce carbon dioxide. If you go to hybrids or electric cars, your cost of carbon reduction is about $100/ton. If you have 10 ways of reducing carbon at $50/ton, why would you spend $100? My beef is not with hybrids; we are investing in hybrid batteries; there is a good market and we can make money at it. But do I believe it’s going to solve the climate change problem? No. (RR: None of the things that have been discussed are going to significantly rein in carbon emissions.) Save yourself the five grand, and instead paint your roof white. You will save more carbon that way.
(RR: He cited this paper by Art Rosenfeld at Lawrence Berkeley Lab: “White Roofs Cool the World, Directly Offset CO2 and Delay Global Warming“).
EC (41:10): Shai Agassi – a long time entrepreneur in Silicon Valley – has a very different approach to batteries. Are you involved in the work he is doing? Does that only work in small countries?
VK: You know, Shai has a very intriguing start-up. (RR: EC interrupts to explain that Shai is developing stations where you can go and exchange batteries for electric cars; he owns the battery and you own the car. See more explanation here.) I mentioned earlier about diversity of opinion; I am glad he is trying it and I am cheering him on. If I can help him I will. It is important to try some of these experiments. He has a particularly clever way to do something that does have a shot at working.
I want to add my Black Swan theory here. Most of you have probably read the Black Swan, or heard about it. The financial crisis is a negative Black Swan. I am a true believer that technology provides positive Black Swans. (RR: VK explains the concept of the Black Swan. Here is a link to the book at Amazon, which I have read and found to be very good). We will redefine energy because of the Black Swans of technology.
(RR: VK then explains his problem with electric cars, and says lithium ion batteries are too expensive, are limited by electrochemistry, and will be for a long time. I would say that while VK seems to have a clear picture in his head on the issues with batteries, he suffers from a blind spot about similar limitations of cellulosic biomass. He then cites all of his investments into different areas, and concludes that sheer numbers mean something is going to work.)
VK: The chance that each approach will succeed is small. The chance that all of them will cumulatively fail is vanishingly small. Mark my words: Vanishingly small, and that’s why we will have unsubsidized market competitiveness with fossil fuels. And the fossil fuel guys won’t know what hit them. I don’t see how by 2030 oil can compete. That’s why I think by 2030 oil will go to $30, because it will be the alternative cost of marginal technologies.
(RR: I think he truly believes this. Yet it shows a failure to grasp issues of scale, biomass density, logistical challenges, and much more. If it were merely a numbers game, we could solve any technology problem by just throwing enough money at it. But there are fundamental issues here regarding biomass that will never – mark my words – never allow it to be produced for $30/bbl. Sugarcane ethanol, yes, can be produced for that in Brazil. But you will never turn cellulosic biomass into a liquid fuel, at scale, for $30/bbl – for the same kinds of fundamental limitations VK mentions for batteries.)
EC (47:40): So by 2030, what will be the primary fuel?
VK: I have a paper on my website that postulates about a technology race between biofuels and batteries. Whichever one makes the most rapid progress will get the larger percentage of the total passenger miles driven in the world.
EC (48:30): Does government risk factor in? There has been a cautionary tale in biodiesel, where there has been great interest, lots of money pumped in, and yet due in part to vagaries of how the environmentalists and government regulations have crashed into each other, you have got more than 100 biodiesel fuels (RR: Biodiesel plants, I presume?) around the country, none of which are producing fuel.
VK: You know, that’s true, but you also have bankrupt financial companies. Look, failure is the natural mechanism of capitalism. But you are right. There is government risk. But we fixed a lot of that last week when the Low-Carbon Fuel Standard passed. It will force the right decisions looking back.
EC (50:18): There have been many technologies – and Kleiner invested in many early on – where the technology, the marketplace, and the government were not in sync. And the technology dies.
VK: I think that’s the wrong way to look at it. Any start-up has risks. It has technology risks, market risks, it has financial risks. It has other risks; it has people risks and management risks. What you are doing as an active investor is balancing those risks. What we are tending to do is increase technology risk so we can reduce market risk. We will generally take on more market risks, have a bigger jump, and a larger probability of failing at the technology such that when we enter the market we have a larger competitive advantage.
EC (51:30): What are you hearing from the limited partners, the people who invest with you? Is there a tolerance for that sort of risk?
VK: Absolutely. My impression is venture capital has gone too far away from real technology risk. The limited partners are thirsting for more technology risk. The limited partners tell me that the earlier stage they can get in on the technology risk, the better they like it.
EC (53:25): I am going to open it up to questions in a minute, but one more question from me. Let’s go back to nuclear for a minute. Aren’t there Black Swans in the nuclear industry? (RR: I was thinking the same thing earlier; Black Swans only appear to have been considered by VK in very specific situations. A positive Black Swan is going to make some of his technologies successful, but he seems to discount any positive Black Swans from other sectors).
VK: There probably are. In fact, Bill Gates is funding one. The problem with nuclear, I think, is different. Because of the NRC it takes 20 years to build one. And I have to give them $100 million to approve every step of the process. The problem with nuclear is that the innovation cycle is very long. If I am building a nuclear plant, I think of something, 20 years later I build something and see how it performs. If I am building a solar thermal plant, six months later I change my manufacturing line. I can even do it half way through building a power plant.
EC (54:40): And if you are building an ethanol plant, two or three years later it’s ready.
VK: Yeah, though every six months people plan on changing the bug in their plant. Every six months you change the bug. Keep evolving it, improve the efficiency. The cycle of innovation – how long it takes – is a really important metric for judging how effective a technology will be in getting to market.
EC (55:20): OK, good. First question.
Q1 from audience (55:30): My question is on nuclear. You said you weren’t interested in building, but how about the services component, i.e., servicing the waste and so forth?
VK: I think it’s a limited investment opportunity. I don’t think it’s an explosive opportunity. (RR: I suppose that depends on whether critical mass is reached.)
Q2 (56:10): What about superconductivity?
VK: It’s an interesting area, I just haven’t seen the pace of innovation. Sometimes it’s self-fulfilling. If you are not interested, nobody funds it, then nothing happens. I would love to see a breakthrough in room temperature superconductivity. (RR: He then said Kleiner invested in a couple of companies in the late 80′s; he mentioned American Superconductor).
Q3 (57:20): With respect to cellulosic ethanol, this question of indirect land use that has ended up in the standards; do you think that will continue?
VK: It’s a fairly complex issue; the science is very uncertain. I think it is figured into the California Low Carbon Fuel Standard. The end result is a reasonable compromise. It’s also something that is fairly uncertain right now. I think the California Air Resources Board (CARB) came up with something that’s a reasonable answer on indirect land use impacts. The corn ethanol guys wanted to have zero. They didn’t get that, so they are now complaining in Washington. I think CARB could have phased it in more slowly because the numbers are so uncertain, so I would not agree 100% with CARB. But I would agree 90% with them.
Q4 (59:10): That’s corn. How about cellulosic?
VK: I think cellulosic should be measured the same way, but I think the impact will be fairly small, and over time it has the potential to be the biggest opportunity to sequester carbon in the soil. I don’t want to get into the details – there are papers on my website about this – but it is possible to change agronomy practices to raise biomass and sequester carbon at the same time. It is the annual crops, where you till up the soil ever year, that you have a problem. Perennial crops, and sugarcane is such a crop, you have a much better chance. Also, a lot of cellulosic crops can be grown without a lot of water and on marginal lands.
EC (60:20): So the amount of land we would need, if we were to truly replace gasoline, how much land would we need?
VK: Under optimistic scenarios we need zero land in this country to replace all of the gasoline in this country. (RR: He referred to this paper – Where Will Biomass Come From? – on his website for a detailed explanation). Look, this is really important. We can’t do linear extrapolation of the past. (RR: Because it doesn’t give the desired answer). If we do, we are sure to fail. We have to do things a new way. The best way to predict the future is to invent it, not extrapolate the past. (RR: Audience starts to applaud). And this is a fundamental difference.
Q5 (61:22): Is the lack of seed capital – especially in Europe – a bottleneck, and how do we reengineer this so that funds are available?
VK: Lack of seed investment in Europe may be a problem for the Europeans, but it’s an opportunity for us. Let me give you an example. I ran into a guy who was a senior director of research at Exxon, who had moved to Europe – Amsterdam – and was struggling with a new idea to make fuel from biomass. He wasn’t producing ethanol. He called me, and said “Nobody in Europe understands me. I have been looking for money for two years.” He had been begging and borrowing space at various labs and universities to do his research. He said that he thought we had it all wrong, that instead of turning biomass into ethanol you should turn it into crude oil. This is exactly the same thing nature does; all crude oil comes from biomass. He said the only problem with nature is that it takes millions of years. He said he could do it in minutes. Now that’s a seed idea. I would have guessed that there was less than a 10% chance that he was going to be able to pull this off. It didn’t take very much for me to write him a check, because if he is right it’s transformative. He moved to Houston and went to work.
I like to joke that I am the only Indian in-sourcing jobs. We have in-sourced three technology companies: One from New Zealand, one from Amsterdam, and one from Australia. The same thing happened with the solar thermal technology in Australia. We funded it and they moved to Palo Alto. Every news channel in Australia carried that story. What was the story? “Why isn’t Australia funding this?”
EC (64:40): Are you seeing more competition at the seed level from other venture capitalists?
VK: It’s starting to increase, but not that much. That’s why we love the seed opportunities. They are the most promising opportunities anywhere. (RR: He then mentioned that the company in Houston is KiOR, which I mentioned previously in Vinod Khosla Scoops Me. Incidentally, VK e-mailed me after I posted that essay and we exchanged several e-mails over KiOR and some of his other ventures.) Nobody wanted to invest in the Internet until the Netscape idea. After Netscape, everybody was interested.
EC (65:40): You have said that you like being a seed investor. Do you think there are enough investors at the 2nd and 3rd tier? These companies are going to need more than just you at that point.
VK: You don’t know for sure, but we see increasing interest. If you see one or two successful IPOs, the amount of money will increase dramatically. Wall Street bounces between fear and greed; we are in a fear cycle.
Q6 (66:40): What are those Ph.D. students looking into right now? In 2005, I did an informal survey at UC Berkeley. Nobody in the engineering department – graduate students or professors – were interested in energy. We did an informal survey in 2006 and suddenly more than 50% were interested in working in energy. That’s why I am very bullish with respect to the new crops of Ph.D. students coming out. It’s the number one choice. Number one used to be nanotechnology, genetics, computer science; it’s now material science, it’s chemical engineering, it’s all kinds of fundamental processes. What I have noticed is physics, chemistry, biology are becoming a lot more important, and that will drive transformation in energy over the next 20-25 years. (RR: I guess I was way ahead of my time since I studied biomass to energy in graduate school at Texas A&M in the early 90′s).
Q7 (68:00): I agree with your urgency about climate change, but it’s interesting to think about other countries, who already realize that we have already baked in about two degrees C in terms of the thermal momentum of the earth. Is there a technology opportunity in adaptation to climate change?
VK: I haven’t spent enough time on adaptation. It’s unfortunate that the people who have the least are the most impacted, like Bangladesh. But there is an interesting area that I have avoided, called geoengineering. I have just been asked to speak at a geoengineering conference, and I haven’t decided. It is a touchy subject; to engineer the climate of this planet. Some people think we have to do it, others think there will be too many unintended consequences. I subscribe to that view.
EC: We will take two more questions.
Q8 (70:22): Could you talk about job creation?
VK: Most of the studies say that job creation per dollar invested is higher for renewable technologies; higher than dollars invested in fossil fuel technologies. I don’t know why that is, but all of the data seem to indicate that this is in fact true.
Q9 (71:52): Do you think Brazil has a chance with sugar ethanol?
VK: Sugarcane ethanol, under the Low Carbon Fuel Standard, comes out looking reasonably good. But, having said that, I think sugar is too valuable a commodity to use. We will get to things other than sugarcane as our source of fuel. I suspect sugarcane will be more lasting than corn ethanol, but even that will be a passing phase. In the end, non-food technologies are likely to be the source of our fuels. Partly because the politics are right; more importantly because the science is right. I evaluate biofuels on one metric: How many miles can you drive per acre? With most food crops, you can get to 10,000 miles driven per acre. Cellulosic technology offers the opportunity to go 100,000 miles on an acre, and then land becomes a non-issue. (RR: Two words: Net energy). Now we promised to take one last question.
Q10 (73:30): A lot of these new technologies are going require someone to install all of this. Are there plans to look at human capital opportunties?
VK: There are clearly opportunities in services. We are not funding them because, partly because I am a techie nerd; I like the technology and everyone should do something they have fun at. But there are clearly opportunities, and others are doing it. Thank you all very much.