Vinod khosla has become the most widely recognized investor in clean technologies—those that generate or save energy with the least environmental impact. He founded Khosla Ventures in 2004 to fund new companies, after being a long-time partner at the giant investment firm Kleiner Perkins Caufield & Byers. His entrepreneurial roots stretch back to 1982, when he co-founded Sun Microsystems, which became a $7-billion workstation and software company. In a one-on-one dialogue, conducted before an audience of energy entrepreneurs and financiers at the recent GoingGreen conference in San Francisco, Scientific American’s Mark Fischetti asked Khosla (an adviser to the magazine) to assess, with his venture capitalist’s eye, which new energy innovations are most likely to succeed and why. Edited excerpts of the conversation follow.

Scientific American: One of your mantras as an investor is: If it doesn’t scale, it doesn’t matter. How are clean technologies doing in this regard?
Khosla: Wind power is scaling, but I’m surprised at how little innovation there is. What wind really needs is energy storage technology, and storage hasn’t yet started to scale. I would call most of the attempts toyish; we need something radical. Solar seems to be doing well, but way too many companies are doing the same thing. Costs are declining, but not enough to reach unsubsidized market-competitiveness—which is also one of my mantras. That’s mostly the fault of investors and entrepreneurs who are trying to do the next marginal thing as opposed to the next radical thing.

The most interesting area is the one that people have soured on the most: biofuels. Amyris just had an initial public offering that was very, very successful. Another one of our companies, Gevo, is on track to do the same. There’s enough proof that half a dozen technologies will make economic sense, some with government subsidies, some without subsidies. The ones without subsidies are the most interesting because they’ll scale infinitely.

Another mantra: Don’t invest in clean tech, invest in main tech. What is “main tech,” and why is that where the most promise lies?
Solar and wind form a very narrow definition of clean tech, probably the least interesting segment. Our firm is investing in radically new automobile engines from EcoMotors that are 50 percent more efficient, at less cost, than today’s engines. We are investing in air conditioners that use a new thermodynamic cycle. That to me is main tech. Air conditioning that doesn’t cost any more than it does now but uses 80 percent less energy—that’s worth talking about. Efficient lighting that pays for itself in the first 12 months, not in 12 years—that’s worth talking about.

We are investing in glass. We are investing in cement—the infrastructure of society, not the sort of fringe, clean tech stuff that depends on subsidies. If you’re going to create, as a community, say, 10 new Googles that are in clean tech, they are going to be main tech companies in mainstream markets that work unsubsidized.

Anything that doesn’t achieve unsubsidized market-competitiveness within seven years of starting to scale is not worth subsidizing. You may need three, four or five years to do R&D before you start scaling, but if you can’t get to market-competitiveness you’re not going to work in China or India. There are no subsidies in India. There are no subsidies for solar in Chile or Africa or most of the world. Most of the interesting energy markets are high-growth, developing world markets. If you’re not competitive there, unsubsidized, forget it. You’re a niche company.

So what is the price point for being competitive? For years investors have said new energy technology must compete with oil at $40 a barrel. Some would say $50. But oil has been well above that price for years, and the big oil states have said they want to make their basement $80 a barrel. Where is the line?
You have to look long term. Let’s say your first production plant for turning biomass into biofuel is producing fuel at the equivalent of $75 a barrel. Chances are by your fifth plant you’ll be at $60. By your 15th plant, you’ll be at $50. So the first-plant economics have to be roughly in the current price range, and then they will keep declining as more plants are built.

When you build the 50th plant, then the ecosystem around it will start to help you: biomass feedstock will get a lot cheaper; John Deere will build custom equipment for harvesting and shredding your biomass. If biofuel can get to $75 a barrel for a first plant, it won’t stop until it goes down to $30. By 2030 the price of oil, in 2006 dollars, will be $30 a barrel—not because we stopped using oil but because oil will face plenty of competition.

Biofuel investments were big two years ago. But before that it was wind and solar. Last year it was the smart grid. Clean tech seems to be a moving target.
One of the problems is that environmentalists have been very, very good about identifying the problems we need to solve. They are horrible at picking the answers. I believe most environmentalists most of the time are getting in the way of answers that make economic sense. Economic gravity always wins. That’s why I have the mantra about unsubsidized market-competitiveness.

For example, I don’t think the electric cars of today make sense economically. Take the Nissan Leaf: a $26,000 electric car, with $20,000 worth of batteries in it? Give me a break. The Chevy Volt, which is a good [hybrid] car, is expected to ship 40,000 units by 2012. The Tata Nano [a small, inexpensive car made in India] had 200,000 orders the first day. We have to make technologies like the Tata Nano, not the Volt, that are low carbon. The bulk of the growth in the world automotive market is in India and China. That economic gravity is key.

In a $100,000 car like the Tesla, a $20,000 battery pack doesn’t matter. It’s a fun, sexy car. But that owner is not a price-sensitive buyer. The bulk of the world is price-sensitive, and those people are going to drive the equivalent of the Tata Nano. If you want to solve a climate problem, make a Nano kind of car that has low carbon emissions.

Or build an electric car battery that costs one tenth as much as lithium-ion batteries do now. Almost certainly the traditional lithium-ion bulk batteries will not be around in 15 years. Nobody else believes that, so I’ll go on the record.

We are investing in solid-state lithium-ion batteries. Maybe that will work. We are investing in magnesium batteries. Maybe that will work. We are doing other unusual things; I call them “quantum nano thingamajigets.” The leading battery in 15 years will be the one that, in people’s eyes today, has a very low probability of being a winner.

This comes back to radical innovation. The companies that will win the big battle of economic gravity, win in unsubsidized market-competitiveness, will be the companies that are trying radical technologies. If an innovative idea has a 90 percent probability of failing, then I like it. Why? Because it is likely to be the one that has a quantum jump in performance. This is my “black swan thesis of energy.” Don’t look for solutions in high-probability areas. Those are all incremental. Look for solutions in the tails of the [bell curve] distribution of “likely to succeed.”

Are those outliers the focus of the two new funds you’ve started: $1 billion for energy and information technology and $300 million for high-risk experiments?
We have plenty of money to invest. The problem is finding the people who want to do something other than incremental, nonscalable innovations. You want to do biodiesel from palm oil? Great. Somebody may succeed. But is it even worth succeeding? Not likely.

Other people are trying to make magic work where it hasn’t worked. Algae for fuel? I’ve looked at two dozen business plans, and I haven’t found one where the economics work. And it’s worse than that; I can’t look at their process costs and identify the hypothetical breakthroughs that could create a five-times improvement. Now, could you build an algae company for a high-value product like a neutraceutical [a food product that has health or medical benefits]? Absolutely. Those of you doing algae, I would suggest shifting your guns to high-value products. Might work.

So there’s plenty of money. There are not enough breakthrough technologies. There are not enough great Ph.D.s in these fields. I actually add up the number of Ph.D.s in each of the companies [we fund or are considering]. I call every CEO and ask, “How many Ph.D.s did you hire last month?” This is my standard question. There’s not enough technical talent for large breakthroughs. Our universities weren’t producing many of these people until about two or three years ago, when interest in clean tech or energy technologies increased. The good news is that now smart Ph.D. students are going into these areas, so in 10 years, innovation will explode.

You say there’s plenty of money, yet critics of clean tech say it takes too much money to scale.
Absolutely wrong. It’s a fallacy. The amount of money needed to reach break-even cash flow or to reach an initial public offering or a reasonable point where you can sell the company does not look any different for our clean tech portfolio—our main tech portfolio—than it did for the last 15 years that I was at Kleiner Perkins, when we were looking at information technology or telecom equipment or enterprise software. During the 1990s a lot of companies needed $50 million to $100 million to reach break-even. Our lighting and air-conditioning companies will need equity in this range.

Are there a few companies that need $300 million or $400 million? Absolutely, but we had biotech companies that needed $300 million or $400 million. The distribution looks about the same.

Capital is crucial. But what about legislation to reduce carbon dioxide emissions? A tax on fossil fuels, a cap-and-trade system or renewable energy standards requiring that a percentage of a state or nation’s energy supply come from renewables are politically stuck. Instead you advocate a “low-carbon-electricity standard,” which would require states to reduce emissions of their electricity sources by a set amount, say, 80 percent by 2030. Why would that work?
Under a low-carbon-electricity standard, the only metric is how much carbon dioxide is produced, not which energy source or clean technology is used to achieve that goal. Each state could pick the technologies most appropriate to it, such as solar in Arizona, wind in Texas or biomass in Arkansas. States with coal or natural gas plants could use carbon capture and sequestration to lower emissions without switching fuels, which also creates a competitive pressure to improve that technology, pressure that does not currently exist.

A low-carbon-electricity standard also allows the conventional fossil guys and new nuclear plants to compete. We need everyone trying to develop radically lower-carbon technologies, not just solar and wind. Carbon capture from natural gas could be almost as low carbon over a plant’s lifetime as solar and also be cheaper. If fossil-fuel plants are not included in a plan to reduce emissions, carbon capture will never become economic. It will be limited to expensive Department of Energy projects.

The added benefit is that the U.S. would develop many economic low-carbon technologies, making the country a leading exporter of carbon-capture technologies to India and China, where much of the world’s coal-fired capacity is located.

You’re enthusiastic about carbon capture. You’ve said, “I only work on things that are intriguing to me, so I switch every few years. It becomes time to learn something new.” Which technologies are most exciting in the next five years?
Every area I look at is exciting. I’m not being facetious. I didn’t think we’d be inventing a brand-new engine type. When we looked at air conditioning, we expected to find better compressors—something marginal. But we discovered a brand-new thermodynamic cycle, developed by engineers at the Caitin company. Nobody has really commercialized a new thermodynamic cycle in 50 or maybe even 100 years. I wouldn’t have expected that.

What’s clear is that people who have been in old energy areas for 30 years are not taking a fresh look, even though the world around them has changed. Take, for example, mechanical engineering. We are still using old mechanical systems like cams to time the valves and cylinders in cars. A cam is a fixed piece of hardware, but driving conditions change, so timing should change. What you need for precise but variable timing are better power electronics devices that can rapidly switch valves on and off. Why aren’t all valves in engines electronically controlled? Why do wind turbines have to run at a particular speed with a gearbox, instead of using power electronics to convert all the energy into power at the frequency and phase you need? These are just two examples; there are 100 cases like that in power electronics that would change applications dramatically.

In every area, no matter how archaic, there’s innovation to be had. That’s a surprise to me. I wouldn’t have expected that.