meanwhile there’s this multi-year delay. So if you ask us 3 years from now, we should have 1,650 patents, roughly speaking.
We’ve done some great licensing deals, we’re very happy with the revenue that we’ve gotten. On the other hand, we’re fundamentally constrained by the fact that [for] most of our best ideas, the patent hasn’t issued yet. It’s hard to get someone to pay you for something where maybe it’ll be there. Plus we deliberately work about 5 years into the future. Another part of our theory here is that we’re about hitting home runs. So much of the world is people thinking incrementally because it’s less risky. People that make products ought to think that way—I’m not saying it’s a bad idea—but I think there’s too few people who really have a blank piece of paper and can start from scratch and think really, really big thoughts. So we try to do very ambitious things.
In our private equity-like business, where we invest in other people’s patents, that is much more straightforward, and so there we’ve gotten a whole lot more assets and they’re much further along because they exist before we buy them…Which is one of the reasons we’re doing both projects. In that context, we’re still feeling our way, we’re still making the model up. We’ve got thousands of assets, but we’ve still spent more than we’ve collected.
We have a lot of techniques to stimulate inventions. But our first invention session was not very ambitious. We looked at digital imaging, which in 2003 was already pretty well developed. We came up with some cool things, and a bunch of those have issued. The world has caught up very nicely with the direction we were going—things like ultra-high-precision analog-to-digital converters—with the basic notion that Moore’s law will continue. Maybe it won’t, but I’m willing to bet on it. But most aspects of imaging don’t follow Moore’s law, so how do you use processing power and chip technology to overcome cheap lenses [e.g., by using software algorithms to enhance images—Eds.]? It turns out there’s a bunch of ways you can, so we have some patents in there.
One of the things we like to say here is we don’t have any commitment to a single problem. The first time I heard this was from one of the founders of Costco, many years ago. I asked what their principles were, and one of them was interesting. It was, “We have no commitment to any product.” I said, what does that mean? He said, “We have a commitment to the customer instead.” So Costco will only sell Pepsi or Coke if they can get you a deal on it. If they can’t get you a deal, screw it, it’s not there. A grocery store has a commitment to every product category. They inverted that. And it gave them flexibility, and it was OK with customers. Well, we have no commitment to any problem. I can’t promise you I’ll invent a great thing to cure cancer. I’ll try. There’s a whole mythology about persistence. I love persistence, it’s very important for some things. But we don’t beat our heads against a wall forever. We’ll give the wall a few good cracks, and then we’ll move along and try to find a softer spot in the wall.
X: Can you talk about the specific role of the new Intellectual Ventures Lab, and some of your favorite projects?
NM: The lab here is one of our more recent projects. At a certain point you need to be able to do a certain amount of direct technical work. You can do a tremendous amount with thinking, pencil and paper analysis, and computer modeling. We love computer modeling. Boy, you can do stuff in computer modeling that would be extremely expensive and difficult [otherwise]. Nevertheless, for some things it’s easier and faster, or more compelling [to do it directly], which is why we have the lab here and we’re very actively engaged with the lab here.
My single favorite project at the moment is our nuclear project. That’s our most ambitious one also. We have invented a fundamentally new kind of nuclear reactor. Which is a nutty thing for a little company to do. But we’ve persuaded some folks that we’re in a position that it makes sense to try it. It’s still very high-risk, but the leverage you could get from having a new carbon-free energy source would be fantastic. My recommendation to the world is, fund 100 really cool new carbon-free energy sources, and fund some that have some diversity—not 100 people trying to do photovoltaics. We have a better chance to get salvation that way than almost any other standard algorithmic approach. Most energy policy things in Washington [DC] involve money from the Department of Energy to do research, and almost always that stuff goes to the usual suspects, and produces nothing. If you really want to stimulate new energy ideas, you need to find a way to get ideas stimulated at a grass-roots level from lots of folks.
Now venture capitalists would say, “Oh yeah, absolutely, I’ve got all that, that’s why I’ve funded all these green-tech things.” And I’d say, no. It’s great that you do that, but everybody you funded had the idea before you met them. What a VC does is turn funding an idea into a company, it does not fund creation of the idea. Now sometimes you get a better idea after you’ve started the company, so accidentally it becomes funding creation of an idea, but that’s not its goal. If you really want to fund creation of ideas, you have to organize differently, and that’s how we’ve tried to organize.
X: The nuclear reactor idea came out of an invention session a couple years ago, and with 30 people working on it, is your largest single project. What stage are you at, and what are the challenges?
NM: The work we’re doing right now is to figure out the technology at ever increasing levels of detail. Now, we could discover that we hit a technical roadblock. And in that particular one, we could massively hit political or regulatory roadblocks, so that’s another factor, which is what makes nuclear tricky. Finally, just because these projects are [of an] enormous scale, we could hit a roadblock that’s just time—oh yeah, everything’s going to work, but it’ll take 20 years to build the first one. If you say that within the nuclear reactor community, they’ll say, “Well of course it should take 20 years to build a new one.” And I’m saying, “No I’m impatient! I want it now, the world needs it now, let’s rethink that.” We’ll see. We might be able to rethink it and get something built in a much more rapid time scale. Or we might not.
We have a whole bunch of physicists and engineers trying to do the next level of design. We’ve got the advantage of computer modeling, which can work very well there. An enormous amount of effort has gone into making great computer modeling tools. We’ve developed some of our own, but we’ve also been able to inherit a tremendous amount from the past. The converse is you really want to make sure this shit works in the computer first, because of the consequences of a mistake! So right now, we’re doing paper [studies] and mostly computational studies to really understand all the aspects of the reactor.
I am hoping that we could get a reactor built inside of 10 years. Now you could quibble about, is that a test reactor, is that the first commercial reactor—and once you ask that level of detail, it’s beyond the resolution that we have now. Of course, to have it built in 10 years, we have to start designing it in 3 years, because it takes a couple years to design it, and then you have to build it. It’s a long process.
Stay tuned for Part Two of the Q&A tomorrow, where Myhrvold talks about tackling global warming, malaria, cancer, and invisibility; what he learned from founding Microsoft Research; and the upcoming Asian expansion of Intellectual Ventures—Editors.
