Interview

The Ultimate Bug Zapper

The Ultimate Bug Zapper

Nathan Myhrvold has an idea to stop malaria: a laser device that would zap disease-carrying mosquitoes before they had a chance to bite. Like an arsenal of other out-of-left-field ideas that Myhrvold’s firm, Intellectual Ventures, is working on to combat the disease, the concept sounded more like Star Wars than a development project. But Myhrvold — whose prior experience includes 14 years as Bill Gates’s ideas guru at Microsoft — and his colleagues built it out of ordinary electronic devices they bought on eBay.

A theoretical physicist and computer geek might seem like an odd candidate to take on a disease that for years has mostly been fought with bed nets and anti-malarial drugs. But that’s exactly the point: Myhrvold and his band of outsiders, working through what they call "invention sessions," are bringing fresh eyes to a problem that has confounded decades of development work. In an interview with FP’s Elizabeth Dickinson on the occasion of World Malaria Day, Myhrvold explains why out-of-the-box ideas might be exactly what’s needed to tackle the scourges of the developing world.

Foreign Policy: How did you first get involved in malaria research — and why malaria?

Nathan Myhrvold: Most of what technologists do is to push technology forward, which is a wonderful thing, but it’s about making toys for rich people. We wanted to do some stuff that would really have impact in the developing world.

Malaria is interesting because it is truly a disease of the developing world — even more than HIV is. Because HIV occurs in the rich world and it’s a popular cause, it gets an enormous amount of research and development dollars. Malaria doesn’t. We thought there might be more opportunity there, and boy were we right. Once you start getting good at stuff on it, you say, "Look, we can do this too — oh my God, we can do this!"

Depending on who you believe, there are 300 to 500 million cases of malaria a year and at least 1 million deaths — most of those are children under age 5. So if we could make some leeway there, it would be really important.

FP: Take me back to that first brainstorming session on malaria.

NM: The Gates Foundation ran a conference on malaria, and Bill asked if I could come. I wound up sitting in a bunch of the technical talks, and I was struck there by how much computing could play a role in really understanding malaria. I decided that there was a new technical approach [to take] called Monte Carlo modeling. It’s [a modeling system] used on Wall Street and in physics research, but it’s not used very much in epidemiology. Using enormous computers, supercomputers, with thousands of processors, [we can] simulate malaria more accurately than it had ever been done before.

[Then] we said, we should look at other ways in directly intervening. There’s two ways you intervene: One is with mosquitoes; the other is with malaria parasite. The malaria parasite is a much more complicated beast in many ways than a virus or bacteria. One of the wonderful things about it from our perspective is that it has a material called hemozoin, which is a very interesting crystalline substance, and it’s totally essential to the malaria parasite’s living. Using that molecule, there happened to be a way in [through which] physicists and optics people could do some very cool work. The most dramatic one is that we’ve built this machine that tracks mosquitoes in the sky and shoots them with lasers, which sort of sounds like a science-fiction fantasy. We thought it was initially, but — damn! — we built the thing and it works.

FP: When you’re coming up with these ideas, when does cost-effectiveness factor into the picture?

NM: We hear about it all along. We also have a strong sense of what’s possible with technology. Your cell phone has computing power that would have cost $1 million to make a few years ago. Giant things in cost-benefit changes can occur. There are people who are [already] trying all of the simple, cheap things. If malaria is going to be defeated by giving away an insecticide-soaked bed net, then we’re late to the party. Part of our whole idea is to say, we are technologists doing highly technological things. Having us volunteer to get out the bed nets is not going to help. Having us think of some outside-the-box ideas just might help.

When you have an area that is terribly starved for resources, as malaria is, the people in that field tend to get very conservative. It’s human nature: If you’re down to your last $10, you’d be pretty conservative about how you spent it. The paradox is that we’re repeating the interventions we’ve been doing for the last 20 years; it’s highly unlikely to do any better than it’s done for the last 20 years. If we don’t have an out-of-the-box solution, what are we going to do?

And we think that there’s room for us to go and do some crazy outside-the-box thinking. Suppose that it’s 100-to-1 odds that we actually come up with something. Hey, a 1 percent chance of solving a problem of this magnitude — that’s pretty good! When I talk about this sometimes, people say, "Is it really realistic to put lasers in villages in Africa?" And [the answer is that] it may well be, under certain circumstances.

The environmental movement is a wonderful thing, but one of the unfortunate things is that it caused DDT to get a bad name in malaria eradication. I don’t want to be rude to Rachel Carson, who did a great thing. Outside my house in Seattle, you can see bald eagles all over, and they survived because we stopped spraying DDT. But remember, that was for agriculture, and for agriculture it wasn’t that important, so thank God that we saved the bald eagle. The trouble is, [now] you can’t spray DDT in Africa without people getting all upset.

Malaria is the only disease to ever be locally eradicated without any medicine. It turns out that if you drain the swamps, spray with DDT or other pesticides, put window screens on [the windows], and have a tremendous amount of discipline, you can defeat malaria. That’s how it was defeated in the United States in the 1930s. In 1935, the peak year, there were 135,000 cases. This was a furious disease in the southern United States. We got that fixed. The trouble is, we got that fixed because we had a combination of a high standard of living and a lot of resources and some discipline. The countries that have it worst [with malaria today] have very low standards of living and very little societal discipline. They can’t put their effort into malaria right now.

So the question is: How can I eradicate malaria cheaply and sloppily, frankly? If I can do that, then I’ve got a real option. Well, this computer program lets us do thousands of experiments to figure that out. Amazing as it sounds, no one knows what strategy to take, and certainly no one knows what strategy is cost-effective.

FP: What are some of the things that the model tells us about how we can combat malaria — things that we aren’t doing now?

NM: It gives you the ability, for example, to [model] tradeoffs between options. Suppose a bed net costs $5, indoor residual spraying costs about $1.50, and malaria pills also cost you a certain amount. How should you trade those off? Also, what penetration do you need to get with a combination of [those interventions] to succeed? It turns out you can get 50 or 70 percent penetration with the nets relatively easily. But you can’t get much above that no matter how hard you try. So some fraction of the homes won’t take [the net] or won’t use it. We have this whole collection of pictures of people fishing with them. The most amusing one is of a woman getting married, and her bridal veil is a malaria net. One of the bad things about a malaria net is that it restricts airflow a little bit, so on the hottest, most humid nights, people don’t use them. And guess which nights you should worry most about!

FP: When you’re thinking about the model and the interventions, are you also thinking about who the ideal clients or customers are?

NM: Yes, very much so. In the case of running a model, it’s for people who are seriously planning malaria eradication or control schemes.

One of the things we’re working on is a small, portable, cheap malaria diagnosis machine [that works like] the glucose meters that diabetics use. You put a drop of blood in, and it says "bing": You’ve got malaria — or not. Today it takes hours and a trained technician who can make a mistake. As a result, most people who have malaria in Africa aren’t diagnosed. If you give them the pills and they have another tropical fever — well, too bad, they’ll die of it. So if we could make a little portable thing like those diabetic glucose meters, I believe we will get the Gates Foundation or some organization to order them in enough volume that someone will actually manufacture them, and I think we could really move the needle. Our second generation [diagnosis machine] won’t even need to prick your finger; it will be totally noninvasive. If it’s something that requires no training and two AA batteries, I think we’ll get it deployed.

There are other, more complicated things like our laser that shoots mosquitoes out of the sky where we’ll have to work with some development partner. Although, it is amazingly cheap at the moment. When we first proposed this, people said, "Oh, it’s like the Star Wars program; it’ll cost you tens of millions of dollars!" But it turns out that every cell phone has a processor in it; every digital camera has a little image sensor in it. All those things that people thought would be so expensive — we just bought on eBay. Still, that project needs to have a commercial partner who will work with us for some application. That application might be malaria. The first one might actually be in a developed-world context. If you want to sell your crops to Whole Foods, you can’t use a crop-duster, but if you use a laser crop-duster — either by flying in an airplane or [a drone], or if you had it on a pole at the edge of the field — you can control your pests in a totally environmentally friendly way.

FP: Looking forward, if there were one goal you wanted to push for in the coming year, what would you say?

NM: We have to get our little diagnostic device to work. We have some other things, but if we could get that to work, that’d be a hell of a year. More generally, as we succeed with these things, we need to find partners for them. We’re not going to manufacture the malaria devices. We need to find someone else who can.