Liz Gannes

Recent Posts by Liz Gannes

DARPA’s Regina Dugan on “The Nation’s Elite Army of Futuristic Techno-geeks”

Dr. Regina Dugan is director of the Defense Advance Research Projects Agency, where she researches, develops and demonstrates high-risk, high-payoff projects for the current and future combat force, prevents strategic surprises for the U.S., and creates strategic surprises for its adversaries. She took the stage for an interview with Walt Mossberg Wednesday morning at D9. Here’s our live coverage:

9:10 am: Walt takes the stage. “Sometimes people wonder who invented the Internet, and the answer is DARPA.”

Regina Dugan: Well, not me personally.

Once a month we have T-shirt and jeans day at DARPA, Dugan says, explaining her rather casual attire. She says the “I cubed” insignia on her T-shirt stands for “impossible, improbable, inevitable”–the natural progression of things.

Dugan: “Our singular mission is the prevention and creation of strategic surprise.” But she won’t reveal any surprises.

9:13 am: Dugan is the first woman to lead DARPA, and previously was a program manager there. She worked on a project on explosive detection based on a dog’s nose. It involved studying things like photoluminescent polymers, and also traveling to mine removal sites in Mozambique, where they drive vehicles through land minefields to blow mines up. She says there was “a black humor to it.” But, “‘the jumpers you never get used to, though, the ones that explode waist-high and hit the windshield.”

Walt: How does DARPA relate to the technology industry?

Dugan: We’re agnostic, but our currency is national security. We look for ideas in universities, small businesses, large businesses. Project managers join DARPA for three to five years and focus and catalyze bridging of ideas across communities.

Walt: GPS and ARPANET started for security purposes, but evolved elsewhere. Do you still have work now that will have that kind of impact on civilian society?

9:21 am: “It’s no small coincidence that the things we would advance would have cascading benefits for society as a whole….We’re working on technology now that would help us to produce vaccines in tobacco plants, and might have implications elsewhere.”

DARPA pilots have millions of tobacco plants growing where we hijack protein production to make vaccines.

Walt: In Silicon Valley there’s less stigma about failure, but that’s not true in the government. Can you take risks?

Dugan: It’s part of the fabric of the agency, this desire to have a big success. That means they can’t fear failure. We have a hypersonic program at DARPA, Mach 20–could get you back to D.C. in an hour, but it might be a little warm.

It’s a big, big reach. Last year we had a flight in August, and we didn’t make it the whole way, but got nine minutes of Mach 17-plus data that we didn’t have before.

“Failure isn’t the problem, it’s the fear of failure.” The nerve you need for success is the exact same nerve until the moment it’s going to be [a failure].

On working with the rest of the defense department: “The reason we can make big reaches is because others work the evolutionary advances.” We have a gang of five too: Army, Air Force, Navy, Marine and DARPA.

Walt: When you look at security issues of the civilian Internet, and the implications for foreign policy, does this make you nervous for national security?

9:30 am: “I think with every opportunity there’s a flip-side risk.” We pay a lot of attention to cyber security and we have a large investment there. We looked back over the state of affairs and observed that if you look at the number of attacks they continue to grow exponentially. We’re in this battle where we’re not convergent with the threat. It’s as if we’re in the middle of the ocean and we’re treading water. If that’s your only strategy, it’s going to be a problem.

We looked at the number of lines of code in security software combined to malware over the last decades. Of the 9,000 different viruses, etc., we looked at, there have consistently been about 125 lines of code.

We have a program called DARPA PROCEED to develop fully homomorphic encryption. So it’s never decrypted. For three decades people used this as an example of something that’s impossible. About a year and a half ago Craig Gentry did this, and then there was another version that was simple and functional. But it’s too slow–about 14 orders of magnitude too slow. If we could accomplish that, it would be a breakthrough for everyone.

Regina Dugan

9:35 am: Walt asks, have we ever lost war plans that would help an enemy? Dugan says she can’t answer.

Dugan also says policy-making is not her responsibility.

Walt asks if DARPA was asked to help with WikiLeaks. Dugan replies: “DARPA has one of the highest densities of significant, technically trained people in the government. We have 120 program managers that are best in class. And we don’t have any entitlements at DARPA.” But she won’t comment directly on WikiLeaks.

Dugan: We have been investigating this whole prospect of how do you design software and hardware inside a computer so it can evolve based on its own experience with a threat. So it’s modeled after the human immune systems. And it manages small threats in ways invisible to the user.

Dugan says DARPA absolutely talks to companies like Google about new platforms and opportunities. One way to interact with the private sector is to transition technology out to the commercial world and then buy it back.

Audience question: Do you talk about this with the President as well, or just the Defense Department? Dugan: We deal with four-star [generals] and civilian equivalents. I have not personally interacted with the president but I would love to tell him about our hypersonic programs.

9:46 am: Audience question: What’s an example of something you came close to realizing but had to pull the plug on? Dugan: An awful lot of our time is spent on this very difficult task of choosing–how to choose between that which is a good opportunity and that which is a better opportunity. We look for places where we are underinvested. An example: In ISR (Intelligence Surveillance and Reconnaissance) we’re swimming in sensors and drowning in data. So we made some decisions to stop doing some of that to maximize speed.

Walt: The potential for data mining creeps me out.

Dugan: I understand the concern, and it’s something that we take very seriously. I don’t think anybody has great answers but it would be inconceivable given the nature of our work that we wouldn’t be investigating areas that make people uncomfortable. But we know we need to handle so-called “creepiness” ethically and responsibly.

In response to an audience question about talent, Dugan says she likes to call DARPA “the nation’s elite army of futuristic techno-geeks.”

Dugan tells an anecdote about students who came in for 90-day program, had a breakthrough at day 82, decided to keep going and three weeks later found themselves in a forward-operating cell in Afghanistan.

Ending line: Dugan says, “I’m at DARPA to serve my country.”

Latest Video

View all videos »

Search »

D Conference Mailing List

Sign up for News about D Conferences

I think it was the right decision for Cisco, but that doesn’t mean I’m not disappointed.

— Flip founder Jon Kaplan on Cisco’s decision to scrap the Flip.