Defense research agency seeks return to 'swashbuckling' days
The Defense Advanced Research Projects Agency-DARPA, originally ARPA-was created in 1958 as the U.S. response to the Soviet launching of the Sputnik satellite. The agency, which among other things developed the technologies leading to the commercialization of the Internet, works to create cutting-edge military technologies.
DARPA Director Anthony Tether was appointed June 18, 2001, and has worked in government and the private sector, serving previously as director of DARPA's Strategic Technology Office from 1982-86. Tether recently spoke with William New of National Journal's Technology Daily.
Q: How does DARPA relate to the Department of Defense?
A: We're a defense agency. We report to the secretary of Defense, as does the Army or the Air Force. ... What this secretary of Defense [Donald Rumsfeld] has done is, from a policy viewpoint, he has decreed that 3 percent of the Defense Department [budget] would go into what is known as the science and technology budget. He also has decreed that 0.8 percent, or roughly 25 to 30 percent of the S&T budget will go to DARPA.
He wants ... an organization that constantly brings forth new things that other people aren't doing. And that's generally DARPA.
Q: You've been here about a year. What were your goals coming in, and are those evolving?
A: Well, the goals I had coming in ... were two things. One was that [the top defense officials] all looked upon DARPA as an organization that was always doing the craziest things. But they felt that DARPA, over the last administration, had really lost a lot of its forward-looking luster.
What they wanted was to not transform DARPA to anything new, but to transform it back to something old. But they wanted it to become like it was in the early '70s and '80s, where it was a swashbuckling place, where program managers were constantly getting the director in trouble with ideas and never taking "no" for an answer.
The other thing was space. DARPA was born in the space age. The other goal was: Go do innovative things in space.
DARPA really does not have an organization in a classical sense. It's 140 program managers all bound together by a common travel agent. Our program managers are only here for four to five years. [Program managers] come up with an idea, they create the program plan, they compete internally for the money. We don't do anything within DARPA; we contract all our money out. They go out and get contractors who execute that plan.
Q: Are you trying to create the next Internet?
A: We're even doing better than that. ... We still have the original ARPA/DARPA thrust of creating a cognitive computer system.
Now we cycle forward 30-some-odd years. In 1965, there were like 50 transistors on the chip. We are approaching a billion transistors on a chip, a chip being something like a third of an inch by a third of inch. This approaches the density of cells in your brain.
Q: How will you make your mark on DARPA?
A: It's the easiest place in the world to change. The reason is we turn over people at a rate of 25 percent a year-which means after I've been here a year, 25 percent of the agency will be people that I personally had a hand in hiring. ...
Q: How has DARPA's mission changed since Sept. 11?
A: DARPA's only real stated charter is to prevent technological surprise. That has not changed, except we have modified the vision so that now we feel our job is not to just prevent technological surprise, but to create technological surprise.
DARPA was [already] working the terrorism and counterterrorism problem. We were developing the capabilities to be able to take inputs, financial transactions, and come up with networks-who's talking to whom.
We also had behavioral-science types of projects where we were trying to predict what a group would do. We also had efforts in foreign-language translation. We had people who were already interested in the subject; we just put them together in the same location and found an office director [former Reagan White House National Security Adviser John Poindexter].
We are going to develop the capability to detect, track, figure out intent and pre-empt terrorists.
Q: What do you need Congress to do?
A: We need Congress ... to understand that patience does pay off. Having an organization like DARPA, it's constantly developing new capabilities that may not turn into real capabilities for five years, 10 years or more.