Apoyan a Conec




“DARPA’S failures are the ones that did not leave a foundation of good research”

American journalist  Sharon Weinberger (San Francisco, 1972) is one of the leading experts in science and national security in her country. At the end of April she visited Valencia invited by the Institution Alfons el Magnànim to launch her latest book, The Imagineers of War, on the history of the Agency for Advanced Research Projects of Defense, known by its English acronym DARPA. The presentation was held in the MUVIM by Vicent J. Martínez, professor of Astronomy and Astrophysics at the University of Valencia and Conec’s director, Carlos Peña, research scientist at at the Institute for Integrative Systems Biology (I2SysBio), and Raúl Jiménez, research professor at the Catalan Institution for Research and Advanced Studies (ICREA).

Founded in 1958, this Pentagon’s scientific body was created by the Eisenhower Administration as a technological response to the challenge of launching the Soviet satellite Sputnik. Some of its initiatives have changed the world as the ARPANET project. Others ended up in crashing failures like the Vietnam War. But to DARPA is due from the gestation of the first nodes of Internet to the autonomous cars and the armed drones. So it should not surprise anyone that Weinberger makes a positive assessment of the agency’s history, which today, according to this journalist, is missing major objectives.

For its failures and extravagant bets, does DARPA deserve being called “the Department of Defense’s mad science«?
— I disagree with this characterization. I find it is not a helpful way of thinking about DARPA. It is absolutely not supposed to be mad science but as a bastion of scientific reason to try far-out ideas, technology and ambition. It is sort of a shorthand way, but it is actually the way to think about the agency, in some way it is counterproductive, because it can excuse the time DARPA has persued just for ‘mad science’.
The question is, and there was always a tension about, if DARPA is a national security agency that does science or a science agency that does national security. Which should be the focus? There was one of the scientists I described in the book,  Nicholas Christofilos, who was sort of a favorite of DARPA. He was involved in the project called Seesaw about a particle beam that was going to blast, nuclear weapons out of space. I guess you call him a mad scientist, but every DARPA director who sponsored it said: “We knew it was not ever going to really work, but it was marvelous science”. At the early 1980s, Ronald Reagan said that they were going to try to make it work. That was to build this thing, which just was sort of an idea, in order to create a space-based missile shield to protect against Soviet nuclear attacks, known as Star Wars project. Researchers were shocked. But it is an example of how for an ambitious idea it is sometimes required technology which was not in access for decades, and actually it should be invested.

“If you want DARPA to be an important agency, you have give them important goals to achieve, and I don’t think they have that in the past few years”

— Why does ARPANET, the first nodes of Internet, represent for you the greatest success of the DARPA’s history?
It is what made agency’s reputation. There were many projects involved, they got some credit and many people were involved. Other agencies were involved, but really it was a DARPA’s creation, they led the great work and funding and set it up. In terms of technology that changed the world, there is no better example than the Internet. There is no aspect in the society than Internet has fundamentally transformed. It is the birth as DARPA getting proper credit for it and also it is the vast example of DARPA’s work that changed the world. And the irony is within the agency, especially with such tiny projects, it was a small part of what the agency was doing, it was not as big as other programs were. For me it is the best example for what the agency was successful.

— DARPA’s projects feed a large number of articles in press with listings about its craziest and most frightening ideas. Is this a triviliatation in order not to take military science seriously?
It is a beatiful way of putting it. People always ask me what the successful and failed projects are, but DARPA is much more complex. Let’s take an example. In the 1980s, DARPA founded the Strategic Computing Initiative, it was basically billions of dollars effort to create artificial intelligence, and absolutely you could put it in the rest of failures. There was a lot of publicity and a little bit of controversy around the program, but they never created artificial intelligence, every technology they tried to produce failed. Most of the programs and companies that were sponsored under the program went bankrupt. But it is accepted that this failure made the foundations for most of great computing science advances in artifical intelligence. It was a vehicle by billion dollars effort that failed. So, what you call ‘success’ or ‘failure’ sometimes just takes time.
Another example is the work on drones. DARPA founded the predecesor of the Predator drone associated with the wars in Iraq and Afghanistan called Amber. And it flew, but the company that aided it went bankrupt. Few years later, another company bought the bankarupted one and the CIA bought the Predator. It depends what you think of it is a success or a failure, and how long it takes. To me, failures are the ones that did not leave a foundation of good research, the failure is when the project is not the vehicle to produce good science.

“They forget sometimes they are a public institution, and they are very much, so I like to remind them of that”

— A very famous DARPA’s failure was also the National Aero-Space Plane.
— It was a single-stage-to-orbit space plane, which basically would take off from a runway, all the way to the orbit, rather than being launched by a rocket, go into space where it could be anything –a spy plane, a space bomber–, re-enter the atmosphere and then land again like an airplane. This became what Ronald Reagan called the Orient Express, which could be a civilian airliner that took you from Washington-Dulles to Tokyo in two hours. The technology was not nearly ready enough to accomplish it. The cost of this very complex program had grown to the billions of dollars, and it was eventually canceled and enlisted as a failure. It was vast research program increadibly costly. Maybe there were better ways to advance the research, but eventually they got a hypersonic airplane of some sort credited back to the DARPA’s work.

— The national air space plane probably was probably the most costly of DARPA’s failures. Is it known how much it cost every project?
We know that its annual budget. You know what the overall budget is, but it is often hard to know on individual programs to figure how much it is. Sometimes it is clear and it sometimes is not, sometimes funding is for classified programs. So, you really have to go back to call for detail.

— American public opinion in relation to DARPA, is the society proud of DARPA advances?
— I have the impression that it is positive. Of course, there are people who oppose any military work, and there are people who oppose any military research. Those people are in oppositon of DARPA, but I think overall in the American public it is a positive example, DARPA enjoys a possitive reputation now.

— Is it important to know the military roots of the current technology?
— I think so. As you look at the current debate in United States, one of the things that the March for Science is protesting for is the cuts in government’s science funding. So I have interest in that someone advocates for science and scientific researchers and funding. If the military budget increases, so probably that would be increased in military research, as during the Cold War, when there was a funding shift into the military institutions. Whether it is a good or a bad thing, it depends on how you see about the military involvement in science and technology. Certainly, ARPANET, a product of military goals, research and interest, is a good thing, but if we look at other areas, it is different. If we talk about the cuts of science funding, nothing is good or bad, we have to try to understand the implications of that shift. It is a little bit naïve or simplistic to say: “Oh, no, government is cutting funding”. It is well known they are shifting it. The government has always funded science and has always national security goals. The question is how broad and narrow the goals are. That is a different debate.

— Some research institutes in Spain refuse to collaborate with DARPA arguing that it’s not necessary military goals for the scientific advances and the case of DARPA misleads people, because the same could be made for civil purposes. Is military investigation a necessary evil?
— It depends where you stand. Any military exist to wait war. Countries have military to protect national interests. So it is a question whether it is a necessary evil. We have to look at the implications after all. There are benefits having a military investment in science, but also some huge disadvantages which I described in the book. DARPA, for example, was embarked in scientific experiments that were incredibly damaging especially in war time. One of them was the introduction of exfoliation experiments in Vietnam. It was applied science, and DARPA was very involved.

— In your book, the most important fact was the Vietnam War.
The most important is how we misunderstand the lessons of prior wars. So, I was interested in the Vietnam War period in part because of the heavily classified history. Discovering something that had been secret was interesting, but also it is so important today because a lot of the work that DARPA founded in the counterinsurgency, discovered betwen 2005 and 2008 by the US military, who started reading these facts, and did not even know had been founded by DARPA in the 1960s, and they were using them as a blueprint to apply in Iraq and Afghanistan. It is amazing. If you choose what you think is successful without understanding the failures of the contacts of where it came from, it is a disaster. And the most important for me is that the original counterinsurgency program for DARPA was about trying to come up with the way to avoid committing US troops in foreign countries, but in the 2000 it was completely undermined the entire point of DARPA program. So, understanding of history is really important, and that is why it is important for me the lessons of history over again as a very modern understanding of what is going on. We learn that we can do everything we want and spend money, technology and troops, but we cannot fix other country problems without better, stable government institutions.

— Despite of criticism, Project AGILE has survived under differents names.
Project AGILE was an umbrella name given for DARPA’s work first in Southeast Asia, in Vietnam and also Thailand, and eventually expanded as a global program for counterinsurgency. And they really looked on it as a global scientific experiment that you could scientifically identify the roots of what causes insurgency and then apply scientific solutions. They looked everything from social studies (antropology, economics, political science…) and then apply technological solutions. And, it was a failure. The best explanation I have got of why it was a failure was from one former DARPA director. He said that with AGILE, DARPA tried to take a system’s analysis approach to a country’s problem, looking at all the component parts and how they were fit together. And the only institution that can apply that within a country is a government itself, not an outside government. If the government itself is corrupt in a long period, there is nothing the US government can do about it. Five decades later, the problems in Afganisthan and Irak were the counterinsurgency for getting involved the US military.

“Right now DARPA is enjoying a kind of golden period without big scandals or big failures”

But AGILE itself is was a program shot down in the end of Vietnam War, firstly administrated by the Overseas Defense Research Office, asked for analysis focused on the strengths and weaknesses of U.S. Government organization and management in the field of counterinsurgency. Congress was becoming very critical of DARPA and Overseas Defense Research Office was unpopular. So, DARPA took all of its technologies from the Vietnam era and renamed the office the Tactical Technology Office, which has nothing to do with Vietnam, and still exist to these days and now becomes the biggest office within the agency. Many of the technologies are associated with DARPA like drones, precision weapons or  stealth aircrafts came out of Vietnam’s error research and Vietnam’s project AGILE. And that is the mostly forgotten, DARPA created these amazing technologies but they were directly related to a warfare. I argue that how we fight our wars today is largely a creation of DARPA. If we pursue wars today, it is enabled by DARPA’s unsuccesful war afford involved. Drones are an amazing and succesful techonology, but they are succesful through warfare, and it is altogether a different question. And what was interesting about AGILE is it was unsuccessful but it really did look at how you solve warfare, and that is something that DARPA doesn’t look at today. As I explained in the chapter called ‘Glorious failures and glorius success’, AGILE was a glorius failure, but it is also important what it is meant by glorius failure which tried something so amazing like Star Wars. At least, it was a noble act in theory.

— Who was the most ambitious of U.S. Presidents?
— It was probably President Kennedy. He took DARPA’s advice and interest in missil defence. The sort rocket program which started DARPA became the basis for the mission of the Apolo moon program. He also approved DARPA’s counterinsurgency program. Missil defence, nuclear tested action, space travel and counterinsurgency, those all four major areas had President Kennedy’s involvement. But I’m not saying it is good or bad, but had involvement from president Kennedy and I don’t think any president prior, which was really Eisenhower, or after had so much interest for DARPA’s work more than Kennedy. He probably had the most personal involvement.

— Why do you state that some of the most spectacular failures reveal the true reach of its leaders’ ambitions?
— One of the things I argue is that if you want to have means for success you must have means for failures too, it is impossible to avoid. DARPA had at its height things like the ARPANET and also the Vietnam War, this is what you have to decide as a country, as a government. I don’t think we have today a leadership that wants to allow DARPA to have these spectacular failures, and maybe that is good. Maybe you don’t want any failures anymore, but that’s a choice to be made, because you are not going to get another ARPANET if you are not going to risk having an agency involved. There is a tendency to trivialize this. DARPA is happily talking about ARPANET, but they don’t want to talk about the Vietnam War. As the popular saying, it would be said that success has a thousand fathers and failure is orphan. So, that could have to go back to DARPA.

“Kennedy probably had the most personal involvement with DARPA”

— What is the most controversial program in recent years?
— In recent years, it was the Total Information Awareness, no doubt. It had huge public interest involved. I think in the future the neuroscience work on the Biological Technology Office could be that controversial. I remember one of the last interviews I did for the book was with the head of the Office.  Neuroscience is the big push they are making, and I asked if they would ever classify work in neuroscience. I expected that he said no, but instead he said: “Well, yes. If we ever came upon something that provided us to seek advantage to US, as well any field, it would be classified”. He was incredibly very honest. But also when we think about neuroscience research related to human clinical studies being classified, it is a very dangerous territory and it could spark a controversy like we had with the IAO. But right now DARPA is enjoying a kind of golden period without big scandals or big failures. It always bears the question of what they were doing at this.

— How would you describe your investigation inside DARPA?
— Luckily, there are different ways. One way we have the Freedom Information Act, which allows access to government records, in theory; in practice, it can be very hard, and so I wasted three years legal battle with DARPA and eventually sued them in court and I won. The Freedom of Information Act says that I can request access to them, and initially they denied access, and then they denied it again, and I sued them in court and then they agreed to give them all. The most complicated were the national archive records. They had along decades declassified a lot of rapports, so you can go in and look files from 1959, the people’s originally handwritten notes, and then I made interviews. I started with people who were at the agency in 1958 in the first days all the way to the current director. So, I interviewed about 100 people. The book was done even in four and a half years, but I started interviewing DAPRA officials in 2007 when I was working in another book for the fiftieth anniversary.

“Many of the technologies are associated with DARPA like drones, precision weapons or  stealth aircrafts came out of Vietnam’s error research and Vietnam’s project AGILE”

— Is transparency possible in DARPA or is it still a utopia?
— DARPA had always a mixed relationship. Even when they spent declassified documents, DARPA has always tried to protect its independence. They always had the view to do their work arguing that they need to be independent and free oversight. And my point was, as all government funding supported by taxpayer dollars, they are still subject to the same law as the rest of Pentagon and there are very clear laws to see that the public have access too. They forget sometimes they are a public institution, and they are very much, so I like to remind them of that. Of course transparency is still a utopia, but I have longer better advocated for transparency, so I do my part.

— How are the research conditions at DARPA?
— DARPA, like other agencies, doesn’t use typically peer review because they want to move quickly, and in technology programs this can often work well because you can make decisions quickly, you don’t have to debate things like that. Using peer review is very problematic, because you need scientific consensus to move forward. This is the condition of applied science, and they want results very quickly.

“ARPANET is the best example for what the agency was successful”

— Has DARPA new challenges or is it stuck nowadays?
— In new areas in neuroscience, it is probably the most recent new push. DARPA was being involved in neuroscience before, but it is the most recent extension because of the creation of the new office with new funding. It is to be seen, first of all, the Trump Administration has not appointed a new director yet. It is really up to the new administration and the new director to send DARPA in a different direction. It is hard to say right now what happen with the budget, what will be the goals, like much government’s setting holding pattern to see where it goes. If the Pentagon budget goes up, and that is what the Trump administration wants, then tipically DARPA´s budget goes up a percentage. So, if their budget goes up, it is just where the new director wants to invest it. I personally hope for anything.
DARPA has been succesful when it had important ambitious goals. And I don’t think it has that right now. I have no personal hope because I have mixed view about my own thinking of what DARPA has done. If you want DARPA to be an important agency, you have give them important goals to achieve, and I don’t think they have that in the past few years.



  1. “Los fracasos de DARPA son los errores que no dejaron una base de buena investigación” - Conec - 29 diciembre, 2017

    […] ENGLISH VERSION: “DARPA’S failures are the ones that did not leave a foundation of good research… […]

Deja una respuesta

Este sitio usa Akismet para reducir el spam. Aprende cómo se procesan los datos de tus comentarios.