The Imagineers of War

by Sharon Weinberger

  By late 2010, DARPA was touting Nexus 7’s successes within the Pentagon, but it was not clear what it had accomplished, if anything. As members of the team worked on a base crunching numbers from military and intelligence data feeds, another team of contractors, the Synergy Strike Force, was working in the provinces of Afghanistan, swapping beer for data and using crowdsourcing techniques honed in the red balloon hunt.

  —

  The Synergy Strike Force was always more a concept than a formal organization, an improbable mix of humanitarians, hacktivists, and technophiles who had set up shop in Jalalabad, Afghanistan, at the Taj Guesthouse, which had previously been occupied by Australian mercenaries. The eclectic group included techno-enthusiasts who wanted to bring the Silicon Valley ethos to Afghanistan. There were a few “burners,” attendees of the annual Burning Man festival, but there were also scientists, security contractors, and a dedicated group from the MIT Fab Lab, short for “fabrication laboratory,” which was building technology, like solar power and Wi-Fi networks, using do-it-yourself engineering.

  For a while the Taj was something of an informal meeting place for Westerners in Afghanistan, or “the tiki bar at the edge of the universe,” as Smári McCarthy, who was part of the Synergy Strike Force, explained in a video interview. McCarthy, a self-described information freedom activist, called the Taj “a little oasis on the edge of Jalalabad where you’ve got this strange mixture of military people, private security contractors, NGO people, and crazy people who are out there to try and build infrastructure in their time off. All sorts of people who under normal circumstances would never meet.”

  When the Synergy Strike Force took over the Taj, the tiki bar began attracting a mix of misfits, artists, and do-gooders, if for no other reason than that it was probably the only bar in all of Nangarhar province. The motley group, described as “super-powered geeks,” set about building do-it-yourself Wi-Fi networks and other small-scale tech projects for Afghans. It was sometimes hard to see what united them, other than the belief that open-source technologies could, if not save the world, then at least substantially improve it.

  In 2010, around the time that DARPA was thinking about data mining in Afghanistan, Todd Huffman, one of the leaders of the Synergy Strike Force, was introduced to DARPA officials at a chance meeting in Washington. Huffman had recently returned from Haiti, where he had been working with Ushahidi, the open-source mapping organization that helped locate victims of the 2010 earthquake. Huffman, a bearded devotee of Burning Man, whose hair on any given day might be died in shades of red and yellow, started talking about crowdsourcing in Haiti and similar work he had done in Afghanistan during the elections. It sounded similar to what DARPA wanted to do with the More Eyes project. Soon, Ryan Paterson, an official in charge of the agency’s newly formed field unit in Afghanistan, showed up at the Taj to spend a month with the burners and anarchists; he even tended bar.

  The Synergy Strike Force was perhaps best known for its Beer for Data program, but it had also done crowdsourcing work in Afghanistan to spot election fraud. Unlike the young computer scientists who sifted through Nexus 7 data from the confines of a military base, the Synergy Strike Force would be on the ground—outside the wire, as the saying went—working with Afghans to collect data. “We were referred to as those weird DARPA people,” said a regional coordinator for the More Eyes program. “Weird for DARPA is a real accomplishment.”

  Soon, DARPA was sponsoring mini-versions of the Network Challenge in Afghanistan. Under More Eyes, members of the Synergy Strike Force in 2011 fanned out over Afghanistan, handing out cell phones to participants in contests to map out areas in the provinces of Nangarhar and Bamiyan. Afghan participants, often drawn from the humanitarian and development community, were provided with GPS-enabled phones and instructed to mark the location of buildings and streets. Like with the red balloon contest, the experiments often had an economic incentive: winning teams got to keep their cell phones. Participants were not told that More Eyes was intended to provide the military with intelligence, and DARPA never publicly announced the program.

  While some of the experiments involved collecting information on politics or health care, the focus was on gathering data useful to military operations. “Generally speaking, U.S. forces have been very successful at intercepting cellular calls and incorporating them into our intelligence framework,” an undated report by one Beltway defense contractor noted. “However, these operations are just the tip of the iceberg of what can be done through cooperative techniques such as crowdsourcing.”

  The crowdsourcing projects promoted by groups like Ushahidi in Haiti were dedicated to humanitarian operations, sometimes even in cooperation with the military. But More Eyes laid bare the overlap between crowdsourcing and intelligence collection. According to an unpublished white paper written by DARPA’s Paterson, crowdsourcing would, for example, allow an Afghan citizen to report an attack on a convoy. That report might then cue a drone and eventually a military strike. Paterson also noted that More Eyes worked directly with the Defense Intelligence Agency on a project called “Afghanistan Atmospherics,” which involved using “selected local persons to passively observe and report on things they see and hear in the course of everyday activities.”

  Paterson described More Eyes as a way “to catalyze the local population to generate ‘white’ data useful for assessing stability at multiple levels (regional, provincial, district and village).” The advantage of this white data, as opposed to the black world of intelligence, is that it is “generated spontaneously by the local population…untainted by influence of outsiders.” In other words, More Eyes was recruiting unwitting spies.

  DARPA was clearly concerned that recruiting local Afghans to provide intelligence could be viewed as citizen spying. More Eyes documents warned against using foreign phones “that stand out due to their appearance or advanced functionality and can be an indicator of collusion with foreigners and can invite threats from local insurgents.” The DARPA white paper suggested that the phones be equipped with an application that can delete data, either by the user or remotely, presumably protecting the information from discovery by insurgents. The members of the Synergy Strike Force at the Taj, some of whom blogged regularly about their experiences, never publicly mentioned the Defense Department, perhaps because many of the team’s hacktivists and technophiles found it difficult to reconcile their self-image as development workers with being military contractors.

  The Synergy Strike Force was a bizarre cultural convergence. As hackers around the world were chafing at the American government’s attempts to crack down on self-proclaimed information freedom organizations, like WikiLeaks, the Synergy Strike Force’s information activists were working on a project to help the defense and intelligence communities collect data in Afghanistan. The group touted its work with headlines like “Afghanistan’s DIY Internet Brings the Web to War-Torn Towns,” and it used DARPA money to pay for solar panels for local universities, but More Eyes was really about intelligence collection. Though the Beer for Data effort was never formally part of the DARPA program, the Synergy Strike Force happily offered the one-terabyte hard drive to the Pentagon. Even the DIY Internet was an opportunity to mine data, providing a treasure trove of Internet traffic in Afghanistan that the NSA could only dream of collecting, according to a scientist leading the project. The program bought laptops, which could be accessed remotely, for provincial Afghan government officials, including the governor of Jalalabad. “Was the More Eyes program successful?” the scientist asked rhetorically. “Well, let’s see, I just put a foreign electronic sensor into the governor’s bedroom.”

  In the end, however, the program fell short of its hopes to demonstrate crowdsourcing in Afghanistan. According to the white paper by DARPA’s Ryan Paterson, a series of experiments showed that More Eyes overestimated the ability of Afghans to access the Internet and the reach of mobile phone services in Afghanistan. “The More Eyes Team quickly learned that only 4 percent of
the population had access and skills necessary to access and exploit the Internet,” he wrote. “Rural populations had even less.” The DARPA contract, which ran out toward the end of 2011, was not renewed.

  The Synergy Strike Force and its tiki bar “oasis” also soon came to an end. Violence in Jalalabad grew steadily worse over the course of 2010 and 2011. Afghans who had worked and socialized with the foreigners at the Taj received death threats, and the insurgency that Western visitors to the bar were trying to forestall enveloped the establishment. On August 11, 2012, two men on motorcycles intercepted a car driven by Mehrab Saraj, the manager of the Taj and a friend to many who had worked on More Eyes. Saraj, who had survived the Soviet invasion, Taliban rule, and then the American invasion, was shot in the chest and killed.

  —

  Almost fifty years after William Godel’s ill-fated trip to Vietnam to collect data for the strategic hamlet program, hacktivists and humanitarians tried and failed to map out Afghanistan. Godel gave cash and gifts for data; the Synergy Strike Force offered free beer and mobile phones. Ultimately, however, what is so striking about DARPA’s efforts in Afghanistan was how marginal they were compared with the Vietnam War. DARPA’s deployment to Afghanistan echoed elements of Project AGILE but never approached its scope. What was forgotten, moreover, was that Godel’s version of counterinsurgency was designed to help local military forces in order to avoid sending American troops. In Afghanistan, counterinsurgency—and DARPA’s contribution—were viewed as tactics for aiding American troops already entangled in a foreign war.

  For several years, Petraeus’s reintroduction of counterinsurgency was hailed as a success, at least in Iraq. The accolades were ultimately short-lived. Similar to Vietnam, the local government’s failure to govern effectively in Iraq ended up fueling the insurgency. In Afghanistan, where the central government was even weaker, counterinsurgency by 2013 was being widely derided as a failed strategy. In the end, Petraeus’s approach suffered from the same fatal flaw as counterinsurgency in the latter days of Vietnam: the local government, not foreign forces, must ultimately provide security. Iraq and Afghanistan were counterinsurgency turned on its head.

  DARPA never publicly discussed More Eyes, and although the Pentagon later touted Nexus 7 as a success, there is no evidence that it had any useful impact on operations. “There are no models and there are no algorithms,” one anonymous official told Wired, griping about the program’s deployment to Afghanistan. A more sanguine assessment was published in The Wall Street Journal, which quoted an unnamed former official claiming that Nexus 7’s predictions about attacks in Afghanistan were accurate between 60 and 70 percent of the time. “It’s the ultimate correlation tool,” the official told the newspaper. “It is literally being able to predict the future.” Neither statement added substance to the debate. One thing was clear, however: like the McNamara Line, Nexus 7 did not change the course of the war.

  In March 2012, Dugan left to take a job with Google’s Motorola unit, where she began working to create a DARPA-like entity within the company. She had been director for less than three years—a not atypical tenure, although brief compared with that of her predecessor, Tony Tether. A Pentagon inspector general investigation into her continuing financial ties to RedXDefense, the family-owned explosive detection company receiving DARPA contracts, put her departure under a cloud. The inspector general’s report, released two years later, determined that Dugan had violated ethics by promoting the company’s proprietary work, though it found no evidence she had tried to direct funding to the company.

  By 2013, DARPA’s work in Afghanistan had drawn to a close, though the Taj, the onetime staging base for More Eyes, lived on as a symbol of the agency’s well-intentioned but ultimately failed efforts to harness science in the service of counterinsurgency. The Taj that year was nominally still open, serving stale cornflakes to the rare guest, who could lounge on a rusting lawn chair perched on cracked concrete overlooking the long-empty pool. At the edge of the pool, solar panels, which powered nothing at all, tilted up wistfully toward the sun. All that was left of the only tiki bar in Afghanistan was a collection of moldering business cards and a “Synergy Strike Force” combat patch stapled to a piece of wood, the last remnants of the Western patrons who had swapped their data for beer. Outside this onetime oasis of DARPA-funded techno-optimism, Afghans lived and fought much as they had for more than a thousand years. Inside, the bar stood empty, an enduring testament to science’s ability to transform warfare but not to end it.

  Epilogue

  GLORIOUS FAILURE, INGLORIOUS SUCCESS

  The intelligence aspects of military operations have grown and grown and grown until it seems the entire operation must collapse of its own mass and sheer awkwardness. In this arena science and technology have seemingly run amok and it is not at all clear that the volume of information has not, by itself inundated the Cray computers, confused the analysts, sacrificed credibility with the consumers, and virtually destroyed the capacity of the system to understand itself.

  —WILLIAM H. GODEL

  When I arrived at DARPA’s new headquarters building in February 2013 to meet with Arati Prabhakar, the agency’s director, I was struck by its location in the backwaters of Pentagon real estate. The building on North Randolph Street in Arlington, Virginia, is DARPA’s fourth headquarters; each move has taken the agency farther away from the Pentagon’s leadership physically and psychologically.

  The agency’s first location was inside the Pentagon’s elite E Ring, steps away from the Office of the Secretary of Defense. The second was in the Architects Building in Rosslyn, Virginia, not far from the Pentagon. The third was in Virginia Square, several stops farther out on the Washington Metro. DARPA in 2012 moved into a custom-designed building near the Ballston station, a project several years in the making. The new building features enhanced security and a customized conference center on the ground level, and each DARPA office now has its own floor. The top floor houses the executive suites for DARPA’s director and staff, who work behind frosted-glass doors. As with all grand projects, construction was plagued by rumors of snafus: bathroom doors that would not close, and windows that at first were not properly tinted to safeguard classified materials. Later, once they moved in, some employees complained that the building’s layout led people to work behind closed doors, interacting less with each other, let alone outside researchers.

  For an agency that claims paternity to some of the past few decades’ most exotic technologies, its offices have in the past been unassuming, even ramshackle. For several decades, visiting DARPA—at least the “white world,” or unclassified parts—was as simple as walking into the building and heading up to the office. That came to a halt in the 1980s, when a man walked in off the street and into one of the offices, dropped his pants, and mooned an unsuspecting DARPA official.

  Those days are long gone, not just for DARPA, but for all government offices. In the era following 9/11, visiting such offices involves metal detectors, X-rays, and, in many cases, confiscation of all electronic devices. My visit to DARPA’s headquarters began with a ritual familiar to those who do business inside the Washington Beltway with its layers of security, badges, and restrictions: I emptied my bag of electronics.

  “We’ll need to disable your camera,” the guard said, examining my iPad. I watched, worried that the security professionals from the agency that created the foundations of the Internet and boasts a number of classified cyber programs would perform some procedure on my iPad that would be impossible to reverse. Instead, the guard took a piece of masking tape, placed it over the camera lens, and handed the iPad back to me.

  “There’s another camera in back,” I volunteered.

  “We’ll disable that, too,” he said, placing another piece of tape over the back camera.

  In DARPA’s top-floor director’s suite, Arati Prabhakar invited me to sit down at a conference table. She had been director less than a year but was well liked and respected, particular
ly after what some felt was a tumultuous few years under Regina Dugan. If there was a criticism of Prabhakar, it was only that no one quite knew what she planned to do with the agency. So, I asked her what she thought was DARPA’s current mission. “I think it is unchanged,” she replied. “It has been and will be to prevent and create technological surprise.”

  She acknowledged, however, that DARPA’s mission has become more difficult in recent years. In the mid-1980s, during her first stint at DARPA, the military was focused on the Soviet Union, and the presumption was that the Cold War would continue indefinitely. “Facing that kind of existential threat, that one monolithic threat sort of drowned out all of the rest of the complexity of national security,” she said. “Of course DARPA, along with everyone else, was almost purely focused on thinking about that one threat. It is not that the world was not more complex, we just treated it as somewhat less complex.” Her perspective is understandable. When Prabhakar worked at DARPA in the 1980s, the agency was already following a path that was narrowing by the year.

  The irony of DARPA is that even as its mandate has shrunk, its reputation has ballooned. The agency that created the foundation of the Internet and stealth aircraft is hailed today as the “gem of the Pentagon,” touted as a model for government innovation, and praised by Democrats and Republicans alike. Prabhakar also knew firsthand the dangers of trying to redirect DARPA. In the 1980s, she had unwittingly presided over DARPA’s failed attempt to act like a venture capital firm, taking an equity stake in a gallium arsenide start-up. When asked about the episode, which had led to the only firing of a director in DARPA’s history, Prabhakar quickly changed the subject.