The following month, Congress voted to end DARPA’s Total Information Awareness program and shut down the Information Awareness Office. Total Information Awareness was, at least officially, dead. Except it was not: it was just about to get transformed into something much different and arguably much worse.
—
In recounting his final days at DARPA, Poindexter stopped to puff on his trademark pipe, just as he had in the Iran-contra hearings three decades prior. “Congress claimed they had closed my office and shut down the TIA program, but in reality what happened was they moved all of the components of TIA out of DARPA and into the intelligence community, and that winds up in the classified part of the DOD budget,” Poindexter said.
Was it possible, then, that Total Information Awareness was really just a cover for a massive black program, as Poindexter had first proposed to Tether in 2002? Poindexter said that in the end they had decided not to pursue the black program, because they wanted to include university researchers who did not have security clearances and it “would take too long to get all of the approvals required for such a classified program.” However, moving the programs over to the intelligence community achieved much of the same effect: Total Information Awareness never died; it just went black, just as Poindexter had proposed at the start. The Advanced Research and Development Activity, part of the NSA, took over almost all the programs that had been in the Information Awareness Office, except the privacy protection research. “The critics we had got the worst of all worlds,” Poindexter pointed out.
All too often, John Poindexter was, and often still is, portrayed as the paragon of the guileful government operative—the “ring-knocking master of deceit,” as The New York Times’s Safire called him. The real Poindexter—unapologetic, but gracious, even with skeptics—was never quite the bogeyman he was made out to be by the critics. His vision, however reprehensible to privacy advocates, was born of a deep, if misguided, concern about national security.
Had Total Information Awareness proceeded at DARPA, the plan was to expand Vanilla World, the simulated world where officials practiced spotting terrorists. Poindexter envisioned Cherry Vanilla World and then French Vanilla World, each with added layers of complexity and realism. But DARPA’s research never got beyond a simplified make-believe world. The fundamental scientific question—whether computer algorithms can be used to spot patterns of terrorist activity from a combination of private and public data—has never been proven, at least publicly.
Total Information Awareness, Poindexter insisted ten years later, was a success that fundamentally changed the focus and direction of how the intelligence community dealt with data. “Although TIA was cut short in 2003 by the Congress, I think basically we accomplished really what I had set out to accomplish, which was to propagate ideas of technology, develop some early versions that could later be improved, and look at a new process for intelligence analysis,” Poindexter said.
Poindexter’s belief in the need to balance privacy with security was sincere but also handicapped by a fundamental misconception of what privacy means to a significant portion of the American public. The concerns about government data collection, computers, and privacy dated back as far as the 1970s, and those concerns had only grown as the scale of data available had increased. “Some government officials attempt to finesse the privacy issue by insisting that individual records and data will not normally be shared or subject to examination by a human observer so, they argue, there is no real infringement on privacy,” wrote Steven Aftergood, a privacy advocate at the Federation of American Scientists, shortly after the debacle. “But that doesn’t get to the heart of the issue. Personal privacy is compromised whenever one is subject to unwanted surveillance, even by a machine.”
In the final accounting, the legacy of Total Information Awareness ran much deeper than any single DARPA program. The controversy not only ended the related privacy research; it also pushed data mining deep into the classified world of the intelligence community, laying the intellectual groundwork for the massive analysis and collection system that would be revealed ten years later by an NSA contractor named Edward Snowden.
For DARPA, the repercussions were also long lasting. Michael Goldblatt, the former McDonald’s scientist turned DARPA manager, found that his super-soldier research, once the darling of Richard Cheney, was also under fire by congressional staffers eager to find more dirt on DARPA. His work to develop vaccines for pain, and soldiers who could survive blood loss, was interpreted as a sinister plot hatched by mad scientists. “They thought we were making people robotic. They totally misunderstood,” he recalled. “I took a lot of heat on that.” But as with Total Information Awareness, the work did not actually end. “The beauty of it is we changed the name, and the program went on,” Goldblatt said, laughing.
Goldblatt might have found the episode funny with hindsight, but at the time he was deeply disappointed. His efforts to build the soldier of the future—embraced by the vice president two years earlier—were now a liability for DARPA. He resigned in 2003, the same year that Poindexter left over Total Information Awareness. Goldblatt even told Tether that the controversy over robo-soldiers could be blamed on him. “We got rid of that crazy Goldblatt,” he instructed Tether to tell Congress. “He was out of control, a cowboy, looking for ways to kill us all, pump us full of drugs.”
The DARPA director soon faced his own crisis. Shortly after the controversy, Newt Gingrich, the former Speaker of the House, paid him a visit with a direct message from Rumsfeld, who rarely communicated directly with DARPA’s director: “I came over to tell you that the Chief says you’re getting close [to getting fired].” Tether kept his job, but DARPA lost what he called the agency’s “greatest strategic thrust.” It also meant that at least for the foreseeable future, DARPA would not be involved in high-profile research related to the war on terror. Futuristic aircraft were fine, but applying the agency’s expertise in computer science to counterterrorism research was not.
For those who believed that DARPA should focus only on futuristic developments, ten or twenty years out, losing a critical counterterrorism mission was no tragedy. But for an agency that pioneered protection for the presidential vehicle in a matter of months, built a revolutionary nuclear test detection system in less than a few years, and led a worldwide counterinsurgency program, it marked a dramatic curtailment of the agency’s mission. The DARPA that emerged from Total Information Awareness would still be revered for its technological accomplishments but largely left out of frontline work in national security. Fearful of criticism, but eager for attention, DARPA turned instead to fantasy.
CHAPTER 18
Fantasy World
“I believe strongly that the best DARPA program managers must have inside them the desire to be a science fiction writer,” wrote Tony Tether, when asked about his philosophy for running DARPA. “Writers such as H. G. Wells, who for example wrote in his 1914 novel The World Set Free about nuclear power and talked about the atomic bomb and gave it the name used today, would have been a great DARPA [program manager].”
Tether’s passion for science fiction had been why he promoted Tony duPont’s fantastical space plane back in the 1980s. And it was also the reason why he selected Michael Goldblatt’s super soldiers and mind-controlled weapons to be the opening act for Richard Cheney’s visit back in 2001. That sort of techno-thriller material was exactly what Tether liked. “Imagine 25 years from now, where old guys like me put on a pair of glasses or a helmet and open our eyes,” Tether said in a DARPATech speech, referring to the work in Goldblatt’s office. “Somewhere there will be a robot that will open its eyes, and we will be able to see what the robot sees. We will be able to remotely look down on a cave and think to ourselves, ‘Let’s go down there and kick some butt.’ ”
More than anything, Tether loved Disneyland. The man who would become the agency’s longest-serving director believed the home of Mickey Mouse and Walt Disney’s Enchanted Tiki Room represented every
thing he envisioned for DARPA. The amusement park was where DARPA had debuted Total Information Awareness in 2002, and it continued as DARPA’s meeting spot. Over his nearly eight years as agency director, Tether would hold all four DARPATech conventions at Disneyland, featuring speeches punctuated by Star Wars theme music and schwag that included DARPA-embellished playing cards, custom DARPA-logo golf balls, and T-shirts decorated with armed drones. “Welcome to our world!” Tony Tether said, beaming to the audience in the opening ceremony in March 2004. “A world where science fiction morphs into reality.”
—
The second DARPATech in March 2004 also marked the one-year anniversary of the U.S.-led invasion of Iraq. What was expected to be a year of rapid reconstruction had instead turned into a bloody conflict, pitting American forces against not the conventional army they were prepared to take on but nearly invisible insurgents. March ended up being a particularly bad month: militants released a video showing the beheading of Nicholas Berg, an American hostage; four employees of the American security firm Blackwater Worldwide were killed in Fallujah, their charred corpses dragged through the streets; and a bomb killed dozens at a hotel in central Baghdad. Fifty-two American soldiers died over the course of that month in Iraq, many while driving in vehicles hit by improvised explosive devices, the homemade bombs being used with increasing frequency.
At Disneyland, however, where program managers were handing out DARPA-embossed M&M’s, the mood was upbeat. Tether believed his mandate was to make the agency “like it used to be,” a place where its employees “were always getting the director in trouble.” That was exactly what had happened with John Poindexter and Total Information Awareness. Tether survived the congressional and public backlash, even though it meant ceding DARPA’s role in counterterrorism research to the intelligence community.
While his hopes of being involved in top-level national security issues were dashed, Tether’s other vision—equal parts fantasy and spy lore—flourished, creating an entirely new perception of the agency. He funded ideas like Blackswift, an unmanned hypersonic fighter plane that could travel six times the speed of sound—a successor to the failed National Aerospace Plane. He also embraced novelties like “polymer ice,” a synthetic substance that could be thrown from the back of Humvees to make an enemy slip off the road. Some of his ideas, however, were more scientifically dubious. He supported work, for example, on a controversial “hafnium bomb,” which used a radioactive material that would potentially be tens of thousands of times more powerful per gram than conventional explosives, if scientists could figure out a way to trigger it. None could.
Tether knew he needed something big to recover from the Orwellian data-mining controversy, and even in the midst of the scandal he had come up with a new science-fiction-sounding idea: a robotic race. The competition would pit robotic cars on a 150-mile course of rough desert that stretched from Barstow, California, to Primm, Nevada, just forty miles short of Las Vegas. The Grand Challenge, as the race was called, would give $1 million to the winning robot creator. As Tether prepared to travel to California to kick off the competition, he had in the back of his mind the realization that the robotic car race, if it went well, might save DARPA from the circling critics. If it did not go well, DARPA’s future was at stake.
—
When Tony Tether flew into Los Angeles on a Saturday to attend an “industry day” for potential participants in the Grand Challenge in February 2003, he was in a panic. The race was a gamble, and DARPA, at huge expense, had rented out the entire Petersen Automotive Museum on Wilshire Boulevard, which housed luxury and vintage cars. “Gee, probably all the people are going to be our own. If there are five other people there, I’m going to have to go out on the streets of L.A. and get the homeless people to come in to fill this place up to make it look like we have people there,” he recalled.
Tether belonged to a generation of engineers who grew up watching Gene Roddenberry’s Star Trek but ended up working on Ronald Reagan’s Star Wars. The future DARPA director graduated with a PhD in electrical engineering from Stanford University in 1969, the year of the Apollo 11 moon mission, and went to work for the growing Cold War military-industrial complex, which was absorbing engineers as quickly as universities could produce them. There was still no Silicon Valley to compete for top graduates; the money, jobs, and excitement were in defense and aerospace. Tether worked for a series of defense companies until he was recruited into DARPA in the 1980s during Ronald Reagan’s military buildup, running the Strategic Technology Office, which was heavily involved in missile defense.
A relentless worker who treated Saturday as just another day, Tether wanted to become director of DARPA in the 1990s, only to lose out to a political appointee. When he got another chance, in 2001, he jumped at the opportunity. With his oversize engineer’s glasses and pomaded hair, Tether seamlessly blended into the Washington Beltway’s techno-industrial complex. But he also had a true, almost childlike love for futuristic technology, which was reflected in his first stint at DARPA, when he pushed people and concepts like Tony duPont and the space plane. His favorite phrase when describing something exciting or surprising was “holy cow!”
Before he became the head of DARPA, or even an engineer, Tether worked full-time for a period as a Fuller Brush salesman, going door-to-door selling personal care products. “I always say that [was] the best education I ever had,” Tether later told an interviewer hired by DARPA. “When you’re a Fuller Brush man and you knock on the door, you only have a second or two to really assess who’s answering that door and how are you going to get in that house.” Selling Fuller Brushes was about reading your audience, or as Tether put it, “telling the right story.”
Tether desperately hoped the Grand Challenge was the right story, but he worried that even a robotic car race, with no obvious connection to weapons or the military, might provoke a backlash, or at least put people off from participating. Despite attending Stanford for graduate school, Tether viewed the West Coast as a strange land, filled with liberals who simply did not comprehend how the world had changed after 9/11. “People, especially people from the Western part of the United States, really, I think, believe that the attack on New York was like a movie, ‘Godzilla Invades,’ you know, ‘New York City,’ or something like that, and didn’t really appreciate that we really were at war,” he later reflected.
At least for the car race, however, his concerns were overblown. When Tether showed up at the car museum, there was a line twisting around the block. Some eight thousand people attended. “Holy cow!” Tether thought to himself. “This can really turn out to be quite something.”
DARPA’s Grand Challenge was the brainchild not of a DARPA scientist but of the agency’s onetime chief legal counsel, Richard Dunn, who had been inventing creative mechanisms to evade bureaucracy. Whether it was finding ways to hire employees on special contracts or circumventing normal government procedures to work with small companies, he had become something of a one-man “fix it” shop for DARPA’s red-tape cutting. The Grand Challenge, or at least the framework of it, was another Dunn innovation, modeled after the Orteig Prize, offered by Raymond Orteig, a New York hotelier who put up $25,000 for the first solo flight across the Atlantic by plane.
Shortly before Dunn retired from DARPA in 2000, he persuaded Congress to give DARPA authority for “incentive prizes,” although there was no specification of the type of contest. Robotics was one of the early suggestions, Dunn recalled, although the idea had been robots that could scale buildings. When Tether took over DARPA, the prize authority was sitting around unused, and he had a different idea: a robot car race. “Well, everybody in this country owns a car. Everybody can buy these computers,” he reasoned. “The sensors you can buy. The actuators are even available through the handicapped market. So, this is something that has a low hurdle for the average guy.”
Originally, Tether had wanted the race to be in Anaheim, part of the Los Angeles metropolitan area. He envisioned a 25
0-mile car race from Los Angeles all the way to Las Vegas along public highways. As with the crossing of the Atlantic by plane in the 1930s, the individual technologies needed to build self-driving cars were all available, in theory, but pulling them together seemed like a nearly impossible feat, or at least something that had simply never been done before. It would be a mental advance, as much as a technical one, and Tether hoped the contest would spark nationwide interest in robotics. In the end, however, an air force colonel hired to manage the logistics told him there was no way to shut down a highway so close to Los Angeles, even at night. Instead, the colonel suggested Barstow, a dying town in the California desert dominated by desert shrubs and methamphetamine labs. Shutting down a road in Barstow would not be too difficult.
On March 13, 2004, fifteen competitors pulled up to a starting line in Barstow: At stake was the $1 million jackpot for the vehicle that could achieve the best time along a treacherous, 150-mile course through the Mojave Desert. National press from around the country had descended on Barstow to document the historic event. The anticipation lasted less than eight miles. One by one, the competitors dropped out, stuck on a rock, caught in an embankment, stumped by software, or, in one case, dramatically barreling through a fence. The race favorite, a Humvee, made it the farthest, logging 7.32 miles before stopping, its “belly straddling the outer edge of a drop-off, front wheels spinning freely, on fire,” as the magazine Popular Science reported, headlining news of the race as “Debacle in the Desert.”
The Grand Challenge was off to an inauspicious start. No one claimed the $1 million prize. Then again, it had taken eight years from the time the Orteig Prize was offered until Charles Lindbergh won it. The saving grace was that unlike the Orteig Prize, which claimed the lives of several contenders, no one died in the desert race. Tether played down the failure, promising to hold another Grand Challenge and blaming the press for hyping the first race. “I know you guys built it up so much [and then] it only went seven miles, you were embarrassed about it,” he said. “But don’t think we were.”
The Imagineers of War Page 37