Kill Chain: The Rise of the High-Tech Assassins by Andrew Cockburn Page A

technology “stealth” and changed the security classification from a low-level “confidential” to the highest levels of “top secret.” Intimations that something new and incredibly sensitive was in the works helped to justify the massive funding while simultaneously making test data inaccessible to skeptics. Meanwhile, Perry pursued his grand vision of stealthy cruise missiles and large stealthy aircraft, even stealthy ships. The services were aghast at the impact the inevitably staggering cost would have on more cherished projects, but Perry calmed their fears by promising that the programs would be “technology driven, rather than funding driven,” meaning that there would be no limits on spending for any apparently promising advance in technology. Behind the cloak of secrecy the multibillion-dollar B-2 strategic bomber and the smaller F-117 proceeded slowly and expensively toward production, their performance and, more important, their budgets screened from the outside world.
    While these technologically ambitious programs were under development, one program founded on radically different principles was quietly entering service. The A-10 Thunderbolt II, to give it its official name, was commissioned to provide “close air support” to troops actually on the battlefield rather than to attack enemy forces far in the rear. Pilots soon renamed it “the Warthog.” True to its core belief in waging war entirely independently of the other services, the air force had no interest in such a weapon, but there came a moment in the early 1970s when it seemed possible that the army might capture the close air-support mission for itself with a costly new helicopter, a development that would have deleterious effects on the air force budget. Consequently, the service turned to an analyst then working in the Office of the Secretary of Defense known for his unconventional view on the importance of the close-support mission.
    Pierre Sprey, a mathematical prodigy whose parents had escaped to the United States from Nice in 1941, had worked summers at Grumman Aerospace from the age of fourteen, the year he was admitted to Yale. By the time he had left Yale at age eighteen and moved to Cornell for his masters, Grumman was paying him as a statistical consultant on programs that included NASA’s Lunar Excursion Vehicle and navy jets. In 1960 he was enlisted to work on an experiment with a direct bearing on “remote sensing” systems of the kind that would one day enable operators looking at a video screen in Nevada to target a truckload of tribesmen on the other side of the world. The experiment was designed to test whether it was worth putting TV cameras in reconnaissance planes to give pilots images better than those they could get with the naked eye.
    In a research facility on Long Island a Grumman team crafted a display filled with models of tanks, jeeps, command posts, and other tactical targets, all carefully camouflaged to appear as they would on an actual battlefield. Combat veterans monitored the effort to ensure realism. An observation platform overlooking the display gave observers the same view they would have from an airplane, with a shutter in front of the platform that could be opened to allow only as much time to view the scene, ten or twenty seconds, as a pilot would have as he flew past. In those few seconds the observers were asked to pick out as many targets as they could. Then, for the same amount of time, observers were shown a TV picture using a high-definition camera of a similar display and asked again to look for targets. The results were surprising. Only when the camera had zoomed in enough to show a narrow view, the equivalent of looking down through a soda straw from 15,000 feet up and seeing an area no larger than two tennis courts, were the TV viewers able to find as many targets as those spotted with the naked eye. But the TV viewers also identified five times as many false targets—tanks and other