Programming and politics are often a poor mix, as the Defense Department's research agency is learning the hard way.
Late last month, the Defense Advanced Research Projects Agency killed a homeland security-related program after a wave of bad publicity. The program, known as Futures Markets Applied to Prediction, was intended to harness the distributed intelligence of financial speculators to gauge the likelihood of terrorist attacks.
DARPA, of course, is also the force behind the controversial Terrorism Information Awareness program (formerly the Total Information Awareness program), which would sift through various public databases in hopes of identifying individuals who threaten national security.
In some ways, the two projects are typical of DARPA's work. The agency often investigates radical ideas for the Pentagon, and they are always on the lookout for that quantum leap in defense operations. Imagine filling a room with the smartest engineers in the nation and asking them to invent an automobile that runs on solar power without any drop in performance. That's DARPA.
But what if that team of engineers actually came up with such a model, the only catch being that any accident would be certain to result in fatalities? Such a proposal, though technically feasible, would likely be deemed a failure.
Advanced research, by its very nature, produces more failures than successes. The only time this becomes a problem, though, is when those failures see the light of day.
DARPA's mission sometimes requires its engineers to consider the most unlikely solutions to a given problem. But somewhere along the line, someone in that organization must provide a reality check and filter out those ideas that are politically untenable before they enter the public realm.
Without an adequate filter, the agency is simply inviting congressional scru—tiny. Perhaps some scrutiny is required. But it would be unfortunate for DARPA — and for DOD — if political pressures put artificial constraints on the agency's intellectual firepower.