Spy vs Virtual Spy


by Tim Moran

I've never taken to video gaming, but I do like spy stories, so reports of the convergence of intelligence gathering and video games piqued my interest -- maybe because the words "intelligence" and "video games" seem so at odds.

It seems that American intelligence analysts tend to be extremely biased, which leads to less than intelligent intelligence decisions. The spy agencies, which have had their share of intel flops, admit to the bias problem. According to the Intelligence Advanced Research Projects Activity (IARPA), which operates within the Office of the Director of National Intelligence: "Cognitive biases in analysis tend to increase with the level of uncertainty, lead to systematic errors, filter perceptions, shape assumptions and constrain alternatives."

IARPA thinks the game's the thing for reducing biases, and, to that end, it has created the Sirius Program, the goal of which is to create a "Serious Game" that will train intel spooks, then measure how well they do at recognizing and acting on the "cognitive biases that commonly affect all types of intelligence analysis."

These different prejudices include: the "anchoring bias" (relying too much on a single piece of evidence); the "confirmation bias" (only accepting facts that back up a pre-made case); and the "fundamental attribution error" (attributing too much in an incident to personality, instead of circumstance). IARPA is looking to enlist research assistance from far and wide: social scientists, computer scientists, statisticians, gaming and virtual-world experts, universities, and companies from around the world.

And if that doesn't work, IARPA is working another angle: computerization. The group started working in 2009 on a system that could replicate, and eventually outdo, human intel decision-making. We hear that the computer excelled at the spy stuff, but it also hit on every cute PC in the department and kept asking for Felix Leiter to fetch it a vodka martini. . .shaken, not stirred.