Read Kicking the Can for Free Online Page A

of his suit coat and draped the coat over a chair. Perspiring, he wiped his face.
    Schilling began the briefing. “The CIA has had several high-profile intelligence failures—not predicting the fall of the Soviet Union, the invasion of Kuwait by Iraq, and al Qaida and Bin Laden. We designed and programmed the OSi software to learn from these mistakes. We call it computational intelligence, CI. We used CI to search data gathered retrospectively, following each of these intelligence failures. This application helped explain the gaps in our intelligence that led to the failures.”
    He claimed the agency was able to improve the predictive accuracy of OSi by making changes to the logic and reasoning sequences. ‘Meta-search,’ its proprietary application, leveraged commercial search engines all over the world. The CIA could customize search criteria by content
—
business, news, people, information type—forum, blog, multimedia, e-mail, and geography
—
Asia, Europe, Africa, and the Americas. The data wereaggregated in a virtual database. It was the comprehensiveness of OSi’s tools that made the application unique. The CIA could search using precise algorithms to parse and aggregate, slicing and dicing the bytes to produce meaningful information.
    “If you’re still on the fence, we can demonstrate OSi’s capability,” Schilling said. Finished, he sat down on a hardwood chair and wiped his forehead.
    “Schilling, please wait in the anteroom. We’ll apprise you of our decision.”
    Schilling gathered up his coat and briefcase and departed.
    “Duncan, make your case.”
    “Mr. President, I recommend we use OSi. It’s no different than the super committee soliciting industry representatives. If we identify the credentials sought—education, experience, training, unique skillsets prospective team members should possess—Schilling’s team can program OSi to create the search algorithms. I’m convinced it’s worthwhile. I would like your approval to work with him. We can reconvene to evaluate the results of the demonstration.”

18
    D r. Janet Duncan illuminated the projection screen and began her briefing. She described the credentials input into OSi. No limitations were placed on information type. Schilling’s team ran two searches. The first was restricted to the United States, the second worldwide. Duncan advanced to the next slide.
    “The software’s computational intelligence feature identified candidates’ motivation as an additional criterion to enhance the overall predictive accuracy. In other words, the software suggested we isolate why a candidate would want to participate in the intellectual contest. When we stripped out ‘recognition’ and ‘financial gain,’ what came back was intriguing,” Duncan said.
    Duncan advanced to the next slide. A picture and biographical information were displayed on the screen.
    “Mr. President, we found someone. Christopher Scott Drummond, age forty-two, married, works as a CPA for Anderson Consulting, a global firm with a significant presence in health care. He’s a graduate student at the University of Washington who wrote a ninety-page thesis espousing a complete overhaul of health care.”
    President Cannon studied Drummond’s picture. He looked younger than his actual age—probablyworked out—but his facial expression conveyed a look of seriousness that reminded Cannon of himself.
    “There were other candidates selected, but we think Drummond is our guy. He’s a no-name, so he should be able to fly under the radar. He doesn’t have any affiliations with industry, so he should be objective. Drummond is our top pick.
    “Some of the strongest candidates include foreigners. It’s impossible to handicap how cultural and ethnic diversity could affect team performance, but the software ranks the multinational team higher than the US team. Among the foreign candidates selected is an industrial psychologist—a bonus for resolving any transcultural