Dear Mr. President

THE DISTRIBUTED BRAIN

As the next president, you should strive for something more substantial than online fireside chats, open-ended forums for public comments, and town-hall meetings in streaming video. Instead of devoting resources to these superficial, large-scale interactions, think small. New York Law School's Noveck has worked to promote a radically different vision of how the opinions and expertise of regular Americans might be tapped to improve government decision-making. Instead of asking people to sound off on whatever bee happens to be in their bonnet, she wants to present the public with a series of specific questions, for which the government needs specific answers.

Her ideas are being tested at a social-networking Web site associated with the U.S. Patent Office called Peer-to-Patent. Here's how it works: Government employees now spend much of their time checking that the ideas contained in patent applications are sufficiently novel and interesting. Peer-to-Patent allows them to recruit unpaid specialists from around the world by posting the applications online. Users migrate toward the technologies they're most interested in; for example, a patent for a novel way to network turbines on a wind farm might attract computer scientists and environmental engineers. They can also rate one another's work or invite colleagues to participate. Then the group hashes out their thoughts over the Web -- not unlike creating a Wikipedia entry -- and passes the best ideas back to the Patent Office. That saves work for the clerks and improves the quality of their research. It might also cut the costs of patent litigation down the road.

The software behind Peer-to-Patent isn't especially complicated or new, nor does the site strive for mass appeal. It simply tries to draw in the particular people who might be able to answer a particular question. Now imagine that instead of building some bloated, mass-interaction site like E-Petitions, you set up a multitude of these smaller, special-interest sites. Each government agency could thus reach out to only those users with the expertise relevant to a particular policy.

Last year, for instance, the Food and Drug Administration proposed a set of rules that would make sunscreen labels more comprehensive and accurate. But the agency has been swamped by several thousand comments in response, and is bound by law to review each one. In June, the FDA conceded that the final rules would be delayed indefinitely.

Now imagine if there were a system for the interested parties to manage themselves. Commenters could debate the fine points of the proposal online and recommend specific changes on a wiki version. Or users might help one another by rating comments up or down, making it easier to find the ones most deserving of a response.