Top 25 Technology Drivers

Paul Resnick, doctor of computer science, is the founder of AT&T Labs - Research's year-old Public Policy Research Department. The department's name sounds innocuous enough, but Resnick isn't. His poetic goal: "more flexible customized tools, rather than blanket rules." That means using technology to mold policies while avoiding unwanted regulation. He sees his mission as both altruistic and pragmatic. Resnick wants to create technologies that will make it "safe, fun and profitable to interact with people we don't know very well." He also wants to help AT&T get a head start conceptualizing the products that will make that happen.

Resnick is mulling technology standards that would let systems act as electronic couriers that explain the degree of privacy sought in an electronic conversation and whether those conditions will be honored at a Web s ite. He's also interested in conveying information about copyright restrictions in a similar fashion. Both approaches, would rely on his PICS work--the effort he considers his most important over the past 12 months. PICS started as a way to establish ratings servers to single-out content that might be inappropriate for children. PICS has grown, thanks to those who support Resnick's belief that technology is flexible enough to convey a variety of metadata about Internet content (see Network Computing Online, June 15, "Effort To Censor Content May Help Business"). For example, with PICS, users transparently can choose only to visit those sites that attest that user information won't be sold; or the site might offer a $5 coupon in exchange for a visit and salable user information.

AT&T also is exploring a concept known as "load-dependent pricing," and it wouldn't be surprising to see Resnick eventually enter that policy debate. AT&T researchers, for example, have shown an interest in a next-generation Internet routing architecture, known as Nimrod, as a way to handle competing bandwidth requirements. The load-dependent pricing concept involves creating a free market in bandwidth, so that users of inelastic applications, such as voice, would agree to pay a higher price for bandwidth, guaranteeing a route around network congestion.

One of Resnick's carry-over interests from MIT is the field of collaborative information filtering--especially for designing price models. One such application, for example, could provide a NetNews reader with articles based on ratings by oth er users that are correlated with the reader's own assessments. He says he's found that the time users spend reading an article correlates well to the user rating--so much so that the entire process could be automated without requiring any wor k from the user.

For Resnick, a lifelong lover of strategy games such as chess and bridge, the molding of technology and policy is the ultimate challenge. "I always wanted to do something that was good for the world with computers," he says. "Of course, as you get older, the simple answers of what is good and bad for the world don't seem so simple anymore ý and it's harder to keep acting on that goal. What has made me feel the best about this year and the prospects for this new department is even though good and bad isn't black and white, even though things are gray, I can still move forward and take actions I think are good for all of us. I'm helping to create an environment where people can interact with others they don't know well, and that's fulfilling."

Contribution Last 12 Months: Co-chair of the Platform for Internet Content Selection (PICS) effort, driver for IP address transferability. n Next 12 Months: Cement AT&T's techno-policy initiatives. n Millennium Forecast: Internet users will lea rn to share the Net: ads will work without people spamming them; all people will be able to use the Net without being offended. n Millennium Disturbance: The Net can't be shared and will fracture into many separate networks. Car: From bicycles and public transportation, to a clunker, to a 1996 Acura Integra--all in one year.