In 1991, computer scientist David Gelernter of Yale University predicted in his book Mirror Worlds
that advances in computing power and connectivity would lead to the creation of virtual cities: micro versions of the real world built out of data streams and algorithms instead of bricks and concrete. The book became an instant classic among the digital cognoscenti and provoked the wrath of Ted Kaczynski, the notorious Unabomber, who nearly killed Gelernter with an exploding postal package in 1993. Fast-forward a decade, and evidence of Gelernter's prescience abounds. Millions of people are active participants in virtual worlds that possess the economic and creative vitality of actual communities. The Net denizens who have built a homestead in massively multiplayer games like The Sims Online
) are the digital world's equivalent of the postwar immigration to California. The worlds are so vivid that the players now take the virtual objects that they've accumulated in these games—swords, houses, entire characters—and sell them in online auctions for real-world currencies. Edward Castronova, an economics professor at California State University at Fullerton, recently calculated that the per capita gross national product of the land of Norrath—part of the Everquest universe (everquest. station. sony.com
)—was $2,266, which is on a par with the per capita GNP of Russia.
Although virtual worlds are now attracting growing audiences that rival those of some television shows, the essential component of Gelernter's vision remains unrealized. Mirror worlds were not supposed to be an escape from reality; they were supposed to reflect
reality. A modern-day city generates plumes of data the way 19th-century industrial towns generated smoke. There are block-by-block crime statistics, test scores for every student in every school, traffic reports updated by the second, demographic profiles by zip code, and so on. All of those numbers exist somewhere in cyberspace, but finding them is next to impossible. Gelernter envisioned a centralized repository for all this data, a virtual reconstruction of a space that would showcase everything going on in reality. Gelernter's simulated worlds were going to be mirrors. By comparison, the simulations we have now are fantasy islands.
In a true mirror world, data would be mapped onto recognizable shapes from real life. For instance, to find information on a local hospital, you would locate the building on a computerized map and click on it with an "inspector" tool. Within seconds, the big-picture data about the facility would come into focus: number of patients and doctors, annual budget, how many patients died in operating rooms last year, and more. If you were looking for more specific information—say you were considering giving birth at the hospital—you could zoom in to the obstetrics department, where you would see data on such subjects as successful births, premature babies, and stillborns. Information about how the hospital connects to the wider city—what Gelernter calls topsight—could be had by zooming out.
Another key feature of Gelernter's vision is what he calls narrative information systems. The data in a mirror world are time-based: The mortality rate at a hospital varies from month to month and from year to year, and a mirror world would record those changes. So with any variable—or combination of variables—you could reverse the data stream to see past conditions. This is a tool not only for making sense of the past but also for predicting the future: If you're in the middle of an economic downturn and you're thinking of moving to a new neighborhood, you might like to see how the real estate values fared during previous recessions. With a mirror world, you would select a neighborhood (or a city block, if you wanted that much detail) with the inspector tool and shuttle the data stream to 1990 or the mid-1970s or the late 1920s, as though you were rewinding a VHS tape.
"My life, like your life, is a series of events in time, with a past, present, and future," Gelernter says, sitting in a conference room in the New Haven offices of Mirror Worlds Inc., the software company he cofounded. "And that's the way my software ought to look. The mirror worlds approach to organizing information is based on reality, as opposed to an engineer's or a computer scientist's fantasy. I don't want my personal life to be stored in an arbitrary UNIX file tree; I want it to be life-shaped—the shape of the way I live it."
He gestures out the window, to the stately spires and ivy-covered buildings of the Yale campus. "I want information on New Haven to be New Haven-shaped, not in 10,000 separate databases."
The good news is that most of the technical infrastructure for creating an authentic mirror world is in place. Consider the latest release of the ubiquitous video game SimCity.
The software interface might have been lifted directly from Gelernter's book: Players create sprawling simulated cities, zoom in on any structure, use an inspector tool, and determine up-to-the-minute attributes. Click on an apartment building and up pops the number of residents, average wealth, land value of the property, crime and pollution rates on the block, and a half dozen other variables. If the game imported those numbers from a real-world database, you'd have a mirror world. And some real data are even easy to find: In 2002, New York City mayor Michael Bloomberg unveiled an impressive program called My Neighborhood Statistics (www.nyc.gov
) that publishes 14 major variables, from homicides to asbestos complaints. Just type in your street address.
Furthermore, game designers are showing an interest in connecting virtual worlds to the real one. A number of recent games allow the importation of weather into the game space. If it's raining outside, it'll be raining on the screen as well. A new immersive world called There (www.there.com
) that's still in beta promises to include virtual shopping malls in its fictional universe so that you can try on that angora sweater in cyberspace—and if you like it, a real one will arrive via FedEx the next day.
A fascinating new project called Game Neverending
) may further blur the line between real spaces and game spaces. Created by a talented team of programmers based in Vancouver, British Columbia, Game Neverending
is a delightfully open-ended virtual space that encourages new sociopolitical structures: Where SimCity lets you build a neighborhood, Game Neverending
promises to let you create a new form of grassroots alliance, or even a cult. The designers are hoping to get away from the traditional game interface, which involves sitting focused exclusively on the computer screen. The idea is to create new ways for information to flow from real space into virtual space.
"I could have lieutenants in my cult who are allowed to draft new members," says cocreator Stewart Butterfield. "And one level up, there are cult members who can kick people out. Now, managing that group can take a lot of time, but the social interactions that are involved in that management can take place outside the game." You could exile an underling without actually launching the game just by sending a short text message from your cell phone. "We really want people to be able to have a 30-second or a two-minute participation that's satisfying, rather than launch a monolithic application and play with it for five hours."
Somewhere in this mix of tools and interactivity, a true mirror world is brewing. Combine the visual interfaces of SimCity,
the up-to-the-minute data of My Neighborhood Statistics, the multiple inroads of Game Neverending,
and you'd be able to create a true alternate universe, one that was mapped to real events. Ten years from now, a massive public planning operation like the one under way for the Ground Zero site might well be unimaginable without a mirror world. Photoshop pictures of the new skyline are nice but can't answer the important question: How will this new space actually be used
once it's built? Will it be dreary, teeming, commercial, or diverse? Just create a virtual model of each proposal, download the latest economic data, populate it with users willing to participate as residents and workers, and press play.
Before I leave New Haven, I ask Gelernter how confident he is that his vision of mirror worlds will be fulfilled. He takes a second to think and then says: "The first city or community that really does it, the whole world is going to take notice. Right now you have these convincing simulated microcosmic universes, and you have all this raw, fresh data pounding in over high-capacity lines." He looks up from the table and smiles. "The combination is too 2 + 2 for people not to eventually say 4."