Can the Future Be Simulated?

By Edward Tenner

More than 20 years ago, when visiting Cornell University on business, I stayed in the hotel-laboratory of the School of Hotel Administration. My host informed me as I checked in that rooms had many advanced features because manufacturers of everything from telephones to plumbing fixtures wanted to test their innovations before putting them on the market.

Now the London Telegraph reports that an entire $1 billion town without residents is under construction in New Mexico for testing innovations not with living people but with computer-controlled simulations:

The town, which will be modelled on a town of 35,000 people, will have roads, houses and commercial buildings, but will have no residents.

It will be built in New Mexico about 15 miles west of the nearest town, Hobbs, which has a population of about 40,000.

Scientists hope to use the new 'town' to research innovations in renewable energy as well as intelligent traffic systems and next generation wireless networks.

Of course there's a lot that can be learned from a "ghost town" (as the article put it) test bed. But can modeling the future work without the human dimension? I doubt it. In a New York Times comment on the project last year, the urbanist Greg Lindsay recalled the disastrous results of Mayor John V. Lindsay's application of a RAND Corporation report in 1968, based on computer simulations, that aimed to reduce response time but proved disastrous in New York's poorest areas.

Simulation software for "evacuation dynamics" of public places has existed for over a decade, yet the difficulty in communication with passengers on the Costa Concordia shows how difficult such models can be to implement in real life.

Today, driverless cars might log hundreds of thousands of miles safely in sparsely populated Western states like Nevada, and now perhaps New Mexico. But the question isn't so much whether individual vehicles can operate safely as whether hundreds or thousands of them, in a mix with conventionally driven cars of all ages and drivers of all skill levels and temperaments, can coexist. 

Simulations may also miss both positive and unintended consequences. Consider the Xerox 914, which ultimately played a big part not only in transforming home as well as office life, but in creating the computer era over 50 years ago. Only the behavior of large numbers of people in the real world disproved the estimates of expert consultants and revealed an almost limitless mass demand for photoduplication. And could the early developers of the cell phone have predicted that texting while driving could become not just a lethal highway menace but a multigenerational habit?

What a sad retreat from the day when at least one great American visionary did plan to build a high-tech experimental community for real people: the original EPCOT. Unlike today's billionaires, whether you love or hate him, Walt Disney was willing to take the risk for building a full-blown urban utopia from the ground up. He died before the project as he conceived it -- a living community of 20,000 -- could be realized. Disney understood what has been too often forgotten, that it's impossible to separate technological innovation from human interaction.

This article available online at:

http://www.theatlantic.com/technology/archive/2012/05/can-the-future-be-simulated/257121/