A Trillion Points of Data

Whether you're pinging a person's location to provide traffic information or using it to chart consumer behavior, privacy quickly becomes an enormous concern.

For this reason, Pentland (a cofounder of Sense Networks) has called for a "New Deal on data." That means, in part, treating data a lot more like personal property—it is owned by the person who originates it, but companies can borrow it. "You should think of data roughly the way you think of money," he says. "It's something you own, it's something you can loan to people, but you want to get something back" in exchange for it, such as a more useful search engine or real-time traffic information.

Just as money can be taxed, there's a case to be made that the government should be able to appropriate data for the public good.

A skilled reality miner could use cell-phone data to pinpoint buildings or blocks where a higher than average proportion of people are at home on a workday. That, in turn, might imply an outbreak of influenza or another contagious disease. The benefits of using this method are enormous—doctors and epidemiologists could track the spread of an illness nearly in real time, and perhaps even issue warnings in communities where infected people have traveled.

That God's-eye view would come in handy in plenty of other scenarios, too. Public-transportation officials could use it to schedule roadwork and determine where infrastructure repairs are most urgent. Similarly, poverty experts could use it to determine whether public policies are working in a given neighborhood. "You can get things out of this data that you wouldn't expect," says Pentland. "Are people working? Are the kids sick? You can tell that from the patterns of communication and mobility that people have."

This might be even more influential in the developing world, where data on all kinds of variables, from population to health, is scarce to nonexistent (but cell phones are plentiful). Developing-world governments could employ reality miners to assess the growth of informal developments like slums, or find out which neighborhoods are less likely to make use of an AIDS clinic. It could, in short, replace the raft of contradictory, costly and sometimes inaccurate polls, censuses and surveys that much of public policy relies on. "That's a really revolutionary idea," says Pentland, "because today we have no clue whether government works or not." A global nervous system would provide more than a clue, and might reshape the business firmament at the same time.

© 2009