Machine Politics

Thanks to the convergence of crowdsourcing, social media and social networks, the world is becoming a more responsive place. While we’re still experiencing the adolescence of social media, the information systems we’re building are starting to get better at taking input from crowds of people and using it to help us mine the data glut for what we, as individuals, are most likely to want or to need.

Case in point: Google’s new Instinct program and Pandora. “I can now program my inbox with Instinct or take Pandora’s Thumbs feature and tell it I dislike a song and it can respond by giving me a better one,” MIT Media Lab’s Riley Crane said at last week’s PopTech2010 conference in Maine.

The great irony, of course, is that despite these systems’ abilities to be more responsive, none of them can yet solve big social problems that require large-scale collaboration or coordination over time.

But we’re getting closer, says Crane. Thanks, in part, to social networks like Twitter and Facebook, we now have—for the first time—the ability to have and maintain, simultaneously, weak social ties of former coworkers and classmates and other acquaintances. “For years, we’ve had ways of communicating with strong and weak ties, but what we’ve really seen with Twitter and Facebook is that we have new ways to keep up with our weak ties and this solves a really hard communication problem,” Crane says. This new class of weak ties, he says, will become increasingly important as we step up our efforts to organize ourselves around shared context. These weak ties, he argues, form a new social class. Can it be mobilized, locally or nationally or internationally, for social good? To achieve political goals? Some argue that in the political arena, the Tea Party movement is an example of weak-tie networks that have been brought together around common goals. [For more on the strength of weak ties, see both Malcolm Gladwell’s October 4, 2010 piece in The New Yorker, “Small Change: Why The Revolution Will Not Be Tweeted” and Jonah Lehrer’s rebuttal, “Weak Ties, Twitter and Revolution,” in Wired.  Also see Twitter Co-Founder Biz Stone’s rebuttal in The Atlantic.]

Not convinced? We’re already seeing what’s possible, though mostly through anecdotes. Patrick Meier, the director of crisis mapping for Ushahidi, told PopTech conferees Friday about an urgent request that had come into Ushahidi from Haiti shortly after the January earthquake. Rescue workers on the ground were scrambling for the GPS coordinates for seven locations in Port au Prince where survivors were known to be trapped beneath fallen buildings. More than 1,000 miles away, in snowy Cambridge, Mass., Meier and one of his grad students got busy and soon were able to find six of the seven coordinates, using Google maps and social networks. But the seventh location remained elusive—until the grad student tweeted for help. “She sent out her query to her followers and got an answer that the building she was looking for might be owned by the Holiday Inn chain,” Crane said Saturday, retelling Meier’s story. “She used that data to look for a building with a pool, and using satellite images, she found it. But she still needed to locate the building nearby.

“Then something incredible happened,” Crane said. Someone in Brooklyn tweeted that he knew someone who used to work in that hotel and knew the area. Soon, that person was walking Meier and his student through a satellite image of the neighboring streets to help rescuers get the proper coordinates. Said Crane: “Think back 5 years. You wouldn’t have been able to do anything like this. We’re getting closer to figuring out what it means to start forming spontaneous communication networks that act like central nervous systems.”

But activists, beware. Thanks to an ever-growing glut of data, we are at once racing to tune out irrelevance and stoke innovation. “If we’re not careful in our race for perfect algorithmic relevance,” says Moveon.org founder and online organizer Eli Pariser, “we may find ourselves at a loss of brilliant breakthroughs.” Pariser told PopTech conferees Saturday that unless we can integrate some form of “curation ethics” into the ways our machines slice and dice the info glut for us, we risk missing need-to-know information that our algorithms may now deem irrelevant. “...This time in the evolution of information systems is really a great time for organizing information but ... innovation requires serendipity. ... An algorithmically curated environment lacks the blind variation needed for creativity. We need media systems that call our attention to what we don’t know. If you have perfectly constructed environments, you’re going to miss out.”

Pariser urged conferees to be vigilant against over-censorship and hyper-stereotyping by data-vetting machines. “What’s creative is something new and useful that you can’t come up with through an algorithm or equation,” Pariser said. “We also need to be open to people and systems that can disprove data—not just to systems that find similarities.”

Read more stories by Marcia Stepanek.