Also in the news today: Howard Rheingold in The Feature on Location-aware devices, privacy, and UI design:
Location-aware devices and services are emerging at the intersection of empowerment and surveillance: the same technology that could let you know if a good Chinese restaurant or old friend is in the vicinity could also betray your location to a totalitarian government, neighborhood spammers, and your vindictive ex-spouse.
Aside from being an apt recitation of all the awful things that ubiquitous location-aware technologies could bring us, this quote reminds me of Scott's point about everyday privacy: when it comes to location-aware devices, the question is not, "What do people want to know about you?" but rather, "Who's asking?"
And also, of course, Louise's study on privacy concerns around location-aware services.
The comments section on the article also brings up dodgeball's lowtech solution. Dodgeball's "location tracking" relies on users self-reporting their location through text messaging. This kind of location reporting works particularly well for d-ball, which is all about connecting friends in NYC through bars and restaurants. And because participation in dodgeball social groups is strictly voluntary, it neatly solves the "who's asking?" issue as well.
In order to "watch" the correct entity, surveillance mechanisms need specific information about the target's social identity or physical location or both. Think about one of Raymond Chandler's gumshoes trailing an elusive blonde through Los Angeles. Either the detective has the car license plate, or the blonde's name, or a really good fix on the roadster's taillights as the it speeds away. And he probably has two out of three. So goes location-based applications.
The initial piece of information is the gateway: either an observer uses location data to get some social information (as in, say, the movie Fulltime Killer, where a contract assassin spies on his housekeeper while she's cleaning his living room), or an observer uses social identifiers to resolve the target's location and activities (as in, say, Minority Report). So on the one hand, the bluejacker gets more social information about an unknown entity whose location is already known. And on the other hand, Rheingold's examples presuppose a remote observer who accesses location/activity info for an already identified target.
This is probably obvious to everyone reading this already, but the two situations are great illustrations of a couple of related difficulties in implementing ubiquitous computing applications. In a Minority Report-type scenario, for example, Rheingold's hypothetical neighborhood spammers don't physically threaten the person whose cell phone they're bombarding*. And in the Fulltime Killer scenario, a hypothetical bluejacker sending creepy messages to nearby women will not also have access to the women's credit records because he doesn't know their legal names. The first scenario implies a loss of privacy -- that is, control over the knowledge other people have about you. The second suggests a mounting vulnerability -- that is, a sense of personal violability. Which scenario you find creepier depends, I suppose, on personal temperament.
*Although an implied threat of this scenario is that a powerful entity like a government or multinational corporation will be able to physically threaten individuals using information about their location and their activities. But for the moment, the most likely location-aware abuse will come from spammers.
Leave a comment