Don’t Be Afraid Of Foursquare, But We Need Circles Of Trust
“Team Foursquare” published a we-get-it-donn’t-worry sort of post, responding to the Please Rob Me induced concerns about geolocational privacy:
Anyway, we definitely “get” the larger issue here - location is sensitive data and people should be careful about with whom and when they share it. And at foursquare, we do everything we can to make sure that our users know with what people and social sites they are sharing their location with.
First, let me say that Team Foursquare and most everyone else is looking at this through the wrong end of the microscope. This isn’t about privacy, that is to say keeping things secret; it is about publicy: making things public.
This distinction seems subtle, but as I expand in a bit, it is a significant shift in frame of reference.
Before that, though, here’s MG Siegel’s take on the Foursquare post:
- MG Siegler, Foursquare Responds To Please Rob Me: Please Shut Up
The point here is: does anyone really think Foursquare alone is going to lead to a rise in break-ins? No (I’m not saying that will never happen, but I don’t see it being an epidemic). The main point, as I saw it, is that tying a closed, symmetric network (Foursquare) to an open, asymmetric one (Twitter) is something that’s potentially troublesome for location-based services. But no one really seems to be talking about it.
What I mean by that is that on Foursquare, just like Facebook, you have to explicitly allow someone to follow you (and you follow them back in return). On Twitter, anyone can follow you without your permission (assuming you have an unlocked account). When you tweet out your Foursquare check-ins (some people even do this automatically), it essentially makes Foursquare an asymmetric network. And believe it or not, some people are doing that without really thinking about it. Or they’re doing it because it’s easier to gain friends/followers on an asymmetric network.
This is an issue that Facebook is undoubtedly thinking about right now as it considers how to enter the location playing field. As I said, right now, the network is largely symmetric, but recent changes are making it more asymmetric. And how they handle location information — with their 400 million+ users is going to be very interesting. And potentially, actually scary.
Leaving the scary aspect to one side, MG is right that the publicy slippage when connecting a closed, relatively stable, symmetric network like Foursquare with an open, relatively fluid, asymmetric one like Twitter can lead to unintended consequences.
Foursquare provides a fixed notion of circles of trust: you have a group of friends to whom you are a friend. Being a friend in this context means you are willing to share geolocational information. This does not, however, obligate someone to actually share their location: they could lurk, only receiving updates. In this way there could be an asymmetry on the sharing of geolocation information, but it would still be going on in a closed, and relatively stable system, because you have to approve all friends and the rate of adopting friends is relatively slow. So asymmetric sharing patterns are always present in Foursquare, with the most voluable and active members sending out more updates than their friends. But that imbalance happens in a circle of trust, so it is accepted.
Note that it could be one of your ‘Friends’ that breaks into your apartment, or who stalks you to your place and rapes you. Most rapes (77%) in the US are by non-strangers according to Bureau of Justice Statistics, and go unreported. While many of these rapists might not be close enough to get friended, my bet is that a lot of them are.
I am not saying this to create concerns about safety, per se. I am suggesting that a single level of ‘friending’ is probably too general to satisfy assumed needs for safety, although there is little evidence that social tools increase the likelihood of burglaries or rape. We don’t have an epidemic of ‘social crime’ to resolve here.
The slippage of geolocational information from a closed, stable network into an open, dynamic one opens up a wider assemblage of contacts, but without the assumed friendship that comes from symmetric following. And it is much more likely that Twitter contacts that opt to follow you are not geographically near by, than Foursquare buddies. One purpose of Foursquare is to increase the likelihood of bumping into known people, for example, Manhattanites bumping into other Manhattanites. So the networks have reasonably high geographic overlap. But that is not the case with Twitter, since the service serves as a way for people to connect through weak ties, and without much concern for geography, aside from the limitations inherent in natural language (Brazilians are more likely to follow Brazilians, but they may live very far away.)
So, I think the slippage of geolocational information from a closed, stable system like Foursquare into an open, dynamic system like Twitter is less problematic than generally considered. I don’t think it, per se, is scary.
While it is possible that a cadre of burglars or a sex slave ring might try to eavesdrop on our geolocational information in these services, history would suggest that our so-called friends and acquaintances are actually the source of most of these dangers.
People are scary, not social tools.
What is needed is a way to create more sophisticated circles of trust, in the final analysis, to minimize the likelihood of various dangers. Consider a young woman, Chloe, who has a close set of confidants — say 15 friends, both male and female — to whom she is extremely close. She also is part of a larger scene of 100 people or people that she sees frequently, but knows less well. And she may part of a even larger sphere — the PR industry, people she knew in college, acquiantances from work, friends of her ex-boyfirend — that she sees occasionally.
Imagine if her geolocational information was propagated in correspondingly less detail as her Foursquare posts moved outward through these circles of trust. Her inner circle might see exactly where she is — a certain corner of a certain bar — and also might receive that information in real-time.
Her 100 or so good friends might learn that she is in the Meatpacking district, or Nolita, but specifics would be blurred. So if one of that 100 had been invited to the same party, for example, they might be able to infer that Chloe was there, too. But they would have to directly ask her to get confirmation, and she could simply opt not to respond. And that information might be delayed by 15 minutes or 30 minutes, also.
And the next few hundred would know only that she was in some larger corner of the city, and that information might be delayed by an hour or more, serving more as an informational marker than an invitation to connect face-to-face: a way that she could transmit her hipster identity, and the identities of those she was out with, as a kind of social marker of her place in the world.
At any rate, we should expect that Foursquare, Gowalla, and others in this niche will experiment with a richer, more fine-grained sort of gelocation based on a tighter implementation of circles of trust. That is the real answer to these questions, because people are doing much more than simply posting their locations through these tools: they are constructing identities that they use to define themselves, and that has to be done in a public sphere. That’s why this is about publicy, not privacy.
Update on Friday, February 19, 2010 at 10:38AM by 
Stowe Boyd
Adina Levin points out (on Twitter) that geo information moving from Foursquare to Twitter would be best captured in a circle of trust model, there, too. But if we default to a simple approach of simply treating contacts in Twitter as relatively untrusted acquiantances — as Chloe’s did with her largest social circle in the example above — then things sort of work, but not in the best possible way.
