| Sun | Mon | Tue | Wed | Thu | Fri | Sat |
|---|---|---|---|---|---|---|
| 1 | 2 | |||||
| 3 | 4 |
5 | 6 | 7 | 8 | 9 |
| 10 | 11 | 12 | 13 | 14 | 15 | 16 |
| 17 | 18 | 19 | 20 | 21 | 22 | 23 |
| 24 | 25 | 26 | 27 | 28 | 29 | 30 |
| 31 |
« Networks in political science | Main | More on networks in political science and sunbelt... »
19 September 2007
There has been a spate of stories (also this story from globe) recently about the use of cell phones to track the locations of your friends. There has also been some talk of linking data from the social networking sites to cell phones, so, for example, you could walk into a room and instantly find someone who you had never met, but who was a friend of a friend (through facebook or some other database that was stored in the phone).
These are clever ideas, but let me throw out another idea: a program for phones that facilitates those “you grew up in Long Island? Did you know John Smith, by any chance?” moments. (A small digression—this exact question (except for the name) was posed to me once. And, surprisingly, I actually did know the person in that case.)
The basic idea is quite simple—if you are talking to someone, and you both have the program, have the phones link via Bluetooth and:
1) find overlapping phone numbers and report them back. A more extensive version of this would also match any incoming or outcoming calls the phones have ever made.
2) collect and match structured data about the owners of the phones—where and when you went to school, where you’ve lived different years, etc., and report back matches. This is all of that information that people initially exchange when they meet. While no substitute, this could be faster and more thorough way to those “Do you know” questions. That is, you would both instantly find out if you both happened to live in Ann Arbor for the same two years in the 1990s.
This would not be a hard program to write, but there is the classic chicken-egg problem of how to get enough people to sign on to make it work. Not my problem—but if you write such a program, let me know.
Posted by David Lazer at September 19, 2007 4:34 PM
I would assume such a program would be even better if it could send a temporary client to the other phone -- thus avoiding the chicken & egg problem. Confidence is a big issue with self-replication applications and urban legend & similar horror stories would kill the ploy. I would recommend to make the send-out client small (brand or device specific) and one way: I send you my list of numbers, and a Java hacks tells you what number we share; you look at it and tell me if anything pops out. If it's anything you should be ashamed off, just lie and don't mention it.
Posted by: Bertil at September 20, 2007 10:48 AM
Isn't there a more dangerous side of this to consider--that very easily people could track down and target certain people for malicious purposes? I understand the whole idea of a connected world, but it is also at the rick of being too connected.
Posted by: Okinawa at September 28, 2007 9:46 PM