June 13, 2004
MIT Serendipity Automates Chance Encounters

Serendipity will help find people you want to meet.

Serendipity, a form of next-generation networking, was developed by Nathan Eagle, a graduate student and Media Lab Europe Fellow working with Alex (Sandy) Pentland, the Toshiba Professor of Media Arts and Sciences in the Media Lab’s Human Dynamics group.

The system uses Bluetooth, an RF (radio frequency) protocol that works like a low-power radio in most cell phones, sending out a short-range beacon. “Think of it as each person having a 16-foot bubble around them, blinking out a unique ID," Eagle said. “When two or more people running Serendipity come into the same ‘bubble,' their IDs are sent to our server, which looks for their profiles. If there’s a match, each gets the other’s name, thumbnail photo and common interests on his or her cell phone." Then it’s only a matter of introductions.

And it’s quick. The server scans for IDs every 60 seconds and only takes about five seconds to find a match, so the whole sequence takes about a minute at the most.

How does the server know about your interests? Just like web-based social network systems like Friendster or match.com, Serendipity depends on profiles that users write about themselves. But Serendipity is unique because it allows the user to “weight” his or her profile to emphasize interests that are of greatest importance to the user’s current social situation.

This reminds me of a proposal I made a few months ago:

Another possible interesting application would be to manage affinity groups. Imagine a traveller who is cruising down a road trying to decide which night club to try out. If people registered with an affinity tracking service then a traveller could choose a club or restaurant whose currently present patrons fit some desired demographic profile. One obvious problem with such a service is that just because one person likes a particular type of person doesn't mean that most who fit a desired profile will like that person in return. Look at celebrities for example. They are loved by all sorts of people who the celebrities would very much like to avoid. So a service would need to develop eligibility criteria that require matching of preferences in both directions before that person driving down the street would get a flashing light on their car LCD pointing them to a particular bar or night club.

Now I'm actually expecting to see this sort of thing to really be implemented and to become widely used. For bar scenes one of the difficult challenges will be the development of image processing software that can analyse the image of a person you haven't even seen yet to decide whether you might find that person attractive. You could just drive through downtown and be told where to stop. In a bar situation the algorithm would have to be fairly sophisticated and use not just images of a person and background info but also your degree of inebriation (higher levels mean lower standards - imagine an embedded nanotech sensor reporting blood alcohol to your cell phone), the time of night (later means lower standards), how long it has been since you last hooked up, and perhaps similar information from the other person to factor in whether you both ought to be told by your avatars to seek each other out.

Heck, the avatar might even tell you how many drinks you'll have to drink to be able to feel that your realistic choices are acceptable. In the longer term as neurobiology and neurochemistry become more advanced you will be able to have embedded implants installed that will release compounds to make you find many more people attractive than you would naturally. Of course in the longer term gene therapies, stem cell therapies, and other therapies will raise average attractiveness that this will be far less of a problem anyhow. As I've argued previously, the female desire for high status males is going to be harder to solve than the male desire for more attractive females.

Of course, some statistical outliers will actually use this technology to meet more interesting people. I'm not trying to argue that the only application for this technology is meeting people for sexual hook-ups. But my guess is it will find wider use for sexual purposes than for intellectual ones.

Share |      Randall Parker, 2004 June 13 06:05 PM  Comm Tech Society


Comments
John Atkinson said at June 16, 2004 2:38 PM:

Christ, imagine the convergence between this kind of technology and 'buzz marketing' techniques. Not to mention the intense social paranoia that will take root in more awkward/old-fashioned types - imagine going through bars, parties, parks and all of life wondering if that 'stranger' looking at you is 'just' looking at you, sizing you up as a potential mate, sizing you up as a potential marketing prospect, looking for business networking, corporate espionage, ex-girlfriend's new boyfriend, &c, &c... Hopefully these systems will co-evolve with anti-anxiety implants.

Sean said at June 17, 2004 3:23 PM:

RE: "imagine going through bars, parties, parks and all of life wondering if that 'stranger' looking at you is 'just' looking at you, sizing you up as a potential mate, sizing you up as a potential marketing prospect, looking for business networking, corporate espionage, ex-girlfriend's new boyfriend, &c, &c..."

...but people could have indulged those same paranoias 50 years ago.

Post a comment
Comments:
Name (not anon or anonymous):
Email Address:
URL:
Remember info?

                       
Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©