Foursquare announced last week that they’d start recommending restaurants based on your location and previous check-ins. Perhaps this is a logical extension of the service that doesn’t surprise many, but I find it interesting for one reason; it brings together the social and algorithmic aspects of the web so well.
I have previously argued that the internet will become much more personal, much more customized to the individual using it over the next decade or so. I argued two things would push this change. First, an ability to know (through social networks) the clubs you belonged to and friends you have and thereby be able to predict your preferences. Second, an ability to predict your preferences through the application of advanced mathematical models to your known information and history. Foursquare making recommendations gives one app the ability to do both.
You can be visiting a brand new city that you’ve never been to before, flip on foursquare and get the following information, “There’s a new Chinese restaurant next door you’d probably like and Mike and Carroll are at the Greek place across the street, while the guy you met on the airplane on the way here is mayor of the sub shop around the corner.” Pretty amazing.