TechCrunch profiles a new music matching service called Pandora. Basically, it is a recommendation engine built on the components of a song, rather than what other users decided to listen to or purchase. They apparently have human listeners who tag each song with a wealth of information about the sounds and style. Then when a user plugs in a song or artist they like, the service will match up other songs with similar characteristics.
It certainly leads to some interesting results. At one point I punched in Enya to see what it would pair it with and up popped a Bette Midler song and another from Kenny Loggins. I was skeptical, but both songs had similar sound qualities to those in much of Enya’s music.
The service is apparently set to go to public beta in just a few weeks. Similar technology could have all sorts of applications in other sectors as well.