Yet more indications that file-sharing is more likely to save music than usher in its demise. A radio station in San Francisco has begun using peer-to-peer music-swapping networks as a way to figure out what's hot amongst listeners. According to ABC:

It's the job of program director Sean Demery to figure out what people want to hear. One new way is by monitoring what file swappers are searching for and sharing most. And he does it with the help of a market-data software company called Big Champagne.

"It basically gives me pretty much what's happening in the mass culture," Demery said. "It tells me what's popular."

Eric Garland, CEO of Big Champagne, compares what his company does online to what the Nielsen rating system does for television.

Actually, peer-to-peer sharing is infinitely better than what Nielsen does. Nielsen's ratings are a mere sampling of a tiny, tiny fraction of the TV-watching audience; networks like Kazaa or Morpheus are filled with millions of millions of folks. (Though they may be biased towards a particular type of music, since file-sharers tend to be younger, I'd bet.) But the point remains: This is an example of people using tech to tap into the unruly spirit of the masses, something that -- for all its vaunted attempts to please its audiences -- the entertainment industry doesn't always do very well.

More importantly, it's an example of how artificial-intelligence techniques are going to become more and more useful. Up until now, they've been interesting curiousities -- but mostly useful to people like bankers, who have oceans of data they need to sift to find patterns. With peer-to-peer networks, everyday people are now generating oceans of data that we ourselves might want to sift.

(Thanks to Boing Boing for this!)