Many years ago I read about a dude who’d converted Unix server logs into a real-time auditory environment — specifically, a rain forest. Server load controlled the level of the rain, CGI calls were bird chirps, potential malicious attacks were the cough of a jaguar, etc. Sadly, I can’t find any info on this anymore. Anybody know anything?
A simpler, easy-to-implement application: you assign musical DNA traits to individual aspects of your collected data stream, and your Pandora/Last.fm/MOG/whatever retrieves music accordingly. For example: whenever I receive new mail, play “excited” music. Or whenever I sell an item on Etsy, play Iggy Pop’s “Success”. It’s not matching music to your mood, it’s matching it to your information landscape.
Once I turn on Stikki’s “local only” feature, I’m thinking of seeding the world with microcompositions — music only available when the user is in a specific location. Like “Soundtrack to the corner of Maryland and Harmon”.
Somebody ought to pay me to think about this stuff. Everybody email Joi and tell him to hire me at the Media Lab. 😉