Panodra, Programming and Personalization


I'm really enjoy Postmodern Jukebox right now, thanks to Jon Samuelson and Curt Rees. I know, I know. You have to get over the creepy clown first.

Every time I listen to Mumford and Sons, I think of Quinn the Business Bohemian. Whenever I hear Sufjan Stevens, I think of Paul the Programmer. When I listen to Of Monsters and Men, I am reminded of the acoustic version of "Little Talks" and the way Jabiz the Teacher Troubadour reminded me of it. I think of Andrew when I listen to Pedro the Lion or Iron & Wine.

On the other hand, when I listen to Band of Horses, I have no one to thank aside from an algorithm. Same goes with The Postal Service. I'm glad I "discovered" this music, but the experience was non-relational. Pandora's algorithm suggested that if I like Ray LaMontagne, I might like Band of Horses. Pandora was right. The algorithm was accurate.

In the first case, a social network my friends recommended music based upon what they knew about my values, my artistic mindset, my tastes, my language. It was complicated and it was inaccurate. Pandora has never told me to listen to Elliot Smith, while Jabiz keeps pushing me to try his music out again. Then again, it took me three years to get into Thrice and I'm glad I did.

In the case of my friends, technology has been a part of the process. It's been the platform that enabled the conversations on music, the sampling of it and the downloading of it. However, it's been socially driven. These conversations might be happening on Twitter, but they might as well be happening over a pint.

Call me crazy, but I'd rather trust the messy, sometimes wrong, personalization that comes from a personal relationship than a precise thumbs-up-or-thumbs-down algorithm.