NYT Article : The Song Decoders at Pandora

The New York Times just posted a really insightful article exploring the algorithms employed by Pandora to customize listeners’ playlists.  It also delves into the discussion of music taste based on social and cultural context rather than the music itself, which is one of my favorite thinking points (see: this post from about a month ago).  Example: if you didn’t know who Celine Dion was, how would your perception change if you were to hear a song of hers?

I’m really into the idea of doing away with a lot of the “extra-musical” context of artists – where they’re from, what social scene is associated with them, whether your friends think their cool, what they wear, etc.  In an ideal world, you would just take music on its merits, but this is clearly impossible in the Internet age.  I’m certainly guilty of contextualizing artists outside of their musical merits – it’s one of the reasons I can’t stand to listen to Kanye West, for example.  It’s not that the music is bad, it’s that he’s such an incorrigible egomaniac that I dismiss it “on principle.”

Pandora seems like the closest thing we currently have to this kind of ideal, because it offers you music based on your taste, not that of your friends (iLike), major label preference (radio stations), or hipness (blogs).  Anyway, the article does a really great job of summing all of these thoughts, so here you go (click image for link):

Graphic : New York Times

Graphic : New York Times

New York Times + Radiohead + Yo La Tengo

new_york_times_logoRadiohead BearYo La Tengo

Two of my favorite bands popped up in the NYT’s RSS feed today.

The Radiohead article talks up the band’s prowess in navigating the current and future music markets, and the Yo La Tengo article is an interview with Georgia and Ira about their influences (and influencing other bands).  Nice.