Pandora: Opening a Better Box PREV / 2 of 2 / NEXT ult following among musicians. I really started thinking a lot about this in 1999, when online music was taking off, and held so much promise. In spite of its inherent ability to connect people one on one, the online music space quickly became an electronic replica of the traditional business. Recommendation systems that were used just compared your shopping or listening preferences to those of others before making suggestions, a methodology that remained inherently biased toward what is already popular. I resolved to make it easier for people to find new music to enjoy, based on what the music they already love sounds like--and to do it in a way that was based on musical qualities, not popularity, or some kind of exclusive editorial taste-making. I wanted a more objective methodology, and one that would not be inherently biased against lesser-known bands. Our journey started more than five years ago, when we created the Music Genome Project methodology and started analyzing thousands of songs, one at a time. Nolan Gasser, a Stanford Ph.D. and musicologist was instrumental part of this effort. Together with a number of other very talented musicians, we broke down music into its component parts--kind of like musical primary colors. It was kind of a crazy idea at the time. TWEED: Did you have any personal bias on what music you wanted to emphasize in either project? TIM: Not really. We just look for high-quality music in all genres. We don't care whether it was created by a well-known band that has already achieved commercial success or one that is just starting out. TWEED: You've been involved with the world of music in many ways--as a musician, composer, and producer among other things--do you still work in these other fields or is Pandora your primary focus for the time being? TIM: I'm afraid my practice hours have been rather diminished over the past few years... But one of the great things about working in a company full of musicians is that there are weekly jam sessions so there are regular excuses to play music, if only for fun. TWEED: What do you see as Pandora and the Music Genome Project's mission for the future and are you going to continue the Music Genome Project until you map every piece of music in existence? TIM: Our primary goal will always be to help music lovers discover and enjoy music they love. In so doing, we hope to affect many positive changes on the industry. We hope for example, that many, many more musicians will be heard--and perhaps that someday because the Music Genome Project is so widespread, making a living as a musician will become a reality for tens of thousands of artists, rather than hundreds. We also hope the sharing of music across continents will bring world peace... but more on that later. As far as the genome goes, we will continue to add new music as fast as we can (thankfully, we'll never catch up)--including expanding into new areas like world music. The playlist algorithms will always be a work in progress and we don't see an end to possibilities for exciting new features. TWEED: Once again, congratulations on taking on such an ambitious and valuable project and making it so accessible. Thank you so much for your time and good luck with this and all your other musical projects from here on out. TIM: My pleasure. --Kevin Michell Monday, 23 January 2006 (c) Copyright 2024 Tweed Media PREV / 2 of 2 / NEXT