Can software have good taste?
Until recently, the idea that a few—or even many—lines of code could predict the music you’ll like even before you’ve heard it seemed impossible. You might have said it couldn’t be done. But entire companies, even industries, depend on that proposition, including Spotify, whose preference technology we explore in the season two debut of our podcast, Actuality.
This week’s episode is inspired by Adam Pasick’s reporting on Spotify’s preference technology. At Spotify, we met with Matthew Ogle, the company’s product owner for “What to Play,” to learn how he teaches software to figure out just that. Then, we talk to musician and music industry thinker Kiran Gandhi about how labels use algorithms and why she’s optimistic about big data and music. Finally, we meet Ted Striphas, a professor at the University of Colorado who has done some deep thinking about what an algorithmic culture really means.
After you listen to this episode, help us build our own playlist of songs in the key of “it couldn’t be done.” And then give us some clues for what you want to see in future episodes. Who else is doing the impossible? Tell us on twitter, or e-mail us.
Actuality is a joint production of Quartz and Marketplace. Every two weeks, we’ll explore the inner workings of the new global economy, combining the best of our economic smarts.