Spotify’s Discover Weekly: How machine learning finds new music

Pretty cool to see them go after the cold start problem with audio models. Often measuring the impact of having cold start vs. not is very difficult because it only affects the long tail of tracks at first, but eventually it could change the dynamics of discovery in the system.

Google Music's "I'm feeling lucky" is awesome. This is very subjective, but I feel it makes much better predictions as to what I'll like than Spotify. Maybe someone else has a different experience?

> Unlike Netflix, though, Spotify doesn’t have those stars with which users rate their music. Instead, Spotify’s data is implicit feedback 

I wish they kept it simpler with ratings, and explicit instead of implicit. The discovery weekly playlists are absolutely horrible for me to the point that I don't even bother checking them nowadays.

What works better for me for discovering new music with spotify is right clicking on an existing playlist and then "Create Similar Playlist" - that gives way more control over what kind of genre/style should the playlist consist of.

Yup. Gotta feel sorry for last.fm, they should have been front and center in this industry but let it slip away.

I use Spotify's Discover Weekly on a regular basis. On average I like more than half of the songs on the playlist, and it's my main way of discovering new music.