Spotify's playlist features may have taken the music app world by storm but long-time rival Pandora is apparently not ready to take that lying down and has a string of server-side updates ready to fight for position. In fact, those new playlist features should have already been hitting some users of the company's premium subscription and have been in testing since back in December. However, the changes aren't substantially different from a user's perspective. Driven by Pandora's long-running and ever-evolving Music Genome database, which takes hundreds of song attributes into consideration before making a recommendation, the company's new playlists take things a step further with machine learning. Of course, the playlists it creates will continue to change over time as the algorithms take user input from thumbs up or thumbs down ratings and effectively learns their musical tastes. But this should fundamentally change the how accurate those playlists are and pave the way for much better-curated content discovery. For free users, the rollout is more gradual but premium access is attainable by watching advertisements for each play session and to share playlists with friends via the Android app's Access feature.
The playlists themselves show up in the Browse section of the app interface and will be based on listening preferences, in addition to containing new music users haven't listened to before within the service. They'll be generated based on likes and dislikes alongside Pandora's existing algorithms and be automatically named following the same standards. For example, a lot of party-worthy upbeat music may show up as "Your Party Soundtrack." Lists will also be curated based on time of year, events, and holidays. With that said, they'll also be swapped out on a weekly basis, so users will need to save a playlist to their own playlists if they want to keep them - with the capability of saving up to 60 playlists at any given time.
The goal here, according to Pandora, is to take features that make Spotify popular and improve on them as much as possible using modern machine learning and a nearly-decade old process. With any luck, the improvements will be enough to, at very least, help the company hold its position as one of the world's most popular apps for music.