The feature is currently being tested with Premium users in New Zealand and will expand to other markets after the beta phase.
According to Spotify, this tool allows users to describe the specific type of music they want to listen to in their personalized playlist. Unlike existing lists, the new feature is not only based on current interests but also exploits the entire music listening history of users since they started using the application. This is a difference from other AI playlistings previously developed by Spotify.
Users can use text reminders to request a music list suitable for each activity or mood. Requests can be longer and more detailed than before, thanks to the upgraded language processing capability. Spotify says the tool can take combined knowledge into account to make better suggestions, not just based on personal music listening habits.
The feature also allows users to review their listening history and select the frequency to refresh playlists, such as daily or weekly. With this approach, each person can create their own versions of famous lists like Discover Weekly, but focus on the theme or stage of the music they want to follow.
Spotify gave examples such as the request for music from the artists I have loved for the past five years or more complex suggestions such as velvet pop and hip hop for a 30-minute run, then switch to a light melody to relax. The playlist will also come with a description and recommendation reasons so that users understand why they received the suggestion.
The trend of giving users algorithmic control is being applied by many platforms. Instagram has just added a short video content management tool. Bluesky, a decentralized social network, also allows users to completely change the display algorithm.
