Spotify Tests AI-Powered ‘Prompted Playlists’ for You
▼ Summary
– Spotify is launching a new “Promoted Playlists” feature, initially for Premium subscribers in New Zealand, giving users more control over its algorithm.
– The feature creates personalized playlists based on long, specific user prompts and considers the user’s entire Spotify listening history, not just recent activity.
– It is an evolution of Spotify’s existing AI playlists, now allowing longer prompts and incorporating “world knowledge” for more nuanced results.
– Users can fine-tune prompts, schedule how often the playlist refreshes (like daily or weekly), and get context for why songs are recommended.
– This launch is part of a broader trend where social apps like Instagram and Bluesky are also introducing features that let users influence their algorithmic feeds.
Spotify is introducing a new feature called Prompted Playlists, giving its Premium subscribers in New Zealand a novel way to shape their listening experience. This beta tool, currently available only in English, represents a significant step in user-driven personalization. It allows listeners to craft detailed text prompts that generate playlists reflecting their entire listening history, not just recent preferences. The company frames this as putting more control over its powerful recommendation algorithm directly into the hands of its users.
Unlike last year’s AI playlist feature, which also used written prompts, this new iteration supports much longer and more specific instructions. A Spotify representative explained that this is possible because the underlying technology now incorporates world knowledge, allowing it to understand complex, nuanced requests. For example, a user could ask for “music from my top artists from the last five years” and then refine it to include “deep cuts I haven’t heard yet.”
The functionality goes beyond simple song requests. Subscribers can design playlists for specific activities or moods with remarkable detail. Spotify provided examples like requesting “high-energy pop and hip-hop for a 30-minute 5K run that keeps a steady pace before easing into relaxing songs for a cool-down.” Another suggestion was to ask for “music from this year’s biggest films and most-talked-about TV shows that match my taste.”
A key differentiator is the feature’s ability to tap into a user’s complete Spotify history “from day one.” Users can also schedule how often a generated playlist refreshes, such as daily or weekly, and continue to fine-tune the initial prompt for even more precise results. This essentially lets people create their own customized versions of popular Spotify staples like Discover Weekly or the genre-focused Daily Mixes, but tailored to a specific theme, era, or activity they define.
To aid users, the playlists will include descriptions explaining the rationale behind the song selections. Spotify will also provide a set of starter prompts to help people begin experimenting with the tool. The feature is seen as an evolution of the platform’s existing AI offerings, with plans to refine it further before a potential wider release.
Spotify’s move aligns with a broader trend among social and content platforms emphasizing user algorithm control. On the same day, Instagram launched a feature letting users dictate the types of Reels they see. Similarly, the decentralized platform Bluesky allows its community to swap out its default algorithm for custom alternatives.
(Source: TechCrunch)
