You Can Now Use AI to Create Playlists on Spotify
You may have noticed that generative AI has made its way into an app or two gadgets lately, and now you can use the technology to create new playlists on Spotify: the AI Playlist feature is now available in the US on Android and iOS. after testing in the UK and Australia, although you’ll need to be a Premium subscriber to access it.
It works the same way as text and image AI chatbots: you come up with your own text prompt, which forms the basis of a new playlist – this could be “songs for a thoughtful mood” or “songs for a birthday party with a fun mood” . a bunch of five-year-olds.” Here’s how to get started with the feature, as well as some ideas for finding old-school music that might work better.
Creating AI Playlists in Spotify
This new Spotify feature is another reminder that the term “AI” is a broad one. The algorithms that are already used in your recommended playlists on Spotify also rely on a different type of artificial intelligence, but this new AI Playlist tool takes that idea up a notch. Instead of just browsing songs related to the music you’re currently listening to and liking, you can be much more specific about what you want.
“The more specific you are, the better,” says Spotify . “The most successful playlists are created with prompts that contain a mix of genres, moods, artists or decades, so get creative! You can also reference places, animals, activities, movie characters, colors, and even emojis.” This gives you a lot of options to work with.
However, there are some restrictions: you can’t create tips using links to brand names or current events. There are also precautions in place to ensure that offensive prompts are excluded, so it’s probably not worth trying to test the boundaries of decency with this new tool.
Assuming you’re a Spotify Premium user, open the app on Android or iOS, then go to the Your Library tab. Click the + (plus) button in the top right corner and select “AI Playlist” from the menu that appears. Now you need to get creative: describe what you want from your new playlist in as much detail as possible. You can base it on mood, events, genres, favorite artists, time of day, activity, or whatever you like.
You’ll also see some suggestions above the text entry field that you can use if you don’t have any inspiration for your own prompts. Swipe left to see more recommendations – you’ll see that you can limit your playlists to a certain length or a certain decade of music, for example, which can then give you more ideas on how to create your own playlists.
Submit a request and the playlist will appear. You can refine it using the hint box at the bottom of the screen, tap any song to hear a preview of it, or swipe left on a song to remove it from the list. When you’re ready to accept the list of songs provided to you, click “Create” to save it.
AI playlists versus traditional music discovery
Based on my limited time testing the AI Playlist feature, the underlying algorithms don’t seem to be too different from those used in the rest of Spotify (such as when you ask it to automatically populate a playlist). Mention “indie rock” and you’ll get a lot of the same songs you’d get in a regular search. However, the text prompt interface gives you more flexibility when creating a playlist.
It’s AI, so of course there are misfires (Spotify also labeled it “beta”). When asked “REM-type songs that aren’t REM,” the AI returned—you guessed it—a playlist including several REM tracks. Some fine-tuning is often required, as is common when generating text or images using generative AI, and the ability to add hints can be helpful.
AI Playlist appears to take your previous listening history into account when creating playlists and mixes songs you’ve heard before with tracks that are popular elsewhere. The uptempo “upbeat songs for a long winter drive” included tunes from the likes of Future Islands, Taylor Swift, Coldplay and 1975 – a mix of what I’d already heard and what I hadn’t heard yet.
It’s certainly an interesting idea, and even if the playlists aren’t much different from the ones you might already find by digging through the recommendations on Spotify, it’s fun to play around with some ideas: suggestions like “songs that will help me forget the past.” or “the perfect playlist for watching a rainy cityscape,” for example (the latter, by the way, was brought back by Bon Iver, Tracy Chapman and Radiohead).
As with most AI, it’s unclear how it actually works, but I can’t help but feel somewhat disgusted by algorithm-based AI recommendations in general. Sure, they’re based on a lot of data, but they often lack the random or thoughtless discoveries that used to come from picking up a music magazine, flipping through music video channels, or listening to mixtapes made by friends.
Personally, I still prefer the old ways of discovering music, including reading album reviews and band interviews, and asking friends for recommendations (Spotify’s Blend feature is actually very good for this). Or, to take another recent example, I browse bands that support bands I already like—a choice driven by people and their tastes that often results in new favorites.