I remember the sound of the record button on my cassette boombox. I waited hours to record a new song without the DJ’s outro.
My prized possessions were mixtapes, curated song by song, and music discovery was an active hunt.
It meant digging through record stores, borrowing CDs from friends, and taking a chance on a band I’d only read about in a magazine.
That world feels like a lifetime ago. Today, finding new music is as simple as opening an app on your Android phone.
Here’s how Spotify’s AI features are making music listening easier for millions.
Spotify’s AI DJ reimagines the radio experience
AI DJ is a personalized, endless radio station. You tap a button, and music starts playing. A natural-sounding voice sometimes cuts in with commentary specific to you, not generic station IDs.
It might say something like, “Coming up next, some 2010s hip-hop you were into last year,” then queue a track you haven’t heard in a while.
It introduces new artists in context, explains why it chose a block of songs, and adapts if you tap the DJ button to change the mood.
AI DJ combines three technologies. The recommendation engine behind features like Discover Weekly drives the core song selection.
It analyses your listening habits, liked songs, and time of day to predict what to play next. It also learns from what you play and skips to refine its choices.
Generative AI powers the commentary. Spotify uses generative AI to write the DJ’s script.
It turns signals like “here’s a 1990s rock block the user will like” into natural commentary and can add artist facts or explain links between songs.
The voice is based on Spotify’s Head of Cultural Partnerships, Xavier “X” Jernigan. Spotify used Sonantic’s technology to model and synthesize his voice.
AI playlists let you build music from prompts
The AI Playlists are a co-creation tool. You can use the feature to generate a playlist by typing a descriptive prompt instead of adding songs manually. You can be highly specific.
Go beyond genres and ask for prompts like “songs for a rainy day in a cozy coffee shop,” “an upbeat playlist for cooking pasta,” or “music that feels like a main character in an indie film.”
You can also use emojis, places, or movie characters as prompts. AI Playlists run on a large language model (LLM), the same technology that powers ChatGPT.
When you type a prompt, the LLM interprets concepts, moods, and their relationships. It understands that ‘rainy Sunday’ implies coziness. ‘Cooking pasta’ suggests upbeat but not frantic.
‘Soulful’ points to specific musical traits and then translates this vibe into concrete musical parameters, such as genre tags, tempo ranges, and instrumentation.
Lastly, Spotify searches its catalog for songs that match those parameters and cross-references them with your listening history to tailor the list.
Spotify’s discovery engine predicts what you like with AI
This is the foundation of the Spotify experience, most visible in Discover Weekly and Daily Mixes. Each Monday, Discover Weekly delivers 30 new tracks it predicts you will like and have not heard.
Daily Mixes are more familiar. They take a genre or mood you like, mix in artists you know, and add a few new finds. They are the reliable mainstays of Spotify’s AI.
My Monday morning routine is simple. I grab a coffee, open Spotify, and press Play on Discover Weekly. Not every week lands, but the hit rate is acceptable.
It is how I found a little-known Australian indie band and why I enjoy synth-pop, a genre I had written off.
The Discovery Engine uses a two-part system.
Collaborative filtering is the friend-recommendation model. Spotify identifies users with listening habits close to yours—your musical twins.
It finds songs your twins play that you have not heard and recommends them. If people with your taste love a song, you likely will, too.
The second system is content-based filtering with the sonic-DNA model. Here, the AI analyses the raw audio.
Spotify breaks each song into hundreds of features, such as tempo, key, loudness, and more, and then finds other songs with a similar profile.
Listening to many fast, minor-key tracks with distorted guitars surfaces songs with the same fingerprint, regardless of artist or genre.
How Spotify Enhance adds songs that fit the playlist vibe
The AI playlist gives you new playlists, while Enhance refines your favorites.
You see an Enhance button at the top of personal playlists. Enhanced adds recommended songs that match the existing vibe.
It adds one suggestion after every two of your tracks, up to 30 recommendations. You can spot these picks, audition them, and add the ones you like with a single tap of the “+” icon.
Unlike broader discovery tools that use your listening history, Enhance focuses on a single playlist.
That is why its recommendations feel more accurate. It is not guessing what you like in general.
Spotify Wrapped has become a yearly tradition
Each early December, social feeds are filled with personalized infographics. This is Spotify Wrapped, an annual campaign that has become a cultural event.
It gives each user a visual listening summary from 1 January to mid-November. It is an interactive story showing your top five artists, songs, genres, total minutes listened, and other useful stats.
For me, Wrapped is a holiday-season highlight. I look forward to it, and there are always surprises.
Like discovering I became an artist’s top fan, all because I looped one song a thousand times after a heartbreak (ouch).
The best part is the IG stories and texts from friends. We compare top artists, laugh at our most played guilty pleasures, and spot overlaps we did not expect.
At its core, Wrapped is a data analytics project powered by machine learning.
Throughout the year, Spotify logs each interaction, such as songs streamed for more than 30 seconds, skips, playlist adds, and artists you follow.
This creates a large dataset for each user. When the November cutoff hits, Spotify’s algorithms get to work.
They process and aggregate billions of data points to calculate statistics such as total listening time and top artists.
Machine-learning models clean the data — for example, ignoring a track left on repeat overnight — and segment users.
Finally, the app turns that data into a shareable story.
AI is reshaping the music listening experience
Spotify still has room to grow if it wants to be the leading music streaming space, but it remains my go-to platform.
The partnership between human taste and Spotify’s AI marks a new frontier in music discovery. Mood-based listening via AI Playlists and AI DJ is breaking rigid genre boundaries.
Listeners are more open to artists from genres they would normally ignore if the music fits a vibe like ‘chill morning’.
And for artists, securing a spot on one of these playlists can be nothing short of career-changing.
