Spotify Coders Haven't Written Code Since December

Imagine fixing a bug in your company's app during your morning commute, using only your phone and a chat app. That's now reality for Spotify's engineers, and it's transforming how one of the world's biggest streaming platforms gets built.

Spotify co-CEO Gustav Söderström dropped a surprising fact during the company's recent earnings call: their best developers haven't written a single line of code since December. Instead, they're using AI to do the heavy lifting.

The company built an internal system called "Honk" that lets engineers deploy code remotely in real time using AI, specifically Claude Code. An engineer can message the AI from Slack on their phone, ask it to fix a bug or add a feature, and receive a working version of the app back before arriving at the office. They can then push that update to production without ever opening their laptop.

The results speak for themselves. Spotify shipped more than 50 new features and changes to its streaming app throughout 2025, including AI-powered Prompted Playlists, Page Match for audiobooks, and About This Song.

Söderström says this approach has sped up coding and deployment "tremendously." But he's clear this is just the beginning of what AI can do for development, not the end of the road.

The Ripple Effect

This shift could change what it means to be a software engineer. Instead of spending hours typing code, developers now focus on directing AI and making strategic decisions about what to build. They become more like architects and less like construction workers.

Spotify is also building something other tech companies can't easily copy: a unique dataset about music preferences. When you ask what makes good workout music, the answer varies wildly by person and place. Americans often prefer hip-hop, though millions love death metal. Many Europeans choose EDM, while Scandinavians lean toward heavy metal.

That messy, subjective data is exactly what makes Spotify's AI special. There's no single right answer, and that complexity gives them an edge as they retrain their models.

The company is being thoughtful about AI-generated music too, requiring artists and labels to indicate in metadata how songs were made while actively removing spam from the platform.

For anyone worried AI will eliminate tech jobs, Spotify's approach suggests a different future: one where humans and AI collaborate, each doing what they do best.