Most of the tech industry is still debating the ethics of replacing programmers with artificial intelligence. Meanwhile, Spotify is quietly implementing a model where the role of the engineer shifts from writing code to overseeing it. During the company’s latest quarterly results summary, Gustav Söderström, co-CEO of the company, revealed a striking statistic: since last December, the streaming giant’s developers have virtually stopped writing traditional lines of code from scratch.
At the heart of this transformation is an internal system called Honk. This tool allows engineers to remotely configure and modify applications in real time, often via simple commands on Slack. From an operational perspective, this is a breakthrough – being able to fix a critical bug in the iOS version during the morning commute to the office drastically reduces iteration cycles and lowers downtime costs. For Spotify’s management, this is not just a gadget, but the foundation of a new efficiency strategy that aims to maximise results while optimising human resources.
But Spotify’s ambitions go deeper than just streamlining the back-end. The company is building powerful data sets to personalise the user experience on an unprecedented scale. New features such as smart playlists generated from descriptive prompts are just the tip of the iceberg. The aim is to contextualise: the system is supposed to understand that a request for ‘workout music’ should produce different results for a user in Stockholm and another for someone in New York, given local trends and habits.
Despite these successes, Spotify is treading on thin ice in its relationship with the creative market. Incidents involving projects such as The Velvet Sundown – bands fully generated by AI that are gaining mass popularity – are causing concern for traditional artists. The company is attempting to balance this conflict by involving music industry leaders in its work on ‘responsible AI’.
