Spotify has been patented with technology that aims to use recordings of users’ speech and background audio to determine what kind of music to curate and recommend, Music Business Worldwide reports. The company applied for the patent in 2018; it was approved on January 12, 2021.
The patent outlines possible applications of technology involving the extraction of “intonation, stress, rhythm, and the like” from the user’s voice. The tech could also use speech recognition to identify metadata points such as emotional state, gender, age, accent, and even environment – i.e. whether someone is alone or with other people – based on audio recording.
The patent filing describes how Spotify is currently using a decision tree – showing users different artists, genres and more – to fine-tune the user recommendation algorithm. “What is needed is an entirely different approach to collecting flavor attributes from a user, particularly one that is rooted in technology, so that the human activity described above (e.g. forcing a user to provide input) is at least partially eliminated and more efficient, ”reads the application Find the patent below.
It is currently unclear whether Spotify has set a roadmap for implementing this technology in its desktop or mobile apps, or what this implementation could look like. It is also unclear whether the technology currently exists or if the patent is speculative. It should be noted that it is not abnormal for technology companies to patent technology that does not end up on the market.
A Spotify spokesperson gave the following statement to Pitchfork:
Read “Can Spotify’s New Detection Mode Be Considered Payola?” over on the field.