A US Spotify patent – “Identification of Taste Attributes From an Audio Signal” – originally filed in 2018, has now been granted from January 12, giving the streaming service the ability to record a user’s daily speech monitor in an effort to adapt recommendations to the user.
This patent also extends beyond everyday speech to background noise, and will help the platform determine the users’ “emotional state, gender, age or accent” and aims to quantify this data by attributing the input to a variety of emotions: “happy, sad, angry or neutral,” reports Music Business Worldwide.
These moods are determined by a number of factors and Spotify will monitor the “intonation, tension, rhythm and similar units” to make the determinations.
Currently, Spotify already collects some of this personal information through a short questionnaire, but this approach is now considered outdated and “boring” for the user. In fact, all the options you have to enter when setting up your profile fall unfavorably so that Spotify can listen to everything you say, whether it’s a bird chirping outside or cars driving.
In the submission, the apparent need for this invasive data collection method is further explained: ‘What is needed is a very different approach to collecting taste traits of a user, especially one rooted in technology, so that the above human activity ( e.g., which requires a user to provide input), is at least partially eliminated and executed more efficiently. “
Whatever Spotify recommends you in the future, we at least still have this AI bot that will roast your taste in music.