Music streaming platform Spotify won a patent allowing it to listen to users’ emotions and background to recommend music. The idea is to use a person’s emotional state to recommend appropriate music. What might seem a good idea to the company is a very creepy feature for users who value their privacy.
We alerted people a couple of years ago about how this was taking place and now a patent has been granted.
The patent covers features that will allow Spotify to use speech recognition and analysis to gauge a user’s demographic attributes and emotional state. The app will also listen to a user’s background to determine where they are; taking a walk, at a party, on the bus, etc. The data the new features will allow the app to collect will help recommend the “perfect” music without requiring the conscious data input from a user.
Through speech recognition, the music streaming platform will be able to determine a user’s gender, age, and other demographic attributes. By analyzing “intonation, stress, rhythm, and the likes of units of speech” in a user’s voice, the platform will determine the emotional state.
The company also plans to use “environmental metadata,” i.e., background noise, to determine the location of users. Many people could find that highly intrusive.
Defending the new features, Spotify argues that the average user can’t input all the details required to fully determine their music preferences and tastes. Additionally, inputting all those details would consume a lot of time, which could be better spent streaming music.
The patent does not explain what emotions Spotify will encourage and what emotions it will try and shift.
According to the patent, the algorithm will factor in a user’s response to the recommended music.
Spotify had previously hinted at the features covered in the new patent.
In a study published in July 2020, the streaming platform claimed to have analyzed over 5,000 users who listened to approximately 17.6 million songs. Apparently, with the help of AI, the music streaming platform was able to predict personality traits and predict musical preferences with a “moderate to high accuracy.”
The study noted the company’s desire to conduct more research on linking “streaming behavior with brain scanning, genetic, and physiological data.”
Last October, Spotify secured another patent for “methods and systems for personalizing user experience based on personality traits,” RT reported.
The patent appears to have laid the groundwork for this one, which was approved earlier this month.
The company low-key acknowledged the potential dangers of the new features. In the study, published last July, it said, “a user’s digital history is extraordinarily personal and sensitive and should be treated with proper consideration of the conceivable misuses and unintended externalities.”
