Spotify’s New Patent Can Listen to Your Speech to Suggest Music
I long ago accepted that our tech overlords own us and the government can come for us any time they damn please. It’s the flipside of modern technology’s connectivity, and it sucks, but railing against it seems pointless: this is the world we live in!
Spotify’s latest patent is sure to cause alarm for privacy wonks, a new feature that utilizes speech recognition to determine a user’s “emotional state, gender, age, or accent” and then uses that data to recommend content.
The new patent, titled “Identification of taste attributes from an audio signal,” was filed in February 2018 and granted on January 12 of this year. The patent provides a “method for processing a provided audio signal that includes speech content and background noise” and then “identifying playable content, based on the processed audio signal content.”
The goal, according to Spotify, is to bring another tool to the service’s arsenal of product recommendation. Instead of relying on user-inputted data such as location, age and gender, or learning a user’s personal musical preferences over time, the service would be able to identify a user’s likely tastes more quickly and adjust on the fly.
The software would monitor a user’s voice to gather information, measuring for “intonation, stress, rhythm and the likes of units of speech” to identify a user’s mood as “happy, angry, sad or neutral,” as well as pinpointing a user’s accent to make recommendations based on that. Environmental data such as “sounds from vehicles on a street, other people talking, birds chirping, printers printing, and so on” could be factors as well.
Content would be recommended based on all of this data, updated in real time, in concert with previous listening history, behavior of a user’s friends on the platform, etc.