Musicians Demand Spotify Not Develop Emotional Speech Recognition Patent
Rage Against the Machine’s Tom Morello, Kliph Scurlock of The Flaming Lips, and all of Harry and the Potters are among the nearly 200 signatures on an open letter to Spotify firmly asking the streaming service not to develop a patent granted earlier this year for tech that can identify emotions in people’s voices.
The patent granted in January describes how Spotify could use its voice recognition tech to infer how someone is feeling by the sound of their voice. The software would also attempt to determine other aspects of the user’s identity, including gender, age, and accent. The resulting profile would then be combined with location data shared with Spotify to generate a playlist of songs that the AI suggests might appeal to the user at that moment. It’s essentially a far more sophisticated and personalized version of the recommendation algorithm used by Spotify right now. Who the listener is, where they are, and how they are feeling would presumably produce a playlist with many more songs the user would like to listen to compared to relying on only their previous listening history.
The nearly 200 musicians, bands, and civil rights groups signing the open letter to Spotify CEO Daniel Ek are concerned with everything else that embedding the tech in Spotify’s new voice assistant might cause. The letter raises several issues, largely stemming from how much the AI can learn about someone and what is done with that data. Worries about privacy, data harvesting, and surveillance are part and parcel of the ongoing debate over using any cloud-connected tech, voice or otherwise, but the Spotify tech adds a few more bullet points that are raised less often in those debates.
The letter alleges the tech could be used not just for monitoring people’s emotions, but manipulating them, giving Spotify’s algorithm the power to alter how people feel, presumably via the songs it queues up to play. There’s also the question of discrimination and whether the AI’s attempt to discern gender or age might lead to, at best, awkward stereotyping. The other element raised in the letter is less about the potential tech and more about the very concept of AI-based music suggestions and if it hurts the industry as a whole by shuffling songs for maximum profit over actual good recommendations.
“You can’t rock out when you’re under constant corporate surveillance,” Morello, one of the highest-profile of the signatories said in a statement. “Spotify needs to drop this right now and do right by musicians, music fans, and all music workers.”
Organized by Access Now and Fight for the Future, the list of musical acts hits a broad array of genres including rock group Eve 6, hip-hop artist Talib Kweli, and ‘wizard rock’ band Harry and the Potters. Access Now sent a similar letter a month ago, basically repeating the talking points, but without the star power of the signers. Spotify responded that the tech doesn’t exist in any of its products and there aren’t any in the pipeline, but that has not appeased them. The musicians and civil rights groups conclude it’s better to not uncork this particular AI genie. It may be too late to try and limit the tech at this point, as Spotify is not alone in working on emotion-detecting AI. Cerence has tested the concept for cars, as has SRI and Toyota. Amazon itself has a patent for similar functionality. Still, the group wants Spotify to pledge that it will never further develop or sell the patent at any point.
“There is absolutely no valid reason for Spotify to even attempt to discern how we’re feeling, how many people are in a room with us, our gender, age, or any other characteristic the patent claims to detect,” Access Now U.S. Policy Analyst Isedua Oribhabor said. “The millions of people who use Spotify deserve respect and privacy, not covert manipulation and monitoring.”
Follow @voicebotai Follow @erichschwartz
Spotify Patents Emotional Speech Recognition Tech for Song Recommendations
Amazon Files for Patent to Detect User Illness and Emotional State by Analyzing Voice Data