Targeted advertising is often incredibly spot-on, leading some people to believe that social networks and advertisers are using their phones' microphones to spy on them. While that's just a myth, Spotify appears to be exploring the possibilities offered from listening to its subscribers. The company filed a patent detailing how it could use microphones to determine people's "emotional state, gender, age, or accent," according to Music Business Worldwide.
The patent was filed all the way back in February 2018 and awarded on January 12. Music Business Worldwide was able to obtain a copy of the full patent (PDF), but the gist of it is that Spotify wants to use speech recognition to make its recommendation algorithms better. The patent describes that Spotify would analyze speech input for emotions, such as happy, angry, afraid, sad, or neutral, which would allow the platform to start playing just what you need to listen to right then. The company also intends to throw in environmental sounds to the mix, like "vehicles on a street, other people talking, birds chirping, printers printing, and so on," allowing for context-based recommendations.
As with any patent filing, we have to remind ourselves that these are often just experiments and ideas floating around in a company, so it's possible Spotify isn't even planning on introducing anything like this at all, or at least not in the near future.
While it's creepy on the surface, this feels like the natural next step following regular speech recognition and might even make communicating with machines easier or more natural feeling. I also wouldn't be surprised if this was something other companies are working towards, as well, particularly those selling smart home speakers like Google, Amazon, and Apple. All that said, I'm still getting some creepy sci-fi vibes à la Her from this patent.
Head to the source link below if you're interested in a deep dive into the technological background of the patent.