Spotify’s New AI DJ Spins Generative AI and Voice Clone into Music Suggestions and Commentary
Spotify listeners can each have a personal DJ to pick songs they may like and even share insight about what they are listening to through a combination of generative AI and synthetic voice cloning. The company envisions the new feature as a way of shortening browsing and search time and enhancing the listening experience with an “AI DJ in your pocket.”
Spotify designed the new feature to curate tracks and provide commentary based on what users have listened to before. The “DJ,” which sounds like it was named by the same committee behind the now-defunct Spotify Car Thing, also learns to refine its choices based on what users tell it they like or dislike. The AI DJ’s comments on the music are produced by one of OpenAI’s large language models, likely a variant of the GPT-3.5 API. The digital DJ offers facts curated by Spotify’s editors about specific songs and artists, as well as the genre being played.
Listeners hear the commentary in the synthetic voice of Spotify’s head of cultural partnerships and host of its morning show, The Get Up, Xavier “X” Jernigan. Spotify processed samples of Jernigan speaking through the synthetic voice platform developed by Sonantic, which Spotify acquired last June. Sonantic’s technology is best known for generating a deepfake of Val Kilmer’s voice for the Top Gun: Maverick film. That’s a notable shift in-house after employing synthetic voice startup ReadSpeaker to create the Car Thing’s text-to-speech feature.
“We have experts in genres who know music and culture inside and out. And no one knows the music scene better than they do. With this generative AI tooling, our editors are able to scale their innate knowledge in ways never before possible,” Spotify explained in its announcement. “To create the voice model for the DJ, we partnered with our own Head of Cultural Partnerships, Xavier “X” Jernigan. Previously, X served as one of the hosts on Spotify’s first (and personalized) morning show, The Get Up. His personality and voice resonated with our listeners and resulted in a loyal following for the podcast. His voice is the first model for the DJ, and we’ll continue to iterate and innovate, as we do with all our products.”
DJ is still in beta and only speaks English, with access limited to Spotify Premium subscribers in the U.S. and Canada, with other markets to follow. Listeners can tap the Play on the DJ card on the Music Feed tab of the Spotify app.