Spotify Explores AI Voice-Cloning To Translate Its Arsenal Of Podcasts
By Nicole Rodrigues, 26 Sep 2023
It seems like almost everyone nowadays is starting a podcast. And with so many springing up across the globe, it can be challenging for listeners to tune in to a series not spoken in their tongue. Spotify is here to save the day as it dabbles with voice-cloning technology to globalize podcasts and offer a multilingual experience while preserving the distinctive tones of your favorite hosts.
The main idea behind ‘Voice Translations’ is to translate the podcaster’s speech into various languages. By doing so, the streamer intends to deliver an authentic listening experience to audiences worldwide, breaking down language barriers.
While specific technical details remain undisclosed, it appears that Spotify’s approach involves voice cloning—a process that trains a computer model to generate speech based on samples and text inputs.
However, this technology has prompted concerns due to its potential for misuse in fraudulent activities.
Spotify, in turn, envisions a more positive application for voice translation. The platform aims to facilitate real-time translation, thus expanding the accessibility of content and introducing listeners to a broader spectrum of podcasters from diverse cultural backgrounds.
At present, Spotify’s Voice Translation is in the pilot phase, featuring a number of podcasts hosted by notable personalities such as Dax Shepard, Lex Fridman, Steven Bartlett, Monica Padman, and Bill Simmons. Notably, this feature is available to both Spotify Premium and free users in French, Spanish, and German for a select number of episodes.
Users can find this new feature on supported episodes within the Now Playing View as it rolls out over the next few weeks.
[via Fast Company and PC Mag, images via Spotify]