Spotify has recently introduced Voice Translation for podcasts, a new pilot service that harnesses the power of AI to translate audio shows into multiple languages. By utilizing machine learning techniques, Spotify’s AI algorithms are able to maintain the original podcaster’s voice, resulting in a more natural sound in different languages.
This innovative Voice Translation feature is built upon a tool developed by Spotify, incorporating advanced AI audio technology, including contributions from OpenAI’s versatile speech recognition model called Whisper. OpenAI, based in San Francisco, has extensively trained Whisper on a diverse dataset of audio, enabling it to perform tasks such as multilingual speech recognition, speech translation, and language identification.
According to Spotify, translated podcasts sound more authentic and natural compared to traditional dubbing thanks to the capabilities of Whisper and its AI model. The distinct speech characteristics of the original podcaster are preserved even when a podcast recorded in English is offered in other languages.
As part of the pilot, a curated collection of AI-translated podcasts has been compiled on a dedicated Voice Translations Hub. This collection features episodes from well-known podcasters, including Dax Shepard, Monica Padman, Lex Fridman, Bill Simmons, and Steven Bartlett. Initially, only a limited number of catalog episodes and upcoming releases will undergo translation. Spotify has committed to providing translations in Spanish, French, German, and other undisclosed languages. The streaming service is also collaborating with other podcasters to include their shows in the translation service.
Ziad Sultan, Spotify’s VP of personalization, believes that offering translated podcasts while retaining the original voice of the creators allows listeners globally to discover and be inspired by new podcasters in a more authentic manner. Spotify aims to foster a deeper connection between listeners and creators by taking a thoughtful approach to AI, aligning with the company’s mission to unlock the potential of human creativity.
Voice-translated episodes will be accessible worldwide for both premium and free subscribers. The translations will initially be available in Spanish, with French and German versions set to be introduced in the near future.
Spotify’s introduction of Voice Translation for podcasts marks a significant step in broadening access to its podcast service for a larger audience of creators and listeners across different languages. By leveraging AI technology, the company aims to make podcast content more inclusive and enable a global community to connect with diverse voices in the industry.
