For your latest Radio World ebook about applications of AI in radio, we went looking for examples of how AI is being deployed by radio and audio companies more than two years after ChatGPT rolled out and changed everything.
We discovered AI tools being used by broadcasters, large and small, for purposes such as language translation; commercial and spec-spot creation; on-air voice imaging; creation of text, audio and video; audio processing; and yes, for on-air hosting.
Here is one example. iHeartMedia — the largest podcast publisher, according to Podtrac — is using AI to bring podcasts to a global audience by streaming them in new languages, accessible on iHeartRadio, Apple Podcasts and other platforms.
Translated episodes began rolling out in June and will be available in Spanish, French, Arabic, Portuguese, Hindi and Mandarin, with plans to expand to more shows and languages. Shows include titles like “Revisionist History with Malcolm Gladwell,” “Stuff They Don’t Want You to Know,” “Betrayal,” “The Girlfriends” and “Murder 101.”
Translations had begun into Spanish on 11 podcasts, listed on a dedicated web page.
iHeartMedia quoted Jay Shetty, the host of “On Purpose,” saying, “One of the questions I get asked most is, ‘When will the podcast be in Spanish? When will it be in Hindi?’”
The program uses SpeechLab, a platform that provides speech-to-speech translation. SpeechLab was created with incubation support from Andrew Ng’s venture studio AI Fund; it is led by entrepreneur Seamus McAteer as CEO.
The platform replicates original speakers or matches native voices; the company promises “nuanced, emotionally resonant vocals that seamlessly uphold your brand’s unique integrity.”
Disclaimers in each episode and in show descriptions let the listener know the show is translated using AI technology. The podcast providers or partners participate in profits from the translated versions.
Conal Byrne, CEO of iHeartMedia Digital Audio Group, said podcast listenership continues to rise, notably in regions such as Latin America, Europe, India and other parts of Asia. The company also said this is part of its goal to make podcasting more accessible and inclusive.
Describing the development process to the website Digiday, Will Pearson, the president of iHeartPodcasts, said SpeechLab cloned host voices from snippets of their shows, then refined the audio to translate it while preserving tone and personality. Clips were reviewed by podcast teams at iHeartMedia, who worked with the podcast creators and with human teams of native speakers to review the snippets.
“The podcast network would share feedback with the SpeechLab’s team,” Digiday reported. “For example, one host who has a slight accent had some words mistranslated. iHeartMedia also tested out the translations with groups of listeners who were native speakers.”
Pearson told Digiday the technology had advanced significantly in the last 12 to 18 months. “There’s a nuance of language … you want the conversational flow to feel right — more for the chat shows than the narrative style shows. When Jay Shetty delivers his content, he has his own delivery. We want to make sure it captures that essence as much as possible.”
[Read More Radio World Stories About Artificial Intelligence]
