Apple Execs Explain How Apple Podcasts’ Transcription Feature Was Created

Back in March, Apple released a new automatic transcription feature for the Apple Podcasts app. Transcripts on the Apple Podcasts app allow users to read the complete text of any new podcast episode. The feature not only improves accessibility for the hearing impaired, but also makes it easier for listeners to catch up on missed details, search for specific information, or follow along with the audio.

In a recently published interview with The Guardian‘s Ari Saperstein, Apple’s global head of podcasts Ben Cave and manager of accessibility policy Sarah Herrlinger explained how Apple Podcasts’ automatic transcript generation feature was created.

“Our goal is obviously to make podcasts more accessible, more immersive,” said Cave. Roughly 15% of adults (approximately 37.5 million people) in the U.S. report difficulty hearing without an aid, with many of these individuals relying on captions and transcripts to follow music, movies, and podcasts. According to Cave, automatic transcript generation was one of the most requested features from creators.

Cave noted that the eventual birth of transcript generation could be traced back to indexing, another feature implemented on the Apple Podcasts app back in 2018. “What we did then is we offered a single line of the transcript to give users context on a result when they’re searching for something in particular,” the Apple exec said.

“There’s a few different things that we did in the intervening seven years, which all came together into this [transcript] feature.”

Herrlinger told The Guardian that Apple worked with both disabled employees and outside organizations to develop the transcription tool. She said that automatic transcription became a priority for Apple Podcasts due to increasing demand from both disabled users and podcast creators.

“In this case, we took the best of what we learned from reading in Apple Books and lyrics in Apple Music,” Herrlinger continued. Apple Podcasts’ transcriptions combine the time-synced, word-by-word highlighting feature for lyrics from Apple Music with Apple Books’ font and contrasting colour scheme to improve readability for the visually impaired.

Apple was pretty late in introducing a transcription feature, but the company hopes to make up for that by supporting a wider catalogue of podcasts. Unlike podcast transcription offerings from rivals like Amazon Music, Apple’s service extends to all the shows it hosts.

“We wanted to do it for all the shows, so it’s not just for like a narrow slice of the catalogue,” said Cave. “It wouldn’t be appropriate for us to put an arbitrary limit on the amount of shows who get it … We think that’s important from an accessibility standpoint because we want to give people the expectation that transcripts are available for everything that they want to interact with.”

Apple Podcasts transcribes every new episode that’s uploaded. Cave told The Guardian that while Apple is currently prioritizing transcripts for new content, the company plans to transcribe its entire back catalogue of shows eventually. “Over time, the entire episode library will be transcribed.”

Both users and disability activists have been impressed with the accuracy and availability of Apple Podcasts’ transcriptions, praising the tech giant for committing to launching a polished product rather than rushing through with a half-baked one.

Apple Podcasts currently supports transcriptions for English, Spanish, French and German-language podcasts, with plans to expand to more languages. While the auto-transcription does struggle with some harder names and complicated jargon, Apple acknowledged that better name recognition is already in the works.