The BBC is using AI to generate subtitles, and it’s about time
The hugely popular BBC sounds audio platform is testing new AI-generated subtitles that play alongside shows, as well as full transcripts, via its website and app. The three-month trial is currently limited to a handful of shows: In Touch, Access All, Profile, Sporting Witness and Economics with Subtitles.
So now when you listen to these specific shows on the website, you’ll see a new subtitles button in the playback bar. Click that and the AI-generated subtitles will play along with the audio.
The BBC is also expanding the trial to the BBC Sounds app, Android firstand later on iOS. So in the app, you’ll see the same subtitle icon appear in the playback bar when you’re listening to these specific shows. It works a little differently, though. Subtitles scroll up on the screen, highlighting the spoken words as the host and guests say them.
The BBC is using a speech-to-text AI called Whisper AI for the trial, after which there will be an evaluation to determine how well it worked. If the BBC considers the project a success (hopefully) it will roll out the subtitles to more shows.
Why so long?
The odd thing is, many podcast apps embraced AI to add subtitles long ago. Spotify, for example, added automatic transcription to its podcasts as early as 2023, with a Read Along feature that displays subtitles as the podcast plays, highlighting each word as it’s said. Apple added transcriptions to Apple Podcasts earlier this year.
While the trial is underway, the BBC will have a human editor check each transcript before it is uploaded, rather than relying on artificial intelligence to produce error-free transcripts. The BBC’s reluctance to fully commit to AI-generated transcripts, preferring to conduct a trial, is illustrative of the caution with which humans generally approach AI.
Given the concerns many people have about what AI could mean for jobs in the future, it’s understandable that the BBC is being cautious. But the fact is that BBC Sounds produces around 27,000 hours of content every month. Much of this is difficult for the 18 million or so people in the UK who suffer from hearing loss or tinnitus to access. And that’s not to mention the significant global audience. It would be impossible to transcribe all that material manually.
The BBC is slowly being dragged into the future in its own way and is embracing AI, but we’re pleased to see the broadcaster making good use of this exciting new technology.