- Google wants AI to automatically transcribe podcast audio into searchable, time-stamped text files.
- Using those files, Google wants to enable listeners to scrub ahead in an episode, seeing what’s coming up as they scrub.
- Ultimately, Google wants to automatically translate podcasts into any given language using text transcriptions.
A few days ago, we told you about how Google is working right now to make podcasts more upfront in Google search results, and how the company sees the medium as having an important future. Now, via the conclusion of the five-part series Pacific Content just published, we have a better idea of what the future of podcasts will look like.
There are three advancements that Google would like to make with podcasts in the coming years. Granted, none of these are actually real technologies yet, but Google Podcasts Product Manager Zack Reneau-Wedeen, pictured above, says that Google is going to “try to make these experiences possible.”
Here’s what Google has in mind for the future of podcasts.
Reneau-Wedeen has a vision that every podcast episode that hits the internet gets turned into a text file. Using artificial intelligence, Google would “listen” to every episode and transcribe it into a time-stamped document that itself could then be searchable.
This would not only make it easy to find podcast episodes surrounding particular phrases and keywords, but it would also enable listeners to immediately jump to one section of a podcast that interests them.
You can do something similar on YouTube right now. If you want to hear a live rendition of a particular song in a full-length concert, you can link directly to that part of the video in the description. You can click here for an example. The difference with podcasts is that the time stamps would be automatically created.
If you could do that with podcasts, it would help not only podcast listeners but also general queries to Google. If you ask Google a question like, “What is Shaq’s favorite color,” it could link you to an audio clip of Shaq on a podcast giving that information directly.
Assuming that a text transcription of podcast episodes becomes a reality, another YouTube feature would then be possible. If you hover over the progress bar of a YouTube video, you will see little images that give you an idea of what’s coming up. This helps you find the content you need in any given video.
Google wants to do the same thing with podcasts, where you’d be able to scrub ahead on a progress bar and see, in text form, what will be discussed later on in the episode.
The ultimate goal with podcast transcription would be to make podcasts from all areas of the world understandable by all citizens.
Right now, a podcast in English doesn’t help you much if you don’t speak English. But Google could transcribe a podcast, translate the text into any given language, and then recreate that podcast in a language you understand. Using artificial intelligence, Google could theoretically deliver that podcast in the original voice tones of the podcasters.
That’s pretty creepy, but also really cool.
The entire five-part series on the future of podcasts is fascinating if you have time to read through it all. Either way, if you’re a podcast fan, it’s nice to know that Google is giving a sharp focus to the medium. That means more podcasts that are easier to get!