Skip to contents

WWDC Notes: Apple Podcast Desktop App, Machine Learning Search

Apple’s week-long WWDC developer conference kicked off yesterday, and there are a few podcast things you should be aware of. Well, two.

The first is the official end of iTunes, which is now going to be split up into separate desktop apps for Music, TV, and Podcasts on the new Catalina Mac operating system. This has been long rumored, speculated, and pre-documented, and hey, it’s actually happening. So long, iTunes. I’m gonna miss ya, buddy. (I recommend NPR’s write-up on the matter if you’re looking for a brief read on why this is happening.)

Nieman Lab’s Josh Benton wrote out loud what I was thinking when I first heard about this development in a post yesterday:

I don’t think that a native podcasts app on the Mac will do much to move the needle of listening; the rise of smartphones has meant that laptop time is now more focused on work than it used to be, and I don’t see much reason most podcast listening would move to the desk.

That said…there’s a reason that many people tie the podcast boom of recent years to Apple deciding to include a separate Podcasts app on in iOS 8 in 2014.

We shall see.

The second thing is what Apple says it will be introducing to the desktop Podcast app: machine learning-driven discoverability features. Here’s TechCrunch on the matter:

The Podcasts app for Mac offers a way to search, discover, subscribe and listen to your favorite audio programs, much as it does on iOS. Your listening data will be synced across devices, and you can listen directly in the new app, as well.

But it’s also got a new trick: it will now use machine learning technology to index the spoken words in podcasts. That will allow you to find more podcasts — or even individual episodes — that reflect your interests.

This, of course, does not come out of nowhere. In December 2017, I reported that Apple was acquiring Pop-Up Archive, an online platform focused on building tools to transcribe, organize, and search audio files. Among its suite of tools was a product called, a podcast search engine which played with similar ideas.

Took a while, but we’re here now. Playlists vs. Search Engines, let’s go.