The question podcasting asks
It's great that podcasting has grown so much. But the original question it asked still hasn't been answered. Until that happens the progress we've made won't be safe from silos.
by Dave Winer Monday, April 24, 2017

The next a continuing series of stories about podcasting.

If an angel tapped me on the shoulder and asked what is the one thing podcasting needs more than anything, I would have an answer. 

What podcasting needs more than anything is a quick easy answer to the following question --

Find me something good to listen to now.

  1. It's personal. Note I didn't say find anyone a good podcast to listen to now, I said find me one. It should be a podcast that I, an individual person, will find good to listen to. 
  2. It's immediate. I don't want to subscribe to a podcast that will someday yield something good to listen to, I want it now.

For the last 17 years podcasting has been answering the wrong question: I want to subscribe to this. That question doesn't come up often, if ever. And podcasting doesn't even answer it very well.

Think about it. Where are you when you decide you want to subscribe to a podcast. That's a trick question. How about nowhere. How about it never happens. I'm never looking at something on the web or elsewhere and think oh that's something I want to subscribe to. 

There are shows that I like so much that it's almost certain that I want to listen to whatever they have available right now. Planet Money, Fresh Air, West Wing Weekly, Radio Open Source are examples. But it's never guaranteed that their latest thing is something I want to listen to now. It might be a rerun. Or maybe I find the subject distasteful or disturbing or the person they're talking with or about is of no interest. 

Okay so if the question is Find me something good to listen to now, then how do you do that? What information do you need to be able to get that done? 

It's collaborative. You need to know what I've listened to that I liked. And you need to know what my friends have listened to that they liked. So we can serve as recommenders for each other.

Facebook's social graph is good enough to get this started. I know because I tried it manually a couple of years ago. I posted a message asking my friends to recommend podcasts. I then made that into a subscription list and fed it into River5 and the result is Podcatch.com. It's highly personalized. It works well enough that I get an answer to FMSGTOLTN pretty much 100 percent of the time. 

To do this for other people we just need to systematize it. Make it so that it works for anyone's friends, and it's dynamic, it's kept up to date. When my friends' interests change, then the recommendations change. 

Also to really work, Facebook isn't the right place to accumulate the data because amazingly you can't listen to a podcast in Facebook. They're too busy reinventing human senses I guess.

 A podcast listening client would be able to accumulate the data. But here's the catch, I don't want to be locked into any one client. So if they aren't giving the user access to their own listening data, and they aren't making it easy to share with others, then I'm not going to use it. It's possible that such clients exist, and if they do I want to know about them. 

How important is this? Very. A similar problem was there for RSS in the early days and it was never solved because the reader vendors were unwilling to accept leadership. They all assumed they were going to dominate, or their investors wouldn't let them share the data. Then along came a silo that didn't use RSS, Twitter, and it solved the problem. RSS languished with no love from the dominant vendor, who eventually left (Google Reader). Today RSS still works, amazingly, but it's not a happy ecosystem. It serves as a cautionary tale for podcasting, in how it can go very wrong. 

I'd say this is the big challenge for podcasting in 2017. It's why I'm worried about it getting siloized. The practical problem is that then there will be a gatekeeper, who behaves like Facebook and controls an opaque algorithm and the medium is closed off to non-Facebook developers, and we lose the freedom that has made podcasting such a wonderful thing.