Podcasting is going to save the world … if only we could find something to listen to!

Tyler Cowen writes on Marginal Revolution:

The key question is what kind of aggregators will take off … The relative returns to “portal podcasts” will be lower than for portal blogs. Glenn Reynolds can read and process material faster than most people, but no one can hear a two-minute comedy routine in much less than two minutes … So you won’t find good podcasts through other podcasts to the same degree, since it is harder to serve as an effective portal. The sorting will work less well, and the categories will be harder to describe and communicate. Advertising will matter more, and institutions such as iTunes will have more influence over selection and content. Podcasting will be more in hock to MSM than are blogs.

The two most prominent podcast portals are Odeo and iTunes. Unfortunately they’ve got it wrong. Both iTunes and Odea organise podcasts into categories, but as Tyler points out the categories aren’t that useful. For example, let’s say I want to listen to some new music. Right now there are only 692 channels under the Odea music tag. That is a few too many to listen to. If I try to be more specific and go to the Indie tag there are 92 channels. This is better but still not great; the term indie covers everything from whiny boys with guitars to stuff I might be interested in hearing. So perhaps Tyler is right and podcasts are doomed.

Or perhaps not. As I posted earler, implicit feedback is much better than explicit categorisation. If you want to see how great implicit feedback can be in the music arena, look no further than last.fm. If I like, say, Mouse on Mars then last.fm correctly guesses I like Plaid, Autechre, and Aphex Twin. All is takes for last.fm to work is a little plugin that tells the site what music you’re listening to. Could the same technology work with podcasts. Of course! Problem solved.