The Last Podcast

Greg Linden has a very interesting article about the current state and future of personalized news services up on IEEE Spectrum online.

He does a great job at looking at the technical issues and limitations behind personalized news services and, I think, makes a good argument for how these algorithms can be a very useful tool for online news sources.

While I find the technology behind this fascinating, I admit that I have a number of deep seated problems with it as well, even though personalized news would definitely be a great time-saver for me.

First of all, the joy in reading news for me is in discovering things I would have never thought of before. Just because I normally read the tech articles in the New York Times doesn’t mean I don’t browse other sections as well.

Basically, I guess I am just afraid to miss something really interesting, but new, just because an algorithm thinks I’m not interested in it. A personalized newspaper, I think, might quickly turn into a personalized echo chamber where my ideas are never challenged.

I am far more likely to trust a network of fellow web-addicts on a service like FriendFeed or the Elite Tech Reddit to share interesting stories they find.

Cartoon courtesy of Gapingvoid.com