Predicting the future is a dangerous business.
I’m reading Emergence, by Steven Johnson, which was published in the heady and optimistic era of 2001. Much of it holds up. But then you come to a passage like the following, which describes what TV watching will be like in the magical, futuristic world of 2005:

The entertainment world will self-organize into clusters of shared interest, created by software that tracks usage patterns and collates consumer ratings. These clusters will be the television networks and the record labels of the twenty-first century. The HBOs and Interscopes will continue to make entertainment products and profit from them, but when consumers tune in to the 2005 equivalent of The Sopranos, they won’t be tuning in to HBO to see what’s on. They’ll be tuning in to the “Mafia stories” cluster,” or the “suburban drama” cluster, or even “James Gandolfini fan club” cluster. All these groups–and countless others–will point back to The Sopranos episode, and HBO will profit from creating as large an audience as possible. But the prominence of HBO itself will diminish: the network that actually serves up the content will become increasingly like the production companies that create the shows–a behind-the-scenes entity, familiar enough to media insiders, but not a recognized consumer brand.

This is why you should never attach dates to your predictions. Admittedly, 2005 still has nine months to go, but somehow, I don’t see Johnson’s vision coming to pass between now and the end of December.
Still, I do see a few scattered seeds that might eventually grow into the kind of thing he’s talking about. The first is podcasting, which has been around for less than a year but has exploded into an alternate radio universe. Somewhere in Colorado, Brian Ibbot records an episode of Coverville, and the next morning, iPod-carrying commutors are listening to it on their way to work. And Ibbot is actually licensing the music he plays, which means that–like most podcasts–it’s fully legal.
Meanwhile, so many people are downloading TV shows (and other media files) via Bit Torrent that it may account for one-third of all Internet traffic. And it’s now becoming possible to set your computer to automatically check for episodes of your new TV show and download them overnight. At the moment this requires some tech savvy, but it’s only a matter of time until somebody comes out with the video equivalent of iPodder–a single program to manage your subscriptions. You couldn’t exactly call it video podcasting, of course, since people would be downloading it to their hard drives, rather than to iPods. “Drivecasting” is an obvious, if unimaginative choice. (I’d suggest calling it “baudcasting”, but it seems that name is already in limited use to mean something different.)
Whatever you want to call it, most of the TV content that’s available this way is illegally distributed network shows. There is, however, a small group of people making TV shows whose primary distribution method is via the Internet: Channel 101 and, recently, Channel 102. By network TV standards, the audiences are tiny. The Bu-arguably Channel 101’s best currently running show–gets an average of 12,500 downloads per episode. Still, that’s pretty impressive for a series of 5-minute episodes put together on a shoestring budget by a cast of (mostly) unknowns. Imagine it linked to an easy-to-use drivecasting program, and Johnson’s vision of user-defined networks doesn’t seem all that far off.
Just don’t expect me to put a date on it.