Television is good for reaching a broad audience. A PBS show is viewed by 2-6 million people; a commercial network show by 10-25 million. Television brings characters and stories to life in a way that is difficult to do in other media. You can humanize characters and lift them out of the realm of statistics. You can use drama to generate emotions and motivate people to action.
Television is terrible at distributing information. The script for a half-hour national network news broadcast would fit into a fraction of the front page of The New York Times. Less than one newspaper page. If you're trying to use TV to distribute information, you're in trouble.
Television is terrible at localization. Most TV is produced for national broadcast. It is extremely expensive to localize content and in general you can't expect to have the footage, much less the editing resources. Even if you had localized content, it isn't clear how you'd distribute it to stations since distribution networks are literally "broadcast" technology.
Enter the Web.
A Web service backed by a relational database can localize an enormous database and make it relevant to anyone willing to type in his or her zip code. Example: www.scorecard.org. A Web service can offer tremendous detail to readers who are willing to read and be engaged.
Given a TV broadcast that will herd millions of users to a Web site and a Web service that localizes data and involves those users, it is natural to think about giving those users an opportunity to collaborate and communicate with each other. For example, along with a show on juvenile crime, it is natural to offer a service that lets people manage volunteers for after-school programs in their neighborhood.
Philip Greenspun has been building Web services since 1993.
I think the major technological constraint for this future of "video on demand" and "localization of the video media" is the mass availability of bandwidth. Just thought I would mention that.
-- Paul Konigsberg, February 23, 2000