The Inverse Network Effect

The always insightful Clay Shirky has just released Social Software and the Politics of Groups. The underlying core: we have no antecedents for how nonlocal groups, and the technology that supports them should function. In other words there is almost no precedent for the way a group of people across the world can interact in a mailing list or chat room. And its equally difficult to design software for those interactions, as there is no precedent for thinking about the problems.

What sprang to mind immediately was Metcalfe's Law which basically states that the usefulness of a network increase exponentially with the number of users. Perhaps there is a reverse law as well. That the larger the network the difficulty in designing new software increase exponentially. How do you test software designed to let 1000 people communicate at once? Its hard enough deal with the tech issues, working with the interface and social use models is a whole other story.

The classic example being that 1 person with a telephone is useless as they have no one to call. 2 people with telephones is useful assuming they want to talk to each other. The more people