Jaron Lanier tries to get his arms around the issue of software scalability.

The reason we're stuck on temporal protocols is probably that information systems do meet our expectations when they are small. They only start to degrade as they grow.

...My engineering concern is to try to think about how to build large systems out of modules that don't suffer as terribly from protocol breakdown as existing designs do. The goal is to have all of the components in the system connect to each other by recognizing and interpreting each other as patterns rather than as followers of a protocol that is vulnerable to catastrophic failures.

This is certainly one of the most important issues in computing and software. As Frederick Brooks showed in his classic The Mythical Man-Month, the diminishing returns in software development are extreme. As you increase the number of people involved in a project, the productivity of each individual plummets.

Anyway, enough ranting. Go read Lanier's essay. It's important stuff. Also, scroll down when you get there, to Charles Simonyi's reply.