Tuesday, February 15, 2011

Moore's Law and Structural Deepening

Apropos Tyler Cowen's book The Great Stagnation: How America Ate All The Low-Hanging Fruit of Modern History,Got Sick, and Will (Eventually) Feel Better.

Matthew Yglesias points to Kevin Drum pointing out that:

World-changing inventions just don’t come around all that often, and when they do it takes a long and variable time for them to become integrated enough and advanced enough to have an explosive economic effect. Steam took the better part of a century, electrification took about half that, and computers — well, we don’t really know yet. So far it’s been about 60 years and obviously computers have had a huge impact on the world. But I suspect that even if you put the potential of AI to one side, we’re barely halfway into the computer revolution yet. To a surprisingly large extent, we’re still using computers to automate stuff we’ve always done instead of actually building the world around what computers can do.

This reminds me of W. Brian Arthur's book The Nature of Technology and Arthur's idea of structural deepening (as described here).

Yglesias has his own theory that advances in computer technology have in fact been "un-optimally" rapid. We have come to expect that new and improved computer technologies will arrive so soon that it isn't really worth refining an existing paradigm. Therefore structural deepening hasn't really happened in Arthur's sense of the term. There has instead simply been a succession of straightforwardly better technologies.