Someone emailed me (Chris) the following question, and I found myself writing a fairly long answer that seems worth posting here.
You have said that molecular manufacturing will be extremely powerful, but very few people know what that really means. What does it mean? What kind of an impact will molecular manufacturing have on politics, economics, law, sociology and the environment?
By "extremely powerful," we mean that molecular manufacturing systems will be able to manufacture a wide range of products direct from blueprints in an hour or so. The active components of the products (motors, sensors, and computers) will be, as a rough rule of thumb, a million times more compact. Manufacturing systems will be able to make more manufacturing systems as easily as any other product, so manufacturing projects can be as big and fast as desired.
The effects on politics will be numerous. Powerful, rapidly developed weapons systems will change the military balance of power. International trade may be reduced or shifted from goods to information. Networked sensors will allow governments (and other groups) to spy on their own or each other's people. Et cetera. This area will probably produce some extreme risks.
We don't really know the effect on economics when both labor and existing manufacturing capital rapidly lose most of their value, when manufacturing capacity is no longer scarce, and when goods can be produced at a cost that is several orders of magnitude less than their value. It seems likely that manufacturers and business owners will try to keep prices up, perhaps through regulation; however, this would stimulate a massive black market. This area needs more study.
Law will still be necessary. New laws and new types of law enforcement may be needed for new products.
Global society probably will be more affected than it was by the automobile, the birth control pill, and the radio, television, and computer. I'm not sure if anyone can predict the forms those effects will take.
The environment could be saved or destroyed; we'll have a lot more of a choice than we do today, and a lot more information on which to base our choices. The phrase "planet-scale engineering" is not an exaggeration. Today we do planet-scale engineering in decades, except it's not really planned; we will be able to do it deliberately, in months.
I'll throw in a bonus question: When will this technology be developed? I recently asked several other researchers for their opinion. The consensus is that the basic technology will probably come into existence sometime between 2010 and 2020. How quickly it's adopted and who will get to use it are as yet unknown.