CRN's Director of Research Chris Phoenix has been pondering, and he's come up with some thoughts about product complexity, self-assembly, nanotech fabrication, and culture clash...
Any manufacturing system can make products of a certain complexity. If the system is purely mechanical, then the product complexity will be limited by what's built into the system. But if the system is programmable, then complexity can be added to the product simply by downloading a new program.
A system using self-assembly must have all the complexity built in to the chemicals. It's hard to build really complex chemicals. Not impossible, but hard. We've barely learned to design one protein that folds the way we plan.
With DNA folding, the sequence of bases is programmable. So you can put a lot of complexity into the system at the start--but only at the start. (At least for now.)
With programmable systems, using actuators controlled by digital signals, the system can build things vastly more complex than its own mechanics.
Those who don't plan to build such complex nanosystems won't care much about this. They'll be happy with self-assembly, because it does everything they need. And if they're not familiar with programming, they won't realize that more is possible--so they'll resist as unworkable suggestions to try for this goal.
Those who are familiar with programming know how much easier software makes things. To them, it's obvious that programmability is a major goal, and that once it's achieved, they can far exceed non-programmable methods.
I think this is a very deep culture clash. It's taken me fifteen years to identify it. It implies that only programmers, or people willing to think like programmers, will comprehend the potential of molecular manufacturing.