Development Team one concentrated on the replication problem. It's really hard to build something that can make copies of itself unless you peek at nature’s answers — violates Gödel’s theorem. What a miracle life is!
There were multiple approaches already under development. They would be ready in time. They wouldn't even try to make each macromolecules self replicating. It's possible with a synthetic DNA but it was sufficient, simpler, much safer, and more adaptable to make the system as a whole able to replicate constitute parts. This would disable an isolated macromolecule from endlessly copying itself without control – a major fear evoked by this technology. This choice of design depended on the communications and computing parts of the system.
Team Two worked on communications. There already were nano transceivers. Work involved making them secure and increasing their bandwidth a thousand fold. Again, military development already had solutions that were incorporated.
A consideration of nano-sized components is that quantum physics plays a part. This presented both problems and tools.
Team Three had the task of developing the computing components. Figuring the smartest solution possible was needed — one better than anyone on earth could come up with. The objective was a computer that would be to general problem solving what Watson was to trivia contests — a superior rival to the best human on the planet. It would not be a conventional processor but a multitude of communicating macromolecules. The macromolecules would be manufactured by other macromolecules. Its processor and memory would be holographic – meaning memories would be held in duplicate parallel connections that weren’t related according to location. Unlike conventional brains, animal or computer, connections between neurons were not physical – they were instead tuned channels of communications, billions of them.
Such low-level parallel computers had been proposed decades before but it never quite worked. It was such a paradigm shift. It needed a new approach. "It's like the voice recognition problem" explained Steven "For decades work went on under one researcher who insisted on his approach. Finally, he was replaced and other approaches were tried. Before you knew it, you were talking to machines on the phone. Now you can not only talk through your cell phone you can talk to it."
Scientists and engineers from around the world worked together. There was a headiness generated by the progress that was being made. Everyone or most at least, worked with an alacrity reserved to the few before the project. There were chinks in the armor of caution that epidemic fear of impending invasion made invisible. They created something which rational contemplation would have labeled supremely dangerous. Criticism was shouted down. The Center for Responsible Nanotechnology was ignored and became mute. Even the Defense Advanced Research Projects Agency, the DARPA responsible for the internet, was sidestepped for reasons of expediency although a couple of its people were hijacked for this project.
The teams created a thing that could grow and change as circumstances demanded. It could process data faster than the human brain – faster than any previous computer. It would be an expert in strategy and tactics. It could see and feel atoms. It was created to save the world.
Few of those working on the project felt the heart pounding pulse of anxiety that comes from holding life or death of a planet in your hands. They didn’t feel this level of anxiety because they didn’t know the implications of their work. They just did their job and were glad to have a job.
They designed their part, figured out how to construct their part and were to deliver their product to others who were figuring out how to combine the pieces. Even the completed pieces weren't completeness. That wouldn't happen until deployment. As long as the molecular manipulators and molecular brain components were kept fifty meters or more apart no spontaneous replication was possible. Humans were still in complete control and could terminate the project by simple inaction rather than by asserted action.
Designing and producing the nanodefense was actually the easy part — so much of it already existed. The hard part was building a monitoring and control system. There's no point building a defense if you can’t control it. Very dangerous technology needs very comprehensive control. It needs to be monitored and shut down or redirected if it wanders off the prescribed objective. This monitor system could listen to the nanos and, if need be, shut them down. The difficulty came in overcoming the extreme measures that were designed into the nanos to prevent communication takeover. There would have to be a lot of these systems because there were going to be a lot of nanos.