Several people have posted comments on our blog asserting that we can't hope to maintain restrictions on advanced nanotechnology, so we should simply plan to defend ourselves.
It may be useful to contrast this position with current efforts to prevent terrorist access to nuclear bombs. "Old weapons, new terror worries," an article in the Christian Science Monitor, discusses the variety of ways that terrorists may try to gain access to nuclear weapons. Possibilities include hacking into the missile launch or warning systems, or attacking in force to seize weapons. Apparently, there used to be a serious "backdoor" in the American system for broadcasting launch orders to Trident submarines. And Chechens have been caught four times scoping out super-secret nuclear sites in Russia.
So should we just give up on nuclear security, and plan to defend ourselves from terrorist-planted nuclear weapons? Of course not. We'll have to be careful not to give up civil rights and liberties in the process of stopping the terrorism. And we should be more creative and proactive in mitigating the conditions that spawn terrorism in the first place. But it would be unthinkable to leave nuclear weapons unguarded and hope that we could figure out a way to defend ourselves from them.
Keeping nanofactories and nano-built weapons out of the hands of terrorists will raise even more difficult questions than restricting nuclear materials and bombs. But it'll also be at least as necessary.
Let's think about defense for a minute. Without massive lifestyle changes, a civilian population will be completely vulnerable to even relatively crude nano-built weapons. But will we be willing to make those changes? How many people don't use sunscreen, or don't install firewalls on their PCs? How can we defend people who are unwilling to take sufficient action to defend themselves, but insist that they be kept safe? This begs for a central concentration of administrative and forceful power. Oops--we don't like that either. What alternatives are there? Distributed engineering? The Internet is a paragon of distributed engineering, with far more "good guys" than "bad guys." But spam and worms are still massive problems.
We don't claim to have the answers. But we do claim that a lot of simple solutions are plainly unworkable. One of those solutions is giving the unrestricted technology to everyone. It's not even stable: the first few successful attacks will inspire really oppressive policy to try to stuff the genie back in the bottle.