Ray Kurzweil discusses some of this in his book, The Singularity. It was at some point during 2006, while reading it, that I recognized that the technologies he was describing, if they came to pass, would entirely obviate nearly all of the "arguments from effect" for government. One of the biggest dangers he foresaw, however, was the potential for the development of "terrorist" nanobots that would essentially "infect" you and could, given the right radio signal, reduce you to a pile of goo.
Worse, said bots might go entirely out of control. Kurzweil's answer? The government will have to become even MORE powerful to defeat this sort of threat.
At the time, my thoughts were: how can you expect a centralized defense to work against a diffuse attack?
Now, however, I don't see how any government would be able to defend itself against such nanobots. It certainly couldn't develop such defenses on its own. The only thing the government will be able to do is to move in after it's already under production and make use of it.
The problem of "violence is immoral" is categorically not answered by technology. If our present society were to obtain such technologies, it's very possible that Kurzweil's worst dreams will come catastrophically true.