My 2 boys have CS careers, and I conversed with a few international programmers on my recent trip to Africa/Middle East/Europe. This is a subject we all discuss often. It seems the technology to create self-aware/self-improving robots already exists, now simply a matter of growth. As such, the "Terminator 2" movie seems frightfully accurate, in that to intercept said self-aware robots we'd not only need to destroy some current technologies, but almost salt some of the technological earth. An example might be "how far are we from creating a drone that can sniff out a rogue terrorist, and selectively kill him/her, or even to robotically improve upon this tech given the result" (the sequiter being that soon all nations might employ this relatively cheap technology). The resultant ironies might be "cleaner" wars, but also robots that kill all humans to protect robots (see "I, Robot"). I'm afraid that we may soon see a Luddite style banning of A.I. elements (and what would that be exactly, while creating an non compliant underground?) Unfortunately it may end up being like the atomic bomb, where we hope that "good" AI prevails over "bad" AI?