Eliezer S. Yudkowsky wrote:
> If I really wanted to be speculative, with respect to "nanofusion", I
> would ask why you couldn't focus nuclear reactions so as to catch
> individual neutrons, or perhaps arrange materials so that a solid
> barrier of neutron-catchers existed instead of a haphazard crystal. I'd
> expect nuclear reactions to be considerably more efficient once we start
> working close to that scale.
In particular, there are three broad classes of possible controlled thermonuclear fusion: magnetic confinement (tokamaks and the like), inertial confinement (laser or beam implosion of pellets), and beam fusion. Beam fusion is the best match for small-scale systems, since the other two systems require a full-surround capture system to extract the energy. The smallest beam reactor that achieves practical break-even might be the size and shape of pencil. I thought we had discussed all this about three years ago, either here on in sci.nanotech. The basic idea is to use nanotech fabrication techniques to radically increase the efficiency of the beam generator and the energy capture systems.
It is this sort of reactor that is the basis of a viable small autonomous nanotech system.