>> Damien B.'s main point seems to be that advanced nanotech would
>> require some damn complex software which embodies or otherwise has
>> access to quite a bit of knowledge, including how things aught to be
>> designed, and how those designs can be assembled. John Clark's
>> response seems to be that it will nonetheless still happen someday.
>> If so, I'm not sure there is much disagreement here.
>
>But wait..New thought just hit. You can use fractal and
self-replicating
>mathematics and the like to program nanobots, could you not. Would
that
>increase the efficiency substantially? Or has this already been
discussed.
Yes. The system (the Overtool) described at
http://www.speakeasy.org/~forrestb
is a fractal-like assembler. Assembly algorithms developed (probably
via GA) at one scale (or simulated) can be applied at all scales.
Forrest Bishop