[extropy-chat] self replicating machine ....

[Adrian Tymes]

--- Dan Clemmensen <dgc at cox.net> wrote:
> For this exersize, we must consider the paper (or other substrate)
> and 
> the rather exotic inks
> as raw materials. At the system level, a small production plant for
> the 
> paper and a small production
> plant for each of the inks is likely to be a whole lot easier to
> build 
> than even the smallest conceivable
> infrastructure based on silicon wafers.

Or possibly thin sheets of rock instead of paper, depending on the
intended environment; the printer would be modified accordingly.

> Can someone point me to
> references?

There is of course von Neumann's work, especially the book "Theory of
Self-Reproducing Automata" (a posthumous collection of his unpublished
work), which you can find at http://www.walenz.org/vonNeumann/ .  Or
check the bottom (majority) of http://www.cs.bgu.ac.il/~sipper/selfrep/
for a starting point.

One problem that systems like this keep running into is the need to,
within their own information space, both encode their own complete
information space and instructions as to how to replicate it.  I think
one could get around this by coding up instructions for subcomponents,
such as transistors or entire gates, and then having the "main" program
just refer to these subcomponents by index.  (Higher level encoding
might be needed; e.g., to write the ROM that encodes a single step of
instructions would need a single step of instructions.)

Perhaps one could write, as the main focus of the paper, an example
self-replicating circuit.  Assuming the existence of an ink jet printer
with which to write it, create a circuit with five outputs (plus x
motion, minus x motion, plus y motion, minus y motion, and deposit ink)
and one input (power, with crude clock signals being provided by
on-chip flip-flops), have it trace through the motions necessary to
deposit itself.  Again, though, the encoding would be the main
challenge here.