YR write to me: This reminds me of an FPGA that was programed
via evolution and not design. They wanted to program an FPGA to
reconize a certain frequency. They started with random designs
and used a computer to select the better designs and breed them
with other better designs. A type of natural selection. It took
thousands of iterations for it to get anywhere, but the end
result was really weird. Out of the 100 gates avaliable, only
over 30 were used (can't remember the exact amount). However the
gates were not useing binary, but more this sort of metastate.
Even weirder was the fact that 8 gates were being used, but
inderpendantly of the rest of the circuit. If they turned these
gates off the circuit would stop working, even though it seemed
like they were doing nothing. They concluded that these gates
created the exact interfeance needed for the rest of the circuit
to work. Not surprising was the fact that this was not able to
be transplanted into another FPGA. The design used the unique
instabilities and errors of that one FPGA.
