Functional Emergence with Multiple von Neumann Computers.
Hideaki Suzuki · 1997
An evolutionary programming methodology in which multiple von Neumann computers evolve through genetic algorithms (GAs) is proposed and simulated. A bit sequence in the program memory owned by the computer is composed of numerous machine codes, each of which has some computational function in the CPU. In the selection operation of GAs, a single machine code is neutral or deleterious in fitness by itself, and an appropriate combination of machine codes (an appropriate subroutine) can achieve the advantageous function which promotes its reproduction in GAs. Through the evolutionary simulation using GAs, these subroutines are created and fixed in the population one by one, and the higher function, which enables the computer to answer the expected value from the environments, finally emerges in the program memory. Under different sets of parameters, the evolutionary speed, determined by the time until this emergence, is examined and the conditions for achieving faster ev...