Massively Parallel Multi-Computer Hardware=Software Structures for Learning

This paper suggests how appropriately structured hardware/software multi-computers might be built and used to explore ways that intelligent systems can evolve, learn and grow. It examines what computers are, the great variety of topologies that can be used to join large numbers of computers together into massively parallel multi-computer networks, and the great sizes that today’s and tomorrow’s micro-electronic VLSI (“very large scale integration”) technologies make feasible. Then it describes several multi-computer structures that appear to be especially appropriate as the substrate for systems that evolve, learn and grow, and begins to sketch out a system of this sort.