Self-Replicating AI : Integrating New Neural Networks Through a Natural Selection Process

EasyChair Preprint 1981

Abstract

Self-replication is any behavior of a system that yields construction of an identical copy of itself. Biological cells, given suitable environments, reproduce by cell division. In this paper, we present a self-replicating neural network that integrates new neural networks through natural selection process. The network replicates itself by integrating other neural networks and by learning to output identical copies of its own components for which it is trained. Also we describe a method of an optimization technique using Genetic Algorithm ( GA ) for updating and optimizing the neural network weights. GA creates multiple solutions and evolves them through a number of generations, and each solution holds all weights in all layers to help achieve higher accuracy. The evolutionary algorithm ( i.e. GA ) was used as an optimization approach that mimics the concept of natural evolution for creating fitter individuals that have higher chance of survival through natural selection. The GA processes integrated with the ANN predictive model ( GA-ANN ) and the network replicates itself by learning to optimize its own weights. We observe from the test results that the networks were able to replicate through natural selection with good accuracy. It is observed that self-replication mechanism for artificial intelligence is convenient because it establishes the possibility of persistent improvement through natural selection.

Keyphrases: Artificial Intelligence, Genetic Algorithm, integrating neural networks, self-evolving, self-replicating

@booklet{EasyChair:1981,
  author    = {Poondru Prithvinath Reddy},
  title     = {Self-Replicating AI : Integrating New Neural Networks Through a Natural Selection Process},
  howpublished = {EasyChair Preprint 1981},
  year      = {EasyChair, 2019}}