System of artificial intelligence as a model of natural intelligence

Abstract

Artificial intelligence systems (AIS) are studied for adequacy and accuracy as models of natural human intelligence. The capabilities of AIS are studied from the standpoint of the theory of computation and logic. The study focuses on the capabilities of artificial neural networks, which are the core of AIS. It is shown that there exist unsolvable for AIS problems, unlike humans. In addition, AIS is unable to transform sets of facts into new knowledge like humans due to the insufficient development of inductive logic methods. The prospects for the development of AIs as models of natural intelligence are assessed depending on the development of super-tuning neural networks, as well as quantum computing.