Algorithmic basis for the functioning of private autonomous combat neural networks: algebraization of language forms

Abstract

It is shown that the modern development of weapons and military equipment will trigger the issue of improving artificial intelligence systems, in particular, their gradual convergence with human intelligence. It is shown that this requires the discovery of the algorithmic basis of artificial intelligence systems, including on the basis of the transition to multivalued logics. A close relationship has been established between the algebraization of natural language and the creation of the language of combat neural networks distributed systems in which processing is carried out based on the same operations that are involved in the human brain. The basis for this is the natural definition of language as a transpersonal structure, which largely determines the nature of human intelligence. A universal means of aggregating police multi-valued logics is provided, many elements can be placed at the disposal of General Galois. Thus, an analogue of the Zhegalkin polynomial is used, which is valid for achieving multi-valued logic with the use of properties (the Zhegalkin polynomial itself applies only to the case of binary logic). A selective case study that proves the constructiveness of the proposed application, a discoverable solution that tests the operation of a ternary adder using a Galois field containing three elements.