A Model of Human Activity Automatization as a Basis of Artificial Intelligence Systems

摘要

In this paper, a human activity automatization phenomenon is analyzed as a process as a result of which a cognitive structure is replaced by the equivalent reflexive structure. Such replacement plays an essential role as a mechanism that optimizes human mental processes according to their energetic and time consuming aspects. The main goal of the studies described in this paper is working out the algorithm that enables us to create the analogous mechanism in artificial intelligence (AI) systems. The solution would enable us to use in real time systems such AI systems, that, so far, could not have been used due to their high time consumption. The information metabolism theory (IMT) is the basis for the analysis. A cybernetic model of automatization, based on IMT, is introduced. There have been specified conditions according to which such solution is profitable. An automatization-type mechanism has been applied to IP traffic scanner and to a mutiagent system. As a result, time and memory properties of the systems have been improved significantly.