Autonomous robots learning and social integration on the basis of modern cognitive technologies

Abstract

The emergence of intelligent autonomous robots as a result of the convergence of scientific and technical knowledge from biology, computer science, cognitive science, physics, materials science and dozens of other sciences, including social and humanitarian knowledge, was a immediate reaction of humanity as a complex self-organizing system to the challenges of the external environment and hence the rising need for continuous development and maintenance of its own viability. The rapid penetration of artificial intelligence technologies in the life of society, causes the need for understanding of scientific methods related to increasing the autonomy and intelligence of robotic systems. The article deals with the problems of the organization of appropriate behavior of robots. The basic requirements to ensure their autonomy are listed. The methodological basis of machine learning is discussed in the problems of processing sensory information and the synthesis of appropriate control procedures. In particular, the problems of constructing models of thinking are described as well as promising neural network technologies, the use of which allows to implement in autonomous robots a number of mental functions previously inherent only to man. Special attention is paid to the epistemological aspects of social integration of autonomous robots, which is important both from the standpoint of improving the efficiency of their training, and from the point of conflict-free coexistence of man and technology. The importance of social integration of autonomous robots in the context of noosphere evolutionary development is emphasized. The evaluation is given for development prospects of intelligent robots and their possible social effect in various spheres of social life.