Features of the problematic field of artificial intelligence. Systemic aspect

Abstract

The paper presents the author’s vision of the AI problematic field. Limited information on AI technologies complicates analytical research, but does not hinder system-theoretical research, especially at a high level of generalization. Digital transformation of public administration actualizes system-theoretical research in the field of AI. The purpose of the work is to identify the features of the AI problematic field that can become the basis for system-theoretical research and possible practical measures. To obtain epistemological emergence, a complex research object was built, which is the unity of three lines of knowledge: on the features of the development of systemic representations, AI technologies and management technologies. Systemic reflection allowed drawing the following conclusions: the accelerating extensive development of AI technologies in the absence of progressive changes in other areas is a hidden factor in the growing imbalance of the social system, requiring urgent elimination; there are no systemic studies of AI using a metasystemic approach that would allow for a holistic disclosure of the AI phenomenon and an explanation of destructive changes in AI systems during their self-learning (neural network hallucinations, AI going out of control, etc.); a new methodology of science and technology in the field of AI could be systemic intensification, capable of eliminating the trend towards extensive development and other shortcomings of AI. The main idea formulated by the author is that it is necessary to intensify a wider system – a metasystem, of which AI systems are a part, so that in such a metasystem, as can be expected, AI systems would acquire their new quality without our direct intervention, due to hidden systemic mechanisms. The key to practical systemic intensification of AI could be the intellectualization of decision-making processes in government bodies.