Технологии искусственного интеллекта: классификация, ограничения, перспективы и угрозы

Abstract

Introduction. Solving the problems of increasing labor productivity, creating transparency of key business processes, and creating new production facilities require not only new production technologies, but the organization of information processes, such as collection, storage, processing, analysis, and issuing responses to requests for information accompanying production processes. One of the most promising tools for solving these problems is AI technologies. Theoretical analysis. A classification of artificial intelligence technologies is presented, the following areas are highlighted: machine learning, natural language processing, computer vision, expert systems, advanced planning, speech recognition, robotics. The characteristics of the main technologies within the selected classification are given. To describe the limitations of the artificial intelligence development, the following levels of development limitation were used: physical implementation, safety of use, interaction with the environment, level of recognition, and the possibility of self-actualization. Prospects and risks are structured as sets of similar levels, complemented by a level of energy analysis. Conclusion. It has been revealed that to ensure the development of Society 5.0, it is necessary to create innovative platforms and mega-regional clusters for cooperation between authorities, entrepreneurs, and research centers in the field of artificial intelligence technologies.