This paper begins by examining concepts of artificial intelligence (AI) and discusses various definitions of the concept that have been suggested in the literature. The nesting relationship of expert systems within the broader framework of AI is described, and expert systems are characterized as knowledge-based systems (KBS) which attempt to solve problems in the manner that humans do. Problem solving is viewed from the perspective of the library user with emphasis placed on nontraditional approaches to knowledge seeking. Sociological, psychological, and cybernetic systems perspectives are then employed in order to see these information technology developments from a library user's perspective. Finally, KBSs are focused on and criteria for their establishment in libraries are listed. The natural language interface is seen as significant, and the pragmatic features of future library operations such as automated cataloging and the use of expert systems in the answering of reference inquiries, and document and information retrieval are examined. The application of AI in library situations is presumed to offer significant benefits and opportunities. (Author/SD)