Planning for and Control of Intelligent Systems

Abstract

The field of Artificial Intelligence (AI) has never before had to cope with the problems experienced in the day-to-day operation of high performance, large throughput information systems (IS). Such environments typically involve the retention and maintenance of large databases with consequent need for good integrity constraints on input and updating of data, plus the careful preservation of program and system integrity through versioning and other special control mechanisms. But the day when there are such production environments involving AI parts may not be far away, we face a possible marriage of database management systems with AI techniques. If this occurs, economic information systems may use AI tools in the production and operation phases of large econometric models and in the normal operation of large statistical data systems. Current practice and present methodologies in IS implementation and operation are therefore examined to suggest how these future systems may evolve from these disparate fields and thereby improve the overall system by their symbiosis.