ARTIFICIAL INTELLIGENCE ] Artificial Intelligence and Learning Environments: Preface William J. Clancey Institute for Research on Learning, 2550 Hanover Street, Palo Alto, CA 94304, USA Elliot Soloway Department of EECS, University of Michigan, Ann Arbor, MI 48109, USA The promise of computer-aided instruction (CAI) has always been individual- ized instruction: providing a learner with an environment that is tailored to his or her learning needs and goals. Although there have been notable successes (e.g., see Larkin et al. [9]), the architecture of CAI systems has been inadequate to provide robust and rich learning environments. Starting in the early 1970s, researchers applied an AI perspective to the problem of creating learning environments. The architecture that evolved during this period sugges- ted that an intelligent CAI system (ICAI) would need: (1) an explicit model of the domain and an expert program that can solve problems in the domain, (2) a model of the student that identifies, at a fine-grained level of detail, what the student understands, and (3) a tutoring model that can provide instruction to remediate misconceptions and/or present new material. Not surprisingly, during this period there was also considerable effort in exploring the psycho- logical questions underlying learning, teaching and understanding. Attempting to summarize and spotlight research in this area, Derek Sleeman and John Seely Brown in 1978 edited a special issue of the International Journal of Man-Machine Studies [10]. The title of the book, Intelligent Tutoring Systems, gave the field its most commonly used name, ITS. Rationale for the Collection Several years ago, we decided that it was again time to assemble a special journal issue that would bring to a general AI audience the advances, insights, and problems of this niche field. ITS research was quite active, with consider- Artificial Intelligence 42 (1990) 1-6 0004-3702/90/$3.50 © 1990, Elsevier Science Publishers B.V. (North-Holland)