Educational technology and artificial intelligence (AI) are natural partners in the development of environments to support human learning. Designing systems with the characteristics of a rich learning environment is the long term goal of research in intelligent tutoring systems (ITS). Building these characteristics into a system is extremely difficult: each requires the use of techniques from AI, including knowledge representation, diagnosis and user modeling, planning, machine learning, and natural language understanding. Artificial intelligence techniques are usable now in practical systems. To illustrate this, several working systems that use artificial intelligence and that have been developed in the ARIES Laboratory (University of Saskatchewan, Canada) are discussed. The SCENT advisor can be used to provide robust diagnosis in a wide variety of problem solving domains. The learning by teaching system inverts the usual instructional paradigm: the system acts as an inquisitive learner, thus stimulating the human learner to refine and extend his/her knowledge. G.E.N.I.U.S. takes advantage of the credibility invested in a programming advisor by human learners in order to provide "ignorance-based" advice on programming errors. Finally, the VCR Tutor provides help to learners on how to program a video cassette recorder. The general lesson is that AI and educational technology can interact in a natural synergy to the mutual benefit of both. (Contains 11 references.) (MAS)