A group of American and Japanese psychologists, anthropologists, linguists, and computer scientists gathered at the University of California, San Diego, to exchange ideas on models of joint problem solving and their special relevance to the design and implementation of computer-based systems of instruction. Much of the discussion focused on microcomputers as instruments for organizing instruction. This report provides summaries of each of the following presentations and the discussions that they generated: "Joint Problem Solving in Functional Writing Environments," James Levin and Margaret Riel; "Modeling Cognitive Strategies with a Talking Microcomputer," W. Patrick Dickson; "Goal-Formation between Users and Computers," Mary Riley; "Quasi-Understanding Induced by Verbal Instruction," Naomi Miyake; "Developing an Automated Tutor for Radar Navigation," Edwin Hutchins; "How to Teach Somebody Something They Don't Already Know," Denis Newman and Andrea Petitto; "Programming in the Classroom: Ideals and Reality," Jan Hawkins and Karen Sheingold; "A Computer Game Environment for the Study of Stress and Performance," Yoshiro Miyata; "Controlling the Relations between Rule Statements and Piagetian Problem Environments," Laura Martin; "Buggy and Beyond," Kurt VanLehn (concerns tools for building computational theories of cognition); "Prescribing Effective Problem Solving Procedures," Joan Heller; "Joint Solving of Physics Problems by College Students," Yutaka Sayeki; "Appropriating an Expert's Understanding," Peg Griffin; "The Nature of the 'Joint' in Joint Problem Solving," Ray McDermott; "Microcomputer Networks and Presuppositions about Modes of Communication," Ron Scollon; and "Collective Scientific Discovery by Day Care Children," Giyoo Hatano and Kayoko Inagaki. A summary of main issues and two references conclude the proceedings. (LMM)