From the Programmer's Apprentice to Human-Robot Interaction: Thirty Years of Research on Human-Computer Collaboration

We summarize the continuous thread of research we have conducted over the past thirty years on human-computer collaboration. This research reflects many of the themes and issues in operation in the greater field of AI over this period, such as knowledge representation and reasoning, planning and intent recognition, learning, and the interplay of human theory and computer engineering. Human-Computer Collaboration Figure 1 illustrates our overall research methodology, which has been to model human-computer collaboration on what is known about human-human collaboration. Furthermore we have focused almost exclusively on the special case of two copresent collaborators, i.e., where each collaborator is able both to communicate with and observe the actions of the other. Examples of such collaborations include two mechanics working on a car engine together or two computer users working on a spreadsheet together. To a first approximation, our approach has been simply to substitute a computer agent for one of the human collaborators, keeping as much else the same as possible. Due to space limitations, we will not attempt to review all research on human-computer collaboration, but limit ourselves to viewing this topic through the lens of our own work and that of our immediate collaborators. Consistent with this, note that bibliography below contains only publications by ourselves and our immediate collaborators. Chronological Summary The chronology of our research begins in 1976 with the publication of Rich and Shrobe’s joint M.S. thesis on the Programmer’s Apprentice [1,3]: “As compared to automatic programming research, the programmer’s apprentice emphasizes a cooperative relationship between the computer and the human programmer...” Shortly thereafter, Sidner began work on modeling how natural language is used in the context of pairs (and later groups) of people achieving tasks together. Her first paper on this topic dealt with the interpretation of discourse purposes in the Personal Assistant Language Understanding Program [2]. Under the direction of Rich and Shrobe, and later Waters, the Programmer’s Apprentice project [4,15,16] lived at the MIT AI Lab from 1976 until Rich and Waters left MIT in Copyright c © 2006, American Association for Artificial Intelligence (www.aaai.org). All rights reserved. focus of attention SharedPlans mutual beliefs