But University of Wisconsin-Madison research psychiatrist Giulio Tononi, who was recently selected to take part in the creation of a "cognitive computer," says the goal of building a computer as quick and flexible as a small mammalian brain is more daunting than it sounds. Tononi, professor of psychiatry at the UW-Madison School of Medicine and Public Health and an internationally known expert on consciousness, is part of a team of collaborators from top institutions who have been awarded a $4.9 million grant from the Defense Advanced Research Projects Agency (DARPA) for the first phase of DARPA's Systems of Neuromorphic Adaptive Plastic Scalable Electronics (SyNAPSE) project. Tononi and scientists from Columbia University and IBM will work on the "software" for the thinking computer, while nanotechnology and supercomputing experts from Cornell, Stanford and the University of California-Merced will create the "hardware." Dharmendra Modha of IBM is the principal investigator. "Every neuron in the brain knows that something has changed," Tononi explains.
This article reviews research in the development of plan generation systems. Our goal is to familiarize the reader with some of the important problems that have arisen in the design of planning systems and to discuss some of the many solutions that have been developed in the over 30 years of research in this area. In this article, we broadly cover the major ideas in the field of AI planning and show the direction in which some current research is going. We define some of the terms commonly used in the planning literature, describe some of the basic issues coming from the design of planning systems, and survey results in the area. Because such tasks are virtually never ending, and thus, any finite document must be incomplete, we provide references to connect each idea to the appropriate literature and allow readers access to the work most relevant to their own research or applications.
The need for presenting useful descriptions of problem solving activities has grown with the size and complexity of contemporary AI systems. Simply tracing and explaining the activities that led to a solution is no longer satisfactory. We describe a domain-independent approach for selectively abstracting the chronological history of problem solving activity (a system trace) based upon usersupplied abstraction goals. An important characteristic of our approach is that, given different abstraction goals, abstracted traces with significantly different emphases can be generated from the same original trace. Although we are not concerned here with the generation of an explanation from the abstracted trace, this approach is a useful step towards such an explanation facility.
One problem with taking advice arises when the advice is expressed in terms of data or actions unavailable to the advicetaker. For example. in the card game Hearts, the advice "don't lead a suit in which some opponent has no cards left" is nonoperational because players cannot see their opponents' cards.