Challenge Problems for Artificial Intelligence

AI textbooks and papers often discuss the big questions, such as “how to reason with uncertainty”, “how to reason efficiently”, or “how to improve performance through learning .” It is more difficult, however, to find descriptions of concrete problems or challenges that are still ambitious and interesting, yet not so open-ended. The goal of this panel is to formulate a set of such challenge problems for the field. Each panelist was asked to formulate one or more challenges. The emphasis is on problems for which there is a good chance that they will be resolved within the next five to ten years. A good example of the potential benefit of a concrete AI challenge problem is the recent success of Deep Blue. Deep Blue is the result of a research effort focused on a single problem: develop a program to defeat the world chess champion. Although Deep Blue has not yet quite achieved this goal, it played a remarkably strong game against Kasparov in the recent ACM Chess Challenge Match. A key lesson we learn from Deep Blue’s strength is that efficient brute-force search can be much more effective than sophisticated, heuristically guided search. In fact, brute-force was so successful that it led Kasparov to exclaim “I could feel I could smell a new kind of int.elligence across the table.” (Kasparov 1996) The experience with Deep Blue shows that a good challenge problem can focus research, lead to concrete