Abstract
AI has generally interpreted the organized nature of everyday activity in terms of plan-following. Nobody could doubt that people often make and follow plans. But the complexity, uncertainty, and immediacy of the real world require a central role for moment-to- moment improvisation. But before and beneath any planning ahead, one continually decides what to do now. Investigation of the dynamics of everyday rou- tine activity reveals important regularities in the in- teraction of very simple machinery with its environ- ment. We have used our dynamic theories to design a program, called Pengi, that engages in complex, apparently planful activity without requiring explicit models of the world.
problem solving and reasoning with representations. We observe that real situations are characteristically complex, uncertain, and immediate. We have shown in [Chapman, 19851 that Planning is inherently combinatorially explo- sive, and so is unlikely to scale up to realistic situations which take thousands of propositions to represent. Most real situations cannot be completely represented; there isn’t time to collect all the necessary information. Real situations are rife with uncertainty; the actions of other agents and processes cannot be predicted. At best, this ex- ponentially increases the size of a Planner’s search space; often, it may lose the Planner completely. Life is fired at you point blank: when the rock you step on pivots un-
expectedly, you have only milliseconds to react. Proving theorems is out of the question.