The authors describe how the application of artificial intelligence technologies to ATC can improve the training of individuals who plan strategic actions, as well as making it possible to interface and cooperate with geographically distributed individuals who plan tactical actions. They also show how such training can affect concerns as diverse as safety, capacity, resource limitations, and traffic system efficiency simultaneously in a real-time environment. By focusing on the learning aspect of artificial intelligence, new dimensions are added to ATC simulation for training. The authors propose the construction of a simulation-based training device, called an intelligent adaptive trainer. This device would be used to teach controllers to reconcile conflicts and simultaneously to build operational confidence within a cooperative decision-making environment. An approach for training controllers to deal with the data discontinuities that frequently occur between levels of planning and operations is discussed. >