John E. Laird

My major research interest is in creating human-level artificial intelligent entities, with an emphasis on the underlying cognitive architecture. A major challenge is to create systems that can work on a broad range of problems, using a wide variety of methods, knowledge, and learning techniques. As part of my research, I study both artificial and natural intelligence. Since 1981, my work has centered on the development and use of Soar, a general cognitive architecture. Over the years, this has led to research in both AI and cognitive science. Within AI my work has included research in general problem solving, the genesis of the weak methods, the origins of subgoals, general learning mechanisms, interacting with external environments, learning by experience and by instruction, and integrating reactivity, planning, and learning, all in the service of constructing complete autonomous intelligent agents. In the past, I’ve done some work on developing human-level AI agents for military simulations, interactive computer games, and most recently autonomous robotics. Within cognitive science, my early research has concentrated on detailed modeling of human behavior (reaction times and error rates) in visual attention, concept acquisition, and dual tasks. Currently I’m concentrating on high-level cognition. Most recently, my students and I have extended Soar to include reinforcement learning, episodic memory, semantic memory, mental imagery, and emotion-inspired processing. We are exploring interactive task learning by developing agents in Soar that learn new tasks from scratch using natural language. The tasks involve real-world robotic interaction and include hierarchical tasks, as well as many simple puzzles and games.