William J. Clancey is a computer scientist whose research relates cognitive and social science in the study of work practices and the design of agent systems. At NASA Ames Research Center, he was Chief Scientist of Human-Centered Computing, Intelligent Systems Division (1998-2013); his team automated file management between Johnson Space Center Mission Control and the International Space Station. His studies relating people and technology include numerous field science expeditions from the Canadian High Arctic to Belize and Polynesia. He is Senior Research Scientist at the Florida Institute for Human and Machine Cognition in Pensacola.
Short abstract
Using AI technology appropriately, improving it, and predicting its implications for society requires a scientific understanding of the relation between people and computer systems. When we anthropomorphize automated systems—describing them as exploring, judging, collaborating, etc.—we are claiming from the start what we have set out to do, as if we already have “expert systems” and “smart phones.” Relating “symbolic” and “deep learning” programs to scientific and engineering models enables us to appraise what we have accomplished and to develop “explainable AI” systems—tools that fit and complement how people think and work and remain under our control. Using examples from NASA applications, I show how modeling and simulating people’s practices enables a scientific design methodology that facilitates explanation and creates reusable frameworks.