This talk describes the modeling of design and innovation through situated computing and computational social science. Computer models generally assume direct causality of behavior from decisions. Where causal theories are not yet known we make use of emergent, phenomenological behavior. Design and innovation are areas where there are no causal models. This talk introduces the principles of situated computing and describes how they form the foundation for a new class models in computational social science. The fundamental ideas of situated computing are:

• situations
• constructive memory, and
• interaction.

The talk describes the development of social, situated, cognitive agents in multi-agent systems based on these foundational ideas. It presents examples of implementations that range from studying doors, through the social production of art and curating an art exhibition to studying innovation.

John Gero Ph.D.
Krasnow Institute for Advanced Study, George Mason University University