Play is a central part of human life. People play throughout their lifespan, and play is involved in human socializing, learning, creative thinking, and intrinsic motivation. The computational play project aims to examine play formally using the lens of computation. While artificially intelligent (AI) agents are able to play against humans and perform well in highly structured, oppositional games such as chess (DeepBlue) or recently Jeopardy! (Watson), they lack the capacity to cope with the free-form, open-ended nature of play. We intended to synthesize research on play to develop agents that are able to engage in play with humans. Developing these agents will benefit cognitive scientists, AI researchers, game designers, and more broadly any person interacting with such playful virtual agents.
The Computational Play project is now in collaboration with Dr. Andrea Thomaz and Dr. Mark Riedl to create a turn-based, real-time computational play experience between a human and an AI system. This research will benefit researchers by enabling a better understanding of how to build co-creative intelligent agents and robotics for unstructured domains.
This research is supported, in part, by funding from The National Science Foundation IIS #1320520.
Zook, A., Magerko, B., and Riedl, M. (2011). Formally Modeling Pretend Object Play. In the Proceedings of the 8th ACM Conference on Creativity and Cognition, Atlanta, GA. (pdf)
Zook, A., Riedl, M. and Magerko, B. (2011). Understanding Human Creativity for Computational Play. In the Proceedings of the 2nd International Conference on Computational Creativity, Mexico City, Mexico. (pdf)