Beyond Adversarial: The Case for Game AI as Storytelling


Roberts David L. Riedl Mark O. Isbell Charles L.
2009 DiGRA '09 - Proceedings of the 2009 DiGRA International Conference: Breaking New Ground: Innovation in Games, Play, Practice and Theory

As a field, artificial intelligence (AI) has been applied to games for more than 50 years, beginning with traditional two-player adversarial games like tic-tac-toe and chess and extending to modern strategy games, first-person shooters, and social simulations. AI practitionershave become adept at designing algorithms that enable computers to play games at or beyond human levels in many cases. In this paper, we argue that the traditional goal of AI in games—to win the game—is not the only, nor the most interesting goal. An alternative goal for game AI is to make the human player’s play experience “better.”AI systems in games should reason about how to deliver the best possible experience within the context of the game. The key insight of this paper is that approaching AI reasoning for games as storytelling reasoning makes this goal much more attainable. We present an overview of traditional game AI techniques as well as a few more recentAI storytelling techniques. We also provide afoundation for describing and reasoning about games as stories, citing a number of examples. We conclude by discussingthe implications forfuture directions.