For years now, video game designers have been trying to provide their players a strong sense of interactivity at the level of the story itself: This would allow players to deeply modifying the story, without diluting it into a confuse and inconsistent experience. While some entire books have been written on interactive narrative, there is no computer system that produces complex interactive narrative or drama, in a strong sense. Some people would conclude from this observation that it is vain to combine interactivity and narration. We claim, however, that the problem is just difficult, but solvable. Artificial Intelligence (AI) techniques could be applied to solve this issue. In fact, the topic has only recently interested research laboratories.

The challenge is more than providing new tools to digital authors or creating better video games. It is about creating a new form of art and entertainment, which might play the same cultural role in the 21st century that the movie did during the 20th.

The goal of this research is precisely to provide new algorithms to really simulate the narrative and make it interactive.

In the talk, we present our approach to interactive drama, and describe the basic components of the proposed model, based on several theories and practice in narrative. We'll present some initial experimental results and discuss future improvement on the system. Finally, we'll have a discussion on how simulation of narrative could have wider applications than entertainment.

Short resume