Exploring the modern possibilities of AI Directors

Games have often used the concept of using an AI to direct gameplay in a way that would be more enjoyable by players in the game. This has most often been targeted at dynamically adjusting the difficulty of encounters so that players will be more successful when playing through a game. However, the general concept can extend more broadly to systems that work to ensure that the player has learned the necessary skills required to advance in the game. We hypothesize that modern learning and planning techniques can be leveraged to create simpler and more robust AI directors that are able to fulfill a broader range of purposes in each game.

Faculty Supervisor:

Nathan Sturtevant

Student:

Kristen Yu

Partner:

Improbable

Discipline:

Computer science

Sector:

Information and cultural industries

University:

University of Alberta