Using generative AI for dynamic difficulty adjustment to maximize player flow in FPS games

Abstract

Dynamic difficulty adjustment becomes more prevalent as game designers focus more on game flow and accessibility. This kind of system promises to keep the player in a flow state by accommodating each individual skill level. Until now, this has been accomplished manually with in-game algorithms, but with the rise of artificial intelligence, a new way of balancing presents itself. Using generative AI for DDA might be the next step in reshaping the concept of game balance. In this project I have created a first-person shooter (FPS) game that uses generative artificial intelligence (AI) to handle dynamic difficulty adjustment (DDA). It will highlight some strengths and limitations of this approach, and demonstrate the possibilities of what AI-enhanced DDA can do.