Author

Aaron Weiss

Date of Award

2016

First Advisor

Aaron Williams

Second Advisor

Eric Kramer

Abstract

Artificial intelligence (AI) has been a topic of fascination ever since the first computers were invented, and a field that continually persists in its popularity. From your Terminator franchises to your phone’s personal assistant program, AI has been a mainstay of modern society over the past three quarters of a century, perhaps a little longer. One area of application for artificial intelligence that has always been both at the forefront of pop culture and academia is gaming. All sorts of video games currently employ varying levels of advanced AI to stimulate their human users, maintaining an equilibrium between challenging the player while not frustrating them (perhaps using multiple difficulty settings). At the same time, we see progress in academic gaming, with projects such as Google’s AlphaGo program, which just recently beat the most dominant Go player in the world using deep neural networks. At this point the abilities and styles of artificial intelligence in gaming vary wildly from one game to the next, and from one developer to the next. It was with that in mind that I set out to create an artificial intelligence for a newfound popular card game, Exploding Kittens . In doing so I also set out to create a framework to adapt the game for the computer, so that it could be played electronically. This thesis chronicles my months of creating a program to 3 do this, while also providing some insight into the history of AI in gaming, and the challenges of adapting card and board games to the digital world.

Share

COinS