My background is primarily in artificial intelligence, especially as it relates to creative design domains. I'm interested in all aspects of "intelligent" computation, including historical, theoretical, and technical perspectives.
One of my lines of work aims at bringing more intelligent tools to the game-design process, mainly at the prototyping stage. The overall vision is a "CAD for game design": an environment that helps a game designer prototype their game, while also providing useful feedback, automation, and visualization where desirable, roughly in the tradition of automated and assisted design in fields such as architecture.
My projects in the direction of automated/assisted game design have included a system to "auto-skin" simple games to be "about" different things, and Ludocore, a game engine in which game mechanics are encoded in symbolic logic. I also maintain an annotated bibliography of systems that generate new videogames.
I've done some work on interactive narrative, particularly on ways to help authors write complex, branching narratives without having to manually specify every possible combination. This uses a declarative, optimization-based approach, and aims for what we call "authorial leverage". Other perennial interests include computer music, generative/algorithmic art more generally, statistical machine learning, and the history of technology.
With some of my ITU colleagues, I am currently writing a textbook on procedural content generation in games (draft chapters are posted free online as we complete them).
In my spare time, I'm a Wikipedia administrator (since 2003). It's partly a hobby I participate in for its own interest (and because I support the goals of the project). But I also think it's quite relevant to my day job as an academic: since it's now a place where a large number of people get their information on a regular basis, I think it's useful for academics to participate in, understand, and be engaged in its community.
I also write essays on various subjects, some of which are preliminary versions of future papers, or less formal expositions of past papers.