Game Development Reference
HTN planners are gaining traction in real-time-strategy games [Brickman08]
[Cerpa08] and other genres. The potential for better performance is always
welcome in computer games. The increase in intelligence is equally welcome; not
only do agents make workable plans, but they re-plan when things go wrong.
The hot topics among professional game AI programmers are always the subject
of presentations at the Game Developers Conference (GDC), particularly at the
AI summits that have been held in recent years. Along with academics and
various other people outside of the industry, this same group of professional
game AI programmers has authored the articles in the AI Game Programming
Wisdom series of topics. They also present at other conferences such as AIIDE
and AAAI. Some of them even run Web sites devoted to game AI.
The first volume of the AI Game Programming Wisdom series, edited by Steve
Rabin, has an entire section devoted to A*. Machine learning is the subject of
the articles in section 10, and neural networks are the subject of the articles in
section 11 in AI Game Programming Wisdom 2 , although by volume 4, the editor
openly asks, ''What happened to Learning?'' [Rabin08]. The entire four-volume
series was written by and for industry professionals, so some of the articles may
be hard going for novice AI programmers. A full list of the articles in the series as
well as from many other sources can be found at http://www.aiwisdom.com/ .
The noteworthy Web sites for game developers are too numerous to keep track
of, so we will mention only a few:
n http://www.gamedev.net . The AI section of gamedev.net is reasonably
n http://aigamedev.com . The aigamedev.com site is focused entirely on game
AI at a professional level, although some of the articles require paid
membership to view.
n http://www.gamasutra.com . The gamasutra.com site tends to collect GDC-
related material; Damian Isla's behavior tree paper is available there.
n http://www.wikipedia.org . While in no way guaranteed to be academically
rigorous or even factually correct , Wikipedia is probably the best place on the
Internet to read about a topic for the first time, including AI-related topics.
Although Wikipedia is a good place to get the first clue about a topic, all