Future Game AI

DOI: 10.5176/978-981-08-8227-3_cgat08-6

Authors: Narendra S. Chaudhari

Abstract:

Game AI has witnessed substantial growth in the last decade. Currently, game AI is used for modeling various aspects of non-player characters (NPC). While such applications (e.g. behavior modeling) have been very successful, the deployment of game AI for serious games having learning as main component poses basic questions, like the representation of an explanation capability as perceived by the human player. In this paper, we propose a general framework for the development of such an explanation capability. Our framework is based on the exploitation of the known results linking (formal) language models and (finite state) machine models in theoretical computer science. It is well-established fact that Turing Model, as a machine model (and its equivalent models like Chomsky’s phrase structured grammar model, or recursively enumerable functions) is a most general model for computation. Many biologically inspired computing paradigms such as traditional neural networks, genetic algorithms, evolutionary computing frameworks etc., aim at addressing the problem of automatic generation of algorithmic solutions for many computationally difficult problems. Various soft computing models have complementary capabilities. This motivates us to propose a framework for their integration. We identify the need for future game AI engines with such capabilities.

Keywords: Game AI, Hybrid soft computing, neural network optimization, genetic algorithms, Memetic Algorithms, Small World Models.