allows a more natural style of development for AI agents in addition to traditional game AI programming. AC agents’ behaviors are trained rather than programmed. Anyone on the development team can sit down with a console or a PC and play in the role of an AC agent. AC agents will learn tactics and strategies directly from its human controller, without the need for coding. Once an AC agent is set up there is no need for recoding: it can be trained and retrained to play in an unlimited number of styles. Additionally, AC may be set up to automatically fine-tune a learned style through self-learning. For example, a learned style may be automatically fine-tuned to increase or decrease its effectiveness.
saves time and money during the development process. The SDK dramatically reduces the amount of time required to program individual characters’ AI. AC agents can also be used for game testing purposes. Self-learning AC agents will find weaknesses in regular AI agents and also problems with game balance.
allows celebrities to participate directly in the game development. As part of a sponsorship deal with an athlete or a movie star the celebrity could train his/her own character or team.
The resources used by are insignificant in comparison to the benefits it brings. On the PS2 platform it consumes under 1Mb of memory and uses less than 20% of processing power. Enhancements to the memory or processing power of game platforms improve the performance of AC technology. This scalability means that the software should not need major adaptations as hardware technology develops.