Rule based systems:

• comprised of a database of associated rules

• relatively simple to use and can be exhaustively tested

• rule sets can be written using simple proprietary scripting languages, rather than full programming languages

• scripting languages available to the general public

Learning systems:

• a set of extremely powerful learning techniques which have been used widely in industry and academia

• it is difficult to frame a learning problem in the dynamic worlds in which games are set

• game developers do not have a great

• understanding of learning techniques

The need for really intelligent solutions (examples):

• units in strategy games that cannot find their way safely across a game environment

• partner characters that get in the way of the player

Cheating the player:

• see through walls

• always have perfect aim

• limitless amunition

• endless recources

Intelligent agents propeteries:

• autonomy (operating without intgervention of human and having kind of control of him)

• social ability (interacting with other agents)

• reactivity (perceiving their environment)

• pro - activness (they are able to exhibit goal-directed behaviour by taking the initiative)

Reactive agents:

• operate in a hardwired, stimulous-response driven manner

• they require very little support infrastructure

• reactive architectures are completely deterministic

• every situation must be encoded within a system’s rules (no learning or adaptability)

• design of a reactive system is extremely difficult

• not capable of any kind of long term planning

Deliberative agents:

• they build internal models of their world, which they then use to formulate plans to achieve goal states

• the main advantage is ability to form long term plans

• cannot make the real-time guarantees

• they require constant maintenance of a knowledge base

Hybrid agents:

• combine aspects of both the reactive and deliberative approaches (ex mixing the time critical behaviours with long term planning)

Reqiurements of ntelligent Agents as Virtual Humans:

• illusion of believability

• choosing reasonable actions and respond realistically

• should have strong personality and be capable of showing emotion

• must appear to make choices which improve their situation

Relationships:

• engaging users in interactions

• maintaining relationships with both the user and other agents

Fidelity - refers to the fact that virtual reality systems need only remain true

Proactivity - considering the moments before and after a character appears in a scene as crucial (theatre). Traditional game agent techniques ignore these two notions completely..