Presentation is loading. Please wait.

Presentation is loading. Please wait.

Game AI Fundamentals. What is Artificial Intelligence (AI)? Not easy to answer… “Ability of a computer or other machine to perform those activities that.

Similar presentations


Presentation on theme: "Game AI Fundamentals. What is Artificial Intelligence (AI)? Not easy to answer… “Ability of a computer or other machine to perform those activities that."— Presentation transcript:

1 Game AI Fundamentals

2 What is Artificial Intelligence (AI)? Not easy to answer… “Ability of a computer or other machine to perform those activities that are normally thought to require intelligence” (from dictionary) “Process” or “science” of creating intelligent machines Artificial “brains” that carries out intelligent behaviour

3 Perspectives of AI People study AI to: ▫Gain insight into nature of human intelligence ▫Create machines that exhibit some form of intelligence So…do game developers need to STUDY AI?

4 How do we know… …if a machine is “intelligent”? ▫How close is it to human intelligence? ▫Requires the machine to have conscience and emotions? ▫Ability to solve a problem that humans can solve? ▫Ability to learn and adapt to new situation? Turing Test

5 Weak AI vs. Strong AI Strong AI – Produce machines that with an intelligence that match or exceed that of humans Weak AI – Produce machines that are able to “act” intelligently

6 Chinese Room Argument Does the machine literally understand Chinese? (Strong AI) Is the machine simulating the ability to understand Chinese? (Weak AI)

7 Deterministic vs. Nondeterministic Deterministic AI ▫Behavior is specified and predictable, no uncertainty ▫Fast, easy to implement, Gameplay limitations Nondeterministic AI ▫Behavior has a degree of uncertainty, less predictable, learn and extrapolate from experience ▫More burden of implementation and testing ▫Gameplay evolution, emergent behavior

8 Established Game AI 1.“Cheating” – most widely used technique, abusing this can destroy gameplay. 2.Finite state machines 3.Fuzzy logic or Fuzzy state machines – improve unpredictability, reduce burden of too many rules 4.Pathfinding – simple shortest path navigation, terrain-based/obstacle-based extensions 5.Rule-based, Probability-based decision making

9 Millington’s Model

10 Future of Game AI The next big thing: Learning – Bayesian techniques, NN, GA ▫NPC behaviors are nondeterministic ▫Game AI can evolve, learn and adapt the more it’s played ▫Implementation/debugging challenge Artificial life and other Emergent AI


Download ppt "Game AI Fundamentals. What is Artificial Intelligence (AI)? Not easy to answer… “Ability of a computer or other machine to perform those activities that."

Similar presentations


Ads by Google