Skip to content

Instantly share code, notes, and snippets.

@iondune
Last active August 29, 2015 14:01
Show Gist options
  • Select an option

  • Save iondune/9dd0a4f6fc669cd9fd45 to your computer and use it in GitHub Desktop.

Select an option

Save iondune/9dd0a4f6fc669cd9fd45 to your computer and use it in GitHub Desktop.
Research Project Topic Proposal

Knowledge-based Authentic AI for an Open-World Game

Date

Thursday, May 22nd (Week 8)

Abstract

Knowledge representation is a fine candidate for implementing a complex and authentic artificial intelligence system for a game, yet many current game systems avoid knowledge-based systems in favor of more traditional artificial intelligence systems and hacks. We explore the flaws of many simple AI systems and discuss the implementation of a knowledge-based authentic AI system for an open-world sandbox game.

Outline

  1. Introduction
    1. Overview
  2. Background 2. Theory 2. Halo
  3. Existing Systems 3. World 3. Game
  4. Problems with Traditional AI Systems 4. Uncanny Knowledge 4. Oblivious Bystander 4. Canonical View
  5. Solutions 5. Bigger-Picture Model
  6. Knowledge Representation 6. Facts 6. Belief 6. Personality 6. Emotion
  7. Interaction Representation 7. Language
  8. Conclusions

References

  • Damian Isla
  • GDC 2005
  • 03.11.2005
  • Presentation
  • Ryan Schmitt
  • Date unknown
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment