Authored vs. Systemic - Finding a Balance for Combat AI in Uncharted 4
Offered By: GDC via YouTube
Course Description
Overview
Explore the development process of Uncharted 4's combat AI in this 57-minute GDC 2017 session presented by Naughty Dog's Matthew Gallant. Delve into the challenges and solutions encountered while balancing authored and systemic approaches to AI design. Learn about macro combat, new features, and AI overview, including connectivity, vantage layouts, and heat search techniques. Discover how the team drew inspiration from PacMan ghosts to create core combat roles and implement global combat parameters. Gain insights into evergreen AI techniques, their limitations, and the lessons learned about partitioning control between design and engineering. Conclude with a Q&A session to further understand the intricacies of combat AI development in AAA game production.
Syllabus
Introduction
Outline
Part 2 Solutions
Macro Combat
New Features
AI Overview
The Problem
Connectivity
Vantage
Layouts
Heat Search
Authored vs Systemic
Heart Points
PacMan Ghosts
Core Combat Roles
PacMan AI Comparison
Role Assignment
Global Combat Parameters
Evergreen AI Techniques
Limitations
Conclusion
Questions
Taught by
GDC
Related Courses
Introduction to Artificial IntelligenceStanford University via Udacity Probabilistic Graphical Models 1: Representation
Stanford University via Coursera Artificial Intelligence for Robotics
Stanford University via Udacity Computer Vision: The Fundamentals
University of California, Berkeley via Coursera Learning from Data (Introductory Machine Learning course)
California Institute of Technology via Independent