Machine Takeover - Improving Your Game With Machine Learning & AI
summary
Forza’s Driveatar uses Bayesian learning plus a neural network to mimic human racers and ships in-game without breaking player expectations, Ubisoft prototypes real-time animation and lip‑sync controllers with ML that successfully speed production, and DeepMind’s AlphaStar boots from human replays then uses self‑play RL to surpass pros (training runs cost millions, ~$3.2M estimated). Practical entry paths include Unity ML-Agents with sample‑efficient training and on‑device inference via Core ML or Unity’s runtime, plus image segmentation with Slicer and OpenAI/Atari baselines for experiments.
AI-written summary · about
chapters (15)
- 0:00Intro
- 1:05Panel Introductions
- 3:35What is Machine Learning
- 6:30Best Examples
- 8:40Seed
- 12:40Best examples of machine learning in game
- 15:40How important are humans
- 20:45AI for eSports
- 22:55Best tools for developers
- 26:40Big drivers for machine learning
- 28:55Opportunities for machine learning
- 32:18Nextgen
- 34:10Monica
- 36:10Jeffrey
- 42:40Open Source
related talks
29mThe Present & Future of AI and Games
10mMachine Learning & Unity ML-Agents update
32m..Stories About Industrializing In-Game Machine Learning Solutions
32mThe Road to General AI
30mBuilding, designing, teaching and training simulation environments for Machine Learning
01h 01mIntroduction to Practical AI
46mAI Takeover
36mImplementing Machine Learning the Unity way