"Google's AI Masters Cooperative Gameplay in Unfamiliar 3D Worlds"

Google DeepMind has unveiled SIMA, an AI agent designed to learn gaming skills and play video games more like a human rather than an overpowered AI. The AI is not trained to win games but to follow instructions and play alongside human players, with the ability to learn and adapt to various games, including open-world and non-linear ones. Currently in the research phase, SIMA has been trained using games like No Man’s Sky and Goat Simulator 3, and has about 600 basic skills. While not intended to replace existing game AI, it is envisioned as another player that can impact game outcomes and may eventually be able to perform more complex tasks and interact with players in natural language.
- Google's new AI will play video games with you — but not to win The Verge
- Google Deepmind trains a video game-playing AI to be your co-op companion TechCrunch
- Google DeepMind's Latest AI Agent Learned to Play 'Goat Simulator 3' WIRED
- Google DeepMind's new AI can follow commands inside 3D games it hasn't seen before Engadget
- Google AI learns to play open-world video games by watching them New Scientist
Reading Insights
0
1
2 min
vs 3 min read
77%
509 → 119 words
Want the full story? Read the original article
Read on The Verge