Microsoft’s Muse is a generative AI video game model that can generate gameplay and controller actions in real-time. It’s claimed to be the world’s first World and Human Action Model (WHAM) and will help developers learn as well as experiment with the weights, sample data, etc.
WHAM basically trains on “ground truth” gameplay video and input data provided by actual players. The data for Microsoft’s demonstrator comes from Bleeding Edge, a four-on-four online brawler released in 2020. It required around 1 million training updates before the WHAM model started showing basic understanding of complex gameplay interactions, like a power cell item exploding after three hits from the player. Ultimately, Muse should enable game developers to prototype their game without needing actual human beta testers.
- Dive into every mission, open world, game match, and more with speeds up to 120 FPS with the best value in gaming.
- Quick Resume: Seamlessly switch between your favorite games and pick up right where you left off.
- Lightning-fast load times: Jump into your favorite games like Minecraft, Fortnite, and Roblox instantly.
I look forward to seeing the many ways in which the community will explore these models and build on our research. I cannot wait to see all the ways that these models and subsequent research will help shape and increase our understanding of how generative AI models of human gameplay may support gameplay ideation and pave the way for future, novel, AI-based game experiences, including the use cases that our colleagues at Xbox(opens in new tab) have already started to explore,” said Katja Hofmann , Senior Principal Research Manager of the Microsoft Research Game Intelligence team.
[Source]
