Microsoft has recently launched a new type of AI that can essentially act as the ChatGPT for video game development where it can analyze old data and basically create gameplay. It’s called Muse AI and Microsoft even partnered up with developer studio Ninja Theory to help Xbox developers. One of the main reasons for Muse AI is apparently to help preserve classic games and to optimize them for modern hardware, not unlike what remakes and remasters do.
The news of this development for the new AI model came from Microsoft itself, announced on their official blog. Muse was originally called World and Human Action Model (WHAM) but it was given a more palatable nickname. In any case, it’s a new AI model that can generate game visuals, actions, and of course, mechanics based on prompts and data.
“This allows the model to create consistent and diverse gameplay rendered by AI, demonstrating a major step toward generative AI models that can empower game creators,” as explained by Microsoft’s Fatima Kardar.
Now, despite the implications– as Muse can potentially be used to bypass the need for actual game developers (and therefore rush the game development process)– developer Ninja Theory assures everyone that the new AI model won’t be utilized for creating content.
“We don’t intend to use this technology for the creation of content. I think the interesting aspect for us that’s exciting, is how can we use technology like this to make the process of making games quicker and easier for our talented team, so that they can really focus on the thing that’s really special about games: the human creativity,” according to Dom Matthews of Ninja Theory
This assurance is then backed by Phil Spencer himself, CEO of Microsoft Gaming. Currently, Microsoft is exploring its options for Muse in how they could port previously un-portable and old games into current-gen platforms, supposedly as some kind of preservation effort.
Muse Isn’t Quite Ready Yet
Despite the promising development of Muse, it’s still far away from being able to generate actual remasters or remakes in 1080p and early demonstrations could only produce 3D imagery at 300×180 pixels running at 10 FPS (frames per second). Still, Phil Spencer remains optimistic about Muse’s potential.
“You could imagine a world where from gameplay data and video a model could learn old games and really make them portable to any platform where these models could run.
We’ve talked about game preservation as an activity for us, and these models and their ability to learn completely how a game plays without the necessity of the original engine running on the original hardware opens up a ton of opportunity,” says the Microsoft CEO.
The more important bit about Muse’s demonstrations, however, is that they can actually create gameplay and 3D imagery in real-time. Gamers could potentially just chat with Copilot regarding what game they want to play and Muse will presumably bring up the AI-generated game.
While it sounds great on paper, there’s certainly more to video games than cookie-cutter gameplay and copied visuals, most of which are just an amalgam of old data with no human judgment as to what can be considered original or novel. That could be one of the reasons why Microsoft is limiting Muse’s use case for old games and preservation
Beyond Muse, we already had examples of AI being used for restoration and remasters namely the GTA Trilogy remaster. If you can recall, AI was used to upscale the low-res 3D imagery in the game and the results were less than ideal but were still sold to the masses– with a disastrous reception.