Part of this transition will involve developers getting used to the capabilities of Google's cloud infrastructure, Harrison said. For decades, game developers have had to be “device-centric,” Harrison noted, designing games for the specific and limited characteristics of home hardware. “With cloud gaming, especially the idea of sharing computing resources across multiple processors in a data center, the move to data-centric gaming is going to be a really fundamental shift,” he said.
Developers who try to take advantage of this will be able to implement everything from “distributed physics” to “complex multiplayer, ranging from hundreds to tens of thousands in a very complex world,” he said. In Stadia's multiplayer games, “every change I make to my world can be instantly, in microseconds or less, communicated to all other clients… You can't do that with a separate box.”
This is very similar to Microsoft's “revolutionary” features. it is promised that this will be possible on Xbox One thanks to integration with the Azure cloud.. Completely destructible environment in Acceleration of 3 multiplayer was perhaps the loudest implementation of Microsoft's promise to date, but they it ended up being a bit of a disappointment in practice.
Elsewhere in the talk, Harrison also began creating elevated visions of Stadia games with “conversational understanding” of the player's voice commands transmitted through the controller's built-in microphone. Google's research into artificial intelligence and machine learning could help create games with non-player characters that “conversate contextually,” he said, using data center storage for a “huge database” of potential conversation choices generated in real time. According to Harrison, this kind of machine learning work could one day help Stadia developers create content more cheaply, with less labor and time.
