The world of video games and in particular its big names have always appreciated the technological developments in vogue. While the NFT/cryptocurrency era didn’t last long (unless it comes back with a vengeance soon), it’s now AI that’s polarizing the gaming industry’s attention. Recently, Ubisoft announced that it wanted to use AI and deep learning to help write game scripts, but now it’s Google that wants to exploit AI in video games.
Harnessing AI to better test video games
This patent recently filed by Google provides a system and method for training a locally executed “fictitious” player to perform real-time game actions in a video game based on one or more data models. These models would be generated by a distance learning service, with the aim of improving the effectiveness and efficiency of video game testing. The goal is to solve the problem of adapting video game testing processes to the complexity of modern video games.
It was in April 2022 that Google filed a patent titled “EFFICIENT GAMEPLAY TRAINING FOR ARTIFICIAL INTELLIGENCE,” but the one that was filed last week deepens the process. To make it simpler, Google wants to allow a game character to make decisions and take actions based on the data collected in the video game. The data learned would come from previous AI tests, but also from observation generated by a human player to further modify the model, more precisely.
“Observation data is generated by the local actor component based on the in-game results of artificial game actions performed by the local actor component, based, at least in part, on inferences generated by the actor component using the provided game data model”.
The fictional player would react differently depending on different parameters sent remotely such as player reference position, object position, motion vector, geometric information and game reward indicators for example.
Google is not the only group to be interested in AI to facilitate game development processes as EA recently filed a similar patent centered on an artificially intelligent system to automate game quality assurance testing. video games through learning by imitation. It would seem that Google wants to create its own technology and this could be of interest to many video game companies wishing to streamline the development process, improve the quality and efficiency of their testing, as well as reduce the time it takes to release new video games. .
