Google DeepMind’s New “SIMA” AI Can Play Video Games Like Humans

Tech giant Google’s artificial intelligence division, Google DeepMind, has introduced SIMA, an AI agent that has the potential to play video games like a professional human gamer.

The AI model is under research, and it will behave more like a human than a highly advanced AI while playing games. SIMA, which stands for Scalable Instructable Multiworld Agent, will eventually acquire gaming skills and will be able to play any game. 

SIMA does not aim to become a replacement for the existing gaming AIs, and it will just gel in as a virtual co-op player. This learning AI would be able to understand verbal instructions while interpreting the 3D and virtual worlds as well. Rather than becoming a hard-coded AI bot, SIMA will provide its users with a more human-like gaming experience.

Google DeepMind stated in a blog post,

This research marks the first time an agent has demonstrated it can understand a broad range of gaming worlds and follow natural-language instructions to carry out tasks within them, as a human might. This work isn’t about achieving high game scores.

Notably, Google DeepMind is renowned for its remarkable achievements in the gaming world. It has built AI bots that are capable of being victorious over the world champions in games such as Go, Chess, and Stratego. This research division has also created models that can participate in games and play them seamlessly without even being taught the rules beforehand. Now, it has shifted its focus completely to video games.

As per the blog post, SIMA is trained to perform simple tasks within ten seconds. The AI agent has approximately 600 chief skills, like turning left, climbing a ladder, and opening the menu to use a map.

During the training and testing of SIMA, Google has worked with many game developers, including Hello Games, Embracer, Tuxedo Labs, Coffee Stain, and more.