Recently, it has been a growing trend for internet platforms, such as Bing or Microsoft Azure, to create an AI section of their website.
Like these websites, Google has created its own new AI called SIMA. However, while other AI bots are programmed to be one’s artificial intelligence assistant, SIMA alludes to the actions of a human rather than a robot.
It is designed to learn and grow with these skills rather than retain all knowledge from the internet. This also means that it has not reached its full potential yet; Google expects it to learn video games and even games that don’t have direct rules or a path to follow.
SIMA is unlike any other AI bot. In the video game field, SIMA coordinates itself with the players around it, meshing with other parties and using it to understand these games further. It is also considered an AI agent, which differs from ChatGPT’s artificial intelligence. AI models have limitations when operating independently since they are trained on large amounts of data. However, an AI agent is capable of processing data and acting on its own.
This artificial intelligence is a step forward in collaborating with humans in games and real-world tasks. 
SIMA thinks independently, “understanding” the commands or rules and following them. As a Google Deepmind researcher, Tim Harley, said, “SIMA isn’t trained to win a game; it’s trained to run it and do what it’s told.” This means that the more you interact with SIMA, the more it will learn and adapt to your commands.
Google collaborated with eight-game developers to train and test SIMA, including Coffee Stain, Hello Games, Embracer, and Tuxedo Labs. These games prepare the agent’s programming for a multitude of gaming environments and rules that a person may induce, such as menu use, mining resources, flying spaceships, and so much more.
One of the more meaningful games that SIMA was plugged into was Goat Simulator, a free-roam game that induces chaos of goats and unreal situations. Yet this was just what SIMA needed; the unruled gameplay with a witgameplayl ending would help SIMA
grow and learn better. “SIMA is greater than the sum of its parts,” says Frederic Besse, a research engineer at Google DeepMind.
After this training, SIMA can follow through with actions commanded by the player, such as, “Go to the spaceship,” “Chop down ten pieces of wood,” and more. “It’s still very much a research project,” says Tim Harley, another member of the Google DeepMind team. “However, one could imagine one day having agents like SIMA playing alongside you in games with you and your friends.”
According to the researchers, learning different language barriers in the plethoras of gaming worlds comprehend how it ties into gameplay. Our first approach was to record pairs of human players across the games in our portfolio, with one player watching and instructing the other. We also had players play freely, rewatch what they did, and record instructions that would have led to their game actions.”
SIMA AI could ultimately change the gaming world and community for the better. It is growing with over 600 skills, eventually leading your party to victory. SIMA is still picking up gaming basics and adjusting to new ones. According to Google, SIMA might subsequently be able to perform more challenging tasks and be the ideal AI party member to help you win whatever game you decide to play.
