Game-playing AI isn’t exactly a new thing but given Google’s deep-rooted ventures in the industry, they have come up with something that can follow natural language models to follow instructions in a video game environment and the real world to some extent.
The AI gaming agent is named SIMA, abbreviated from Scalable Instructable Multiworld Agent. Unlike what OpenAI or Google’s very own Gemini offers, SIMA has been trained to follow instructions in certain controlled environments, most notably inside video games where they claim it is capable of working alongside another human.
At the time of writing, the model is highlighted to be effective in video game environments, but long-term goals would likely include effective utilization of an instruction following the AI language model to be effective in multiple scenarios in real life while assisting humans in a variety of tasks.
“We partnered with game developers to train SIMA on a variety of video games. 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.” read their statement.
“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, then rewatch what they did and record instructions that would have led to their game actions,” they added.
Google Deepmind partnered with game developers to enhance their AI agent (SIMA) by exposing it to diverse virtual environments. Nine video games, including Teardown and No Man’s Sky, provided a variety of challenges that expanded SIMA’s capabilities. Each game introduced a new interactive world and a set of skills to master.
Stay updated with the latest news in gaming and esports by signing up for early access to TalkEsport’s Mobile App. Follow us on Twitter and Google News to ensure you never miss the exciting updates.
