Google DeepMind is assembling a new team of artificial intelligence researchers to develop “world models” that can simulate physical environments. The initiative will be led by Tim Brooks, a former co-lead for OpenAI’s Sora project who joined DeepMind in October to work on Google’s video generation and world simulators.
World models are a relatively new development within AI that could serve a variety of purposes, such as creating real-time interactive media environments for video games and movies, and realistic training scenarios for robots and other AI systems. It’s also part of Google’s push to achieve an artificial general intelligence system, or AGI, before its competitors.
“DeepMind has ambitious plans to make massive generative models that simulate the world,” Brooks announced in an X post on Monday. Brooks included two open job listings for research engineers and scientists who will help to advance AI “world models” capable of simulating real-world scenarios by solving problems around training “at massive scale,” curating training data, and studying how they can be integrated with multimodal language models.
DeepMind has ambitious plans to make massive generative models that simulate the world. I’m hiring for a new team with this mission. Come build with us!https://t.co/pqvALtAvLs https://t.co/vtwgeXl9Dl
— Tim Brooks (@_tim_brooks) January 6, 2025
“We believe scaling pretraining on video and multimodal data is on the critical path to artificial general intelligence,” DeepMind said in the job descriptions. “World models will power numerous domains, such as visual reasoning and simulation, planning for embodied agents, and real-time interactive entertainment.”
The race to be the first to declare AGI is heating up, so Google’s focus here isn’t surprising. OpenAI CEO Sam Altman recently said that the company has cracked how to achieve the tech industry’s long-sought benchmark, and that autonomous AI agents may start to meaningfully join workforces this year.
There are several competitors who already have a head start over DeepMind’s world modeling project, including Nvidia’s Cosmos platform for advancing physical AI, autonomous vehicle, and robot development; and the World Labs startup created by “the godmother of AI,” Fei-Fei Li.
Introducing Genie 2 – our most capable large-scale foundation world model, which can generate a diverse array of consistent worlds, playable for up to a minute. We believe Genie 2 could unlock the next wave of capabilities for embodied agents . pic.twitter.com/AfL3EbOMeB
— Jack Parker-Holder (@jparkerholder) December 4, 2024
The new DeepMind team will work alongside existing Google AI projects including its flagship Gemini AI models, Veo video generator, and Genie — Google’s prior world model for simulating playable 3D environments in real-time.
