Brooks, who joined Google DeepMind in October, announced the initiative on X, stating, “DeepMind has ambitious plans to make massive generative models that simulate the world. I’m hiring for a new team with this mission.”
Job listings linked in Brooks’ post reveal the new team will collaborate with Google’s Gemini, Veo, and Genie projects to tackle emerging challenges and scale models to unprecedented computational levels. Gemini, Google’s flagship AI series, is designed for tasks like text generation and image analysis, while Veo focuses on video generation. Genie, Google’s take on world modeling, can simulate dynamic 3D environments in real time, with its latest model showcased in December.
According to the job descriptions, Brooks’ team will develop tools for real-time interactive generation and integrate their work with multimodal models like Gemini. “We believe scaling AI training on video and multimodal data is critical to achieving artificial general intelligence,” one listing reads, referencing AGI, a term for AI capable of performing tasks on par with human intelligence. World models, the team notes, could drive innovations in visual reasoning, robotics simulation, and interactive media.
The competition in world modeling is heating up. Companies like Fei-Fei Lee’s World Labs, Israeli startup Decart, and Odyssey are all racing to harness this technology for applications ranging from video games to robot training environments. However, the creative industry remains skeptical.
Game studios such as Activision Blizzard have faced criticism for using AI to streamline production while cutting jobs. A study commissioned by the Animation Guild predicts over 100,000 US-based film, TV, and animation jobs could be disrupted by AI by 2026. Some startups, like Odyssey, have committed to working alongside creatives rather than replacing them, but Google’s approach remains unclear.
Copyright issues add another layer of complexity. Many world models reportedly train on gameplay footage, which could lead to legal disputes if the material is unlicensed. While Google claims its use of YouTube content complies with the platform’s terms of service, it has not disclosed specific details about the videos used for training.