Google is forming a new team to work on AI models that can simulate the physical world.
Tim Brooks, one of the co-principals of OpenAI's Video Generator, SoraWHO the left For Google's AI research lab, Google DeepMind, last October, he will lead the new team announcement In a post by X. It will be a part of Google DeepMind.
“DeepMind has ambitious plans to build massive generative models that simulate the world,” Brooks wrote Monday morning. “I'm recruiting for a new team with this mission.”
According to job list As Brooks linked in his post, the new modeling team will work with and work with Google Gemini, i seeAnd Genie teams to tackle “significant new problems” and scale models “to the highest levels of computation.” Gemini is Google's flagship series of AI models for tasks like image analysis and text generation, while Veo is Google's own video generation model.
In Genie's case, it's a world model – AI that can simulate games and 3D environments in real time. Google's latest Genie model, preview Last December, a huge variety of playable 3D worlds could be created.
“We believe in scaling [AI training] Video and multimodal data are on the critical path to artificial general intelligence,” reads one of the job descriptions. Artificial General Intelligenceor AGI, generally refers to AI that can perform any task a human can perform. “World models will power numerous domains such as visual reasoning and simulation, planning for embodied agents, and real-time interactive entertainment.”
According to the description, Brooks' new team will look to develop “real-time interactive generation” tools on top of the models they've built and study how to integrate their models with existing multimodal models such as Gemini.
A number of startups and big tech companies are chasing the global model, including influential AI researcher Fei-Fei Li. World LabsIsraeli upstart DescartesAnd The Odyssey. They believe that world models could one day be used to create interactive media such as video games and movies, and to run realistic simulations such as training environments for robots.
Work on massive world simulation models with Tim and the DeepMind team 🙂
On the complicated path of AGI. https://t.co/4Zuju5eMHb
— Logan Kilpatrick (@OfficialLoganK) January 6, 2025
But creatives have mixed feelings about the technology.
A recent A Wired investigation found that game studios like Activision Blizzard, which has laid off many workers, are using AI to cut corners, boost productivity and compensate for attrition. and a 2024 Study Run by the Animation Guild, a union representing Hollywood animators and cartoonists, estimates that 100,000 US-based film, television and animation jobs will be disrupted by AI by 2026.
Few startups in the new world modeling space, like Odyssey, have pledged to collaborate with creative professionals — not replace them. We'll see if Google follows suit.
There are also unresolved issues of copyright. Some world models appear to be trained on clips of video game playthroughs, which could make companies developing those models the target of lawsuits where the videos were unlicensed.
Google, the owner of YouTube, claims that it is allowed to train its models on YouTube videos, according to the platform's terms of service However, the company did not say which specific videos it is sourcing for training.