Google, which owns YouTube, asserts it complies with the platform’s terms of service when training its AI models
Google DeepMind is assembling a new team to create advanced AI models capable of simulating the physical world. This initiative will be led by Tim Brooks, a former co-lead on OpenAI’s video generator, Sora, who joined Google DeepMind in October. Brooks announced the development on social media platform X, highlighting the team’s mission to build “massive generative models that simulate the world.”
The project will integrate existing work from Google’s Gemini, Veo, and Genie teams to tackle innovative challenges. Gemini, Google’s flagship AI, excels at tasks like image analysis and text generation, while Veo focuses on video creation. Genie, on the other hand, is a model designed to generate interactive 3D environments in real-time. These technologies will form the foundation for Brooks’ new team.
Scaling Towards Artificial General Intelligence
According to job postings linked to Brooks’ announcement, the team aims to scale AI training on video and multimodal data, a critical step towards achieving artificial general intelligence (AGI). AGI refers to AI systems capable of performing any task a human can.
The models will serve diverse applications, including visual reasoning, simulation, planning for robotics, and real-time interactive media like video games and movies. A primary focus will be on developing tools for “real-time interactive generation,” allowing seamless integration with multimodal models such as Gemini.
Industry Race & Ethical Questions
Google isn’t the only player in this space. Startups like Fei-Fei Lee’s World Labs, Decart, and Odyssey are also pursuing world models. These models promise a range of uses, from creating immersive video games and films to training robots in realistic environments.
However, not everyone is optimistic about the implications of this technology. A 2024 study by the Animation Guild estimated that AI could disrupt over 100,000 jobs in the U.S. film, television, and animation industries by 2026. Reports suggest companies like Activision Blizzard are already leveraging AI to cut costs and address workforce shortages.
Some startups, like Odyssey, have pledged to collaborate with creative professionals rather than replace them. It remains to be seen if Google will take a similar approach as the industry grapples with concerns about job displacement and ethical AI usage.
Copyright Concerns Persist
Another challenge facing world models is copyright. Many are trained on video game playthroughs and other media, which could expose developers to legal risks if unlicensed content is used. Google, which owns YouTube, asserts it complies with the platform’s terms of service when training its AI models. However, the company has not clarified which videos are being utilised for this purpose.
Google’s investment in world modelling underscores its commitment to advancing AGI while navigating the ethical and legal complexities of the emerging technology. Whether this initiative can balance innovation with responsible AI development remains to be seen.
