After Mira Murati, another OpenAI executive leaves company

Tim Brooks, a key figure at OpenAI, is leaving the company to join Google’s DeepMind. At OpenAI, Brooks co-led the development of a video generation system called “Sora” alongside William Peebles. Sora is an AI model designed to generate video content, but it hasn’t been released yet due to some technical challenges. Brooks announced his departure on X (formerly Twitter), expressing gratitude for his two years at OpenAI, where he contributed significantly to the Sora project.
Brooks will now work at Google DeepMind, the AI research arm of Google, where he’ll focus on developing video generation technologies and “world simulators.”
These simulators are AI-driven models that can create and simulate entire virtual worlds, potentially for use in games, movies, or even training AI agents. Demis Hassabis, CEO of DeepMind, welcomed Brooks and said that his role would help make the “dream of a world simulator” a reality. Though the term “world simulator” may sound abstract, it refers to AI systems like DeepMind’s Genie, which can generate interactive virtual environments from images, sketches, and real photos.
What Was Tim Brooks’ Role at OpenAI?
At OpenAI, Tim Brooks played a critical role in the development of Sora, a video generation AI model. He helped launch and steer the project’s research direction and led efforts in training large models for video creation. Sora was unveiled to the public in February 2023 but faced several technical difficulties. One major issue was the system’s slow processing speed—reportedly taking more than 10 minutes to produce a 1-minute video clip. This put it at a disadvantage compared to rival systems developed by companies like Luma and Runway, which have more efficient video generation tools.
Brooks’ departure comes at a time when OpenAI is working on an improved version of Sora to address these setbacks. The company is reportedly trying to enhance the system’s speed and overall performance to make it competitive in the rapidly evolving field of video generation technology. Despite these efforts, Brooks has decided to move on, perhaps seeking new opportunities and challenges at DeepMind, where his expertise in video generation and AI models can be further developed.
What Will Tim Brooks Do at Google DeepMind?
At Google DeepMind, Tim Brooks will be focusing on video generation technologies and “world simulators.” World simulators, although not fully defined, are models that can create interactive, controllable virtual worlds. These simulations can be applied to various industries, including gaming, movies, and AI training. For example, DeepMind’s recent model, Genie, can generate entire virtual worlds from a mix of synthesized images, real-world photos, or even simple sketches. These worlds can then be used for controllable content creation, where users can interact with and explore the simulated environments.
The potential applications of these world simulators are vast. In gaming, they could revolutionize how virtual worlds are created and explored by players. In film production, these tools could simplify the creation of visual effects or entire scenes. Moreover, in AI research, they could be used to train robots or other AI systems in virtual environments before deploying them in the real world.
Brooks’ move to DeepMind signals his interest in advancing these cutting-edge technologies, and his expertise in video generation from his time at OpenAI will be invaluable in pushing the boundaries of what’s possible with world simulators and AI-driven video creation.
In summary, Tim Brooks is a talented researcher leaving OpenAI, where he co-led the development of the Sora video generator, to join Google DeepMind. At DeepMind, he will work on video generation and world simulators, which have the potential to create interactive virtual environments for various uses in games, movies, and AI training.