Google DeepMind has brought in Tim Brooks, the research lead behind OpenAI's Sora video generation system, for work that points directly at one of the most closely watched areas in artificial intelligence: AI-powered video and world simulation.
Brooks announced on X that he is joining Google DeepMind. In the post, he said he will work on video generation and world simulators, thanked OpenAI for an "amazing two years at OpenAI making Sora," and wrote, "Can't wait to collaborate with such a talented team."
A notable hire in AI video
Brooks' move matters because Sora has become a reference point in the AI video market. OpenAI's introduction of Sora in February created significant buzz and put pressure on competitors, making the researcher associated with that system a high-profile addition for Google DeepMind.
The hire may indicate that Google DeepMind wants to strengthen its position in AI-powered video generation. The company has worked on video AI and vision projects before, including Project Astra, Veo, Imagen Video and Phenaki. Even so, Google has yet to make a breakthrough in video AI comparable to Sora.
That makes Brooks' new role especially relevant. His stated focus is not only video generation, but also world simulators, a term that connects generated video to broader questions about whether AI systems can model environments in useful ways.
Why world simulators are now part of the video race
The discussion around Sora has never been limited to polished clips. OpenAI sees Sora as an important step towards artificial general intelligence (AGI). It also claims that Sora can simulate and understand the physical world, which the source describes as a controversial theory.
That claim helps explain why a video model can become strategically important beyond media generation. If a system can produce coherent video, the next question is whether it is merely arranging pixels or representing something deeper about objects, motion and physical environments.
Google DeepMind CEO Demis Hassabis welcomed Brooks in language that pointed beyond ordinary video tools. He wrote, "So excited to be working together to make the long-standing dream of a world simulator a reality!!"
That framing suggests Google DeepMind views the work as more than a feature race. Video generation is the visible output, but world simulation is the broader ambition attached to it.
What this means for OpenAI and Sora
Brooks' departure is a setback for OpenAI because of his role in Sora. The system has been presented as an important part of OpenAI's broader AGI direction, and losing a research lead associated with it gives a rival company relevant experience at a sensitive moment.
OpenAI is reportedly working on an improved version of Sora. The goal is to generate longer, higher-quality video clips more quickly, though there is no concrete timeline for release.
That leaves OpenAI in an active but unresolved phase. Sora has already changed expectations in AI video, but the next version has not arrived on a fixed schedule. At the same time, Brooks is moving to a company that has both existing video AI projects and explicit interest in world simulators.
The market is moving quickly
Since Sora's unveiling in February, the AI video market has developed rapidly. New competing products have launched from Chinese companies, while US companies have released model updates.
The source identifies Luma AI, RunwayML and Pika among the US companies involved in those updates. That wider activity shows that Sora did not simply draw attention; it helped intensify a competitive field around AI video generation.
The current landscape includes several overlapping threads:
- Talent movement: Tim Brooks is moving from OpenAI to Google DeepMind.
- Model competition: Sora remains a benchmark while rivals continue to release products and updates.
- Research ambition: video generation is increasingly discussed alongside world simulators.
- Uncertainty: OpenAI is reportedly improving Sora, but no concrete release timeline is available.
For Google DeepMind, the hire adds a researcher directly linked to one of the most discussed AI video systems. For OpenAI, it removes a key Sora figure while the company is reportedly working on a more capable version. For the AI video market, it is another sign that generated video and world simulation are becoming central areas of competition.