Tim Brooks, who led the development of the AI video generator Sora with William Peebles at OpenAI, announced his move to the Google DeepMind AI laboratory. There he will conduct research in the field of video creation using AI and “world simulators.” Brooks’ departure is rumored to be due to Sora’s technical problems and performance lagging behind competing systems from Luma, Runway and others.
Image source: Pixabay
Google DeepMind CEO Demis Hassabis is confident that Brooks’ arrival will help “make the long-time dream of a world simulator a reality.” The term refers to AI models like the recently released Genie, which can generate playable, controllable virtual worlds from synthesized images, real photographs, and even sketches.
At OpenAI, Brooks was one of the first researchers to work on the Sora AI model, unveiled in January 2023. Sources in the know link his departure to technical problems that allegedly plagued the system, which took more than 10 minutes to create a 1-minute video clip. OpenAI is reportedly in the process of training an improved Sora model that will be able to generate videos much faster.
In the spring of 2024, Google introduced its own video generation model called Veo. Veo is expected to soon be available to content creators on the short video service YouTube Shorts.
It appears that OpenAI is still lagging behind its competitors in advancing video content creation developments. Early last month, Runway signed a deal with Lionsgate to train a custom video model based on Lionsgate’s catalog of films. At the same time, Stability, which develops its own set of video generation models, added Avatar, Terminator and Titanic director James Cameron to its board of directors.
OpenAI demoed Sora to filmmakers and Hollywood studios earlier this year, but no long-term partnerships have been announced.
It’s interesting that Brooks is actually returning to Google, having previously worked on Pixel phones. It should be noted that he joined the series of high-ranking employees and founders who left OpenAI:
- Renowned researcher Andrej Karpathy left OpenAI in February.
- A few months later, OpenAI co-founder and former chief scientist Ilya Sutskever left, along with former security chief Jan Leike.
- In August, co-founder John Schulman said he would leave OpenAI.
- CTO Mira Murati, chief scientific officer Bob McGrew and vice president of research Barret Zoph announced their departure at the end of September.
- Company President Greg Brockman is on sabbatical.