Elon Musk’s xAI targets gaming with ‘world models,’ hires Nvidia experts

Elon Musk’s artificial intelligence venture, xAI, is making a bold move into the video game industry, leveraging next-generation 'world models' designed to understand and interact with the physical world. This strategy places the company in direct competition with tech titans like Meta and Google, who are also racing to develop AI systems capable of learning from real-world physics and environments.

Unlike current AI tools that primarily predict text or images, xAI’s advanced 'world models' aim to comprehend three-dimensional environments and simulate life-like gaming environments. These systems focus on understanding causal relationships and how objects behave, such as how a ball bounces or how light moves, moving AI capabilities beyond simple pattern prediction.

To rapidly accelerate this ambitious project, xAI has been quietly recruiting top researchers with expertise in world modeling, including former Nvidia researchers Zeeshan Patel and Ethan He. Nvidia has been a pioneer in this field with its Omniverse platform, which is designed for building and running realistic simulations.

xAI is now on an aggressive hiring spree for its ‘omni team,’ which is centered on creating AI experiences across images, video, and audio. The company is seeking talent for positions with high-end salaries, ranging from $180,000 to $440,000 annually.

A key role in this effort is a ‘video games tutor’ tasked with training Grok, xAI’s existing AI chatbot, in game design principles and mechanics. This move signifies that Grok will be a central tool in the development of the future AI-generated content.

Musk himself has confirmed on X (formerly Twitter) that xAI plans to release a “great AI-generated game” before the end of next year (2026). The dive into gaming, and the fusion of AI with the physical world through 'world models,' marks one of Musk’s most ambitious commercial moves to date in the evolving AI landscape.