NextFin news, Elon Musk's artificial intelligence company, xAI, announced on Sunday, October 12, 2025, that it has recruited AI experts from Nvidia to develop advanced AI world models. This move positions xAI alongside competitors such as Meta and Google in the race to build AI systems capable of understanding and designing physical environments.
According to reports from the Financial Times and other sources, xAI hired Nvidia researchers Zeeshan Patel and Ethan He, who have experience in world model development. These AI models are trained using video and robotic data to comprehend real-world dynamics, a field where Nvidia leads with its Omniverse simulation platform.
World models represent a significant technological leap beyond current large language models (LLMs) like ChatGPT and xAI's own Grok, which are primarily text-based. The new models aim to generate interactive 3D environments, with applications in gaming and robotics AI systems.
Insiders revealed that xAI's goal includes creating AI-generated games with rich interactivity, confirming Elon Musk's earlier statement on the X platform about launching powerful AI-generated games by the end of 2026. On October 7, 2025, xAI also released an upgraded image and video generation model available for free use.
Experts in the tech industry view world models as a potential breakthrough that could expand AI applications from software to physical products such as humanoid robots. Nvidia has estimated the market potential for world model technology to be comparable to the size of the global economy.
xAI is actively recruiting additional talent in image and video generation technologies for its 'omni team,' offering annual salaries between $180,000 and $440,000. The company is also seeking a video game director to train Grok in producing AI-assisted game designs, with hourly wages ranging from $45 to $100.
This strategic hiring and development effort underscores xAI's commitment to advancing AI capabilities beyond text-based models, aiming to create AI systems that understand causal physical phenomena and real-time object interactions in diverse environments.
Explore more exclusive insights at nextfin.ai.
