xAI Hires Nvidia Researchers to Develop Advanced AI World Models

Elon Musk’s AI company xAI has hired two researchers from NVIDIA to advance its work on “world models,” AI systems designed to understand and simulate real-world environments. The new hires, Zeeshan Patel and Ethan He, bring experience from NVIDIA’s Omniverse platform, which specializes in physics-based simulations.

World models are trained using video and robotics data to help AI systems comprehend motion, causality, and physical interactions. Unlike language models such as Grok or ChatGPT, these systems aim to extend AI capabilities into physical applications like robotics and gaming.

xAI plans to use the models to generate interactive 3D environments and enhance AI-driven robotics. Musk has stated that the company intends to release an AI-generated game before the end of next year, aligning with its focus on integrating real-world simulation into AI development.

The company recently introduced a new image and video generation model available for free, and it continues to recruit experts for its “omni team,” which works on multimodal AI systems capable of processing text, images, video, and audio. Salaries for these roles range from $180,000 to $440,000, with additional specialized positions such as a “video games tutor” to train AI systems in game design.