World Labs

World Labs is an AI company founded by Fei-Fei Li — the Stanford AI professor known for her work on ImageNet and computer vision — focused on building AI with spatial intelligence. The company's research centers on creating AI systems that can understand three-dimensional space and generate realistic 3D environments from images.

The company's spatial intelligence research addresses what Fei-Fei Li has described as one of the next major frontiers in AI — moving beyond 2D image and language understanding to true 3D spatial comprehension. This capability would enable AI to understand and reason about physical space in ways that current systems cannot.

World Labs' initial demonstrations showed the ability to generate explorable 3D worlds from a single image, creating coherent 3D scenes that maintain consistent spatial relationships and allow virtual camera movement through the generated environment. This capability represents a significant advance toward AI that truly understands physical space.

The company's approach combines advances in neural radiance fields (NeRF), 3D Gaussian splatting, and large-scale AI training to create models that can infer and generate 3D spatial content from limited image observations. These technical approaches represent the current frontier of 3D AI research.

For applications in AR/VR content creation, robotics, autonomous systems, and the future of spatial computing, World Labs' research represents foundational work on how AI will understand and interact with three-dimensional physical reality.