World Labs, founded by Fei-Fei Li, is developing “spatial intelligence” to enable artificial intelligence to understand and interact with the three-dimensional physical world. The startup raised US$ 1 billion in funding from investors including Nvidia, AMD, and Autodesk to build models that incorporate geometry, physics, and time, moving beyond current text and image generation capabilities.
What Is Spatial Intelligence in AI?
Spatial intelligence allows artificial intelligence to build three-dimensional models of the environment, accounting for depth, movement, and physical causality. According to Fei-Fei Li, a professor at Stanford University and former vice president and chief scientist of AI and machine learning at Google, current AI models are limited by their inability to grasp the physical laws that govern our reality. By developing this capability, machines can shift from merely recognizing objects in an image to predicting how those objects behave in a 3D space.
Fei-Fei Li is widely recognized as the “madrinha da IA” (godmother of AI). Her 2006 creation of ImageNet provided the massive, cataloged image database that became the foundation for modern deep learning and computer vision.
How Will World Labs Apply These Models?
The primary goal for World Labs is to create “world models” that serve various industrial and creative sectors. These models aim to simulate physical environments, which can be used to train robots, accelerate the development of video games, and assist in medical or industrial simulations. The company’s first product, Marble, is being utilized by professionals in effects visuals, architecture, interior design, games, robotics, and researchers interested in tracking the psychological aspect of immersive environments.
The Evolution of Computer Vision
The pursuit of spatial intelligence builds directly on the progress of computer vision, which Li helped pioneer. Her work on ImageNet allowed for the training of algorithms at an unprecedented scale, directly enabling technologies like autonomous vehicles and medical diagnostic imaging. Following her tenure as director of the Stanford Artificial Intelligence Laboratory from 2013 to 2018, Li founded the Stanford Institute for Human-Centered AI (HAI), which emphasizes developing technology that remains aligned with human needs.
When tracking the progress of spatial AI, monitor the integration of physics-based engines in generative models. These engines are what allow AI to move from static 2D outputs to functional, interactive 3D simulations.
Policy and Global Leadership
Beyond her technical contributions, Fei-Fei Li is an active participant in global AI governance. She served as a member of the National Artificial Intelligence Research Resource Task Force for the White House and the National Science Foundation between 2021 and 2022. Since 2023, she has acted as a special advisor to the Secretary-General of the United Nations.
Frequently Asked Questions
- What is the main goal of World Labs?
The company aims to develop spatial intelligence, enabling AI to understand and simulate the physical, three-dimensional world. - What is the Marble product?
Marble is the first product released by World Labs, designed for use in fields like architecture, robotics, and gaming to create and analyze immersive environments. - Who is Fei-Fei Li?
She is a Stanford professor, the creator of the ImageNet database, and a prominent voice in AI policy who served as a special advisor to the UN.
Interested in the future of robotics and machine learning? Subscribe to our newsletter for weekly updates on the latest breakthroughs in spatial intelligence and AI policy.
