World Labs, a startup founded by AI pioneer Fei-Fei Li, has unveiled its first project — an AI system capable of generating video game-style 3D scenes from a single image. Many AI systems can convert photos into 3D models and environments, but World Labs’ scenes are unique for their interactivity and modifiability.
“Our technology lets you step into any image and explore it in 3D,” World Labs wrote in their blog. The system transforms an image into an interactive 3D scene that can be explored using keyboard and mouse in a demo version on the World Labs website. Scenes are rendered in real time in the browser and feature a controllable camera with adjustable depth of field.
The World Labs system belongs to a new category of AI called “world models.” Many such models can simulate games and 3D environments, but often with artifacts and consistency issues. Unlike others, World Labs’ approach ensures that scenes remain unchanged after generation, adhering to basic laws of physics and providing a sense of solidity and depth.
World Labs also plans to create tools useful for professionals such as artists, designers, developers, filmmakers, and engineers. The startup has already raised $230 million in venture capital from investors including Andreessen Horowitz, Ashton Kutcher, Intel Capital, AMD Ventures, and Eric Schmidt. The company, valued at over $1 billion, hopes to launch its first product in 2025.
