For years, AI has learned to write, code, and create images. The next race is teaching machines to understand and interact with the physical world. World Labs just made a move that brings that goal a step closer.
The startup founded by AI pioneer Fei-Fei Li announced Tuesday that it has acquired SceniX, a robotics company known for building high-fidelity simulation systems that help robots learn in virtual environments before carrying those skills into the real world.
“The world is not just made of words, and spatial intelligence was never just about perceiving and generating worlds. It’s about interacting with them. Today, SceniX is joining World Labs,” ‘AI godmother’ Fei-Fei Li said in a post on X.
The world is not just made of words, and spatial intelligence was never just about perceiving and generating worlds. It’s about interacting with them.
Today, SceniX is joining World Labs. 🌎🤖👇 pic.twitter.com/qThtkjgd3f
— Fei-Fei Li (@drfeifei) July 21, 2026
The acquisition marks World Labs’ biggest step yet into embodied AI, a field focused on giving AI systems the ability to perceive, reason, and act in physical spaces instead of remaining confined to screens and chat interfaces.
World Labs said the acquisition builds on its long-term goal of developing “spatial intelligence,” or AI systems that can perceive, generate, reason about, and interact with both virtual and physical environments. The company believes future robotics breakthroughs will come from combining world models, learning-based simulation, and real-world feedback into a single training loop.
“Today, we’re excited to announce that SceniX, a robotics company, is joining World Labs,” the company said in a blog post.
Li echoed the announcement in a post on X, writing: “The world is not just made of words, and spatial intelligence was never just about perceiving and generating worlds. It’s about interacting with them. Today, SceniX is joining World Labs.”
The message reflects a broader shift happening across the AI industry. Large language models changed how people interact with computers. The next phase centers on systems that can understand physical environments, predict what happens next, and take actions inside those environments through robots and other autonomous machines.
Bringing AI from virtual worlds to physical ones
The announcement included a 36-second video showing the direction World Labs is heading.
It opens with frozen blue waters before shifting to robotic manipulation of a glossy green sphere moving through a series of marble tracks. Robotic arms guide the object through increasingly complex environments before the video transitions to a humanoid robot operating inside a home. The sequence closes with the World Labs logo and the words “Acquires SceniX,” highlighting the company’s plan to connect AI-generated worlds with real robotic systems.
Li later introduced members of the SceniX team, including cofounders Yunzhu Li and Changxi Zheng, whose work focuses on building simulation platforms where robots can safely learn new skills before deploying them on physical hardware.
That “sim-to-real” approach has become one of robotics’ biggest priorities. Training robots entirely in the real world is slow, expensive, and risky. High-fidelity simulations let developers generate millions of training scenarios at a fraction of the cost, then transfer those learned behaviors onto actual robots.
Why SceniX matters
SceniX has built its reputation around hybrid simulation technology for general-purpose robotics.
Its software creates realistic environments where robots can practice manipulation, movement, and interaction before those same policies are tested on physical machines. The company says its systems have shown strong transfer from simulation into real-world deployments, moving beyond controlled laboratory demonstrations.
That expertise fits naturally with World Labs’ flagship technology, Marble.
Introduced earlier this year, Marble generates persistent, spatially consistent 3D environments from text prompts, images, videos, panoramic photos, or rough 3D layouts. Users can explore, edit, and expand those environments, making the platform useful for creative workflows, gaming, virtual reality, simulation, and robotics research.
Adding SceniX gives World Labs technology that could transform those generated environments into training grounds where robots learn how to interact with objects before performing the same tasks in factories, warehouses, homes, or other physical settings.
A bigger bet on spatial intelligence
World Labs emerged from stealth in September 2024 with a mission to build AI systems capable of perceiving, generating, reasoning about, and interacting with three-dimensional environments.
The company has attracted some of Silicon Valley’s largest investors. It raised $230 million shortly after launch, followed by a $1 billion funding round in February 2026 that valued the startup at $1.23 billion. Investors include Nvidia, AMD, and Autodesk, which invested $200 million and secured an advisory role.
The acquisition of SceniX signals that World Labs sees robotics as the next proving ground for its spatial intelligence research.
Many AI companies are racing to build larger language models. World Labs is pursuing a different challenge. Its focus is teaching machines to understand geometry, movement, physics, and cause-and-effect so they can operate reliably in the physical world.
That goal reaches far beyond chatbots. Success could influence warehouse automation, manufacturing, household robots, industrial inspection, autonomous systems, and digital simulation platforms used across multiple industries.
World Labs has not shared a timeline for integrating SceniX or announced new products tied to the acquisition. The company says it plans to reveal more about SceniX’s technology and future roadmap in the coming months.
One thing is already clear. World Labs is betting that the next leap in AI will come from machines that can do more than generate text or images. They will need to understand space, predict physical outcomes, and interact with the real world. SceniX gives the company another piece of that puzzle.
