Intelligence grows through experience.

The physical world is rich with lessons. Every interaction reveals something about movement, materials, space, and cause and effect. We believe these experiences can become a foundation for intelligence that helps people in the world we share.

Our vision is a future where human-like systems perceive, learn, think, adapt, and act in the physical world. By human-like, we mean capabilities inspired by how people understand and interact with the world, across many forms of machines and systems.

Our mission is to advance physical intelligence by enabling machines to learn from real-world experience.

We’re starting with an open learning layer for Physical AI. Our goal is to help developers and researchers turn observations, actions, and outcomes into useful learning data, and connect experience, training, and evaluation in a continuous cycle of improvement. Each deployment should help make the next one more reliable.

Our guiding principle is to make physical intelligence open and composable. We want developers to combine, reuse, and improve capabilities for perceiving, learning, and acting in the physical world. Through open-source foundations and modular design, we aim to make it easier for individuals and teams to experiment, contribute, and build on one another’s progress.

We believe that progress must serve people. Physical intelligence should reduce dangerous and exhausting work, support human judgment, and expand what people can do. Human dignity, safety, and the freedom to make meaningful choices will guide what we build and how we build it. We will measure progress by how it improves people’s lives as well as what it enables machines to do.

Built on Earth by humans.