THE robotics company behind the NEO humanoid, 1X, has unveiled a new AI model it says understands real-world dynamics and can help its robots learn new actions on their own.

Called the 1X World Model, the physics-based system combines video and natural-language prompts to give NEO robots new capabilities. According to 1X, the video inputs let NEOs learn tasks they were not explicitly trained to perform.

The announcement comes as 1X prepares to ship NEO humanoids to consumers. The company opened preorders in October and says demand exceeded expectations.

With this update, users give NEO a simple voice or text prompt, and it uses what it’s looking at to generate visualizations of future actions, and a built-in inverse dynamics model then translates these into precise movements for NEO to complete the request.

Demonstrations in 1X’s latest video showcase NEO’s ability to generalize beyond training data. For simple prompts like packing a lunch box, NEO visualizes and executes fluidly, even with unfamiliar objects. More impressively, NEO handles completely novel tasks, such as operating a toilet seat, opening a sliding door, ironing a shirt, brushing a human’s hair and more without any prior examples in its dataset. This highlights the transfer of broad human knowledge through the World Model.

“After years of developing our World Model and making NEO’s design as close to human as possible, NEO can now learn from internet-scale video and apply that knowledge directly to the physical world,” founder and CEO Bernt Børnich said in a statement. “With the ability to transform any prompt into new actions — even without prior examples — this marks the starting point of NEO’s ability to teach itself to master nearly anything you could think to ask.”

A 1X spokesperson clarified that while the World Model enables the robots to attempt a wide range of tasks, in practice the system has so far learned only simple actions. Still, those small gains are an early step toward more complex behaviors. The World Model also provides 1X with insight into how NEO plans to behave in response to prompts, generating data that the company can use to improve future training and capabilities.(SD-Agencies)