1X has officially released its physics-grounded World Model to empower the Neo humanoid robot with a...
The AMW Read
Updates the 1X case study by demonstrating a shift toward embodied scaling via internet-scale video training, moving beyond teleoperated demonstrations.
NoveltySignificance
Robotics · Player MapScaling Laws
1X has officially released its physics-grounded World Model to empower the Neo humanoid robot with autonomous self-learning capabilities. By processing internet-scale video and prompts, the 20,000 USD consumer bot can now generalize actions across unfamiliar domestic tasks without specific teleoperated demonstrations. This marks a major shift toward embodied AI scaling, utilizing over 10,000 hours of robot data to predict real-world outcomes. As 1X prepares for 2026 shipping of its 30kg home robot, this technology accelerates the transition from pre-programmed tools to truly adaptive household assistants. 🤖🏠
