New Unitree G1 AI Robot Just SURPASSED Human Skills

Unitree G1 is learning complex human skills from surprisingly little training. UC Berkeley's Workhorse system teaches it to catch boxes, kick packages, climb obstacles, and recover from pushes using recorded human demonstrations alone. Another G1 learns to catch flying boxes through Reflex, while Westlake's WR1 robots handle household chores and a new expressive robot face mimics human emotions. 📩 Brand Deals & Partnerships: [email protected] ✉️ General Inquiries: [email protected] What You'll See: 0:00 - Intro 0:35 - How UC Berkeley's Workhorse teaches Unitree G1 to catch boxes, climb obstacles, and recover from pushes. SOURCE: https://hybridrobotics.github.io/workhorse/ 5:28 - Why the new Reflex system helps Unitree G1 catch flying boxes using vision, rapid prediction, and coordinated full-body movements. SOURCE: https://toberj.github.io/reflex/ 7:41 - How Westlake Robotics' WR1 enables humanoids to handle everyday household chores and cooperate to fold actual clothes. SOURCE: https://pandaily.com/westlake-robotics-wr1-humanoid-general-brain-dexterous-hand-clothes-folding 10:05 - How XRobotDoll's expressive robot face displays happiness, anger, sadness, and surprise on command, without proving it actually feels emotions. SOURCE: https://pasqualepillitteri.it/en/news/22475/xrobotdoll-robot-emotions-viral-video-confirms 🚨 Why It Matters Humanoid robots are moving beyond walking and scripted demonstrations. Unitree G1 now learns complex physical tasks directly from human examples, while Reflex enables split-second reactions to flying objects. Alongside new household robots and expressive faces, these advances point toward more useful humanoids, although real-world reliability remains largely unproven. #ai #unitree #robots