Skild AI trains Unitree G1 to play soccer with 140 years of simulated self-play
Distinct-Question-16 · reddit · 2026-09-24
Skild AI showed a Unitree G1 humanoid robot trained to play soccer via self-play reinforcement learning: the model competed against increasingly tough versions of itself, accumulating the equivalent of 140 years of practice in simulation. The demo highlights large-scale simulated self-play as an effective approach for embodied robot skill training.
More from Embodied
- Guava distills GPT-5.4 into a 4B robot agent with just 2,000 simulated trajectories — jbhuang0604 · 2026-11-09
- Dex-One2Many learns dexterous manipulation from one human video, 95%+ success across 4 hands — jbhuang0604 · 2026-11-09
- Fly-by-Code: coding agents pilot drones via active visual and physical feedback loops — jbhuang0604 · 2026-11-09
- μ0 world model learns robot physics from video only, beats π0 by 5 points on RoboCasa365 — jbhuang0604 · 2026-11-09
- NVIDIA unveils GR00T 2 open-world robot action model, hitting 95% success with 5 demos — CyberRobooo · 2026-11-09
- CoRL keynote to explore open-source data and hardware for robotics — micoolcho · 2026-11-09