Google DeepMind uses Gemini 4 Argon to train a humanoid robot from video
Original titleTeaching a humanoid robot to play cornhole with Gemini 4 Argon
AISummary
Google DeepMind researchers gave Gemini 4 Argon a video of a human playing cornhole. The model built a MuJoCo simulation and calculated the physics and whole-body control needed for Apptronik's Apollo 2 humanoid to sink seven consecutive shots in the real world. The source says this Video-to-Sim-to-Real pipeline compresses weeks of manual modeling into a few hours.
AIWhy it matters
The source shows a concrete video-to-simulation-to-real workflow, a reusable pattern for training robots from ordinary footage, with a stated time saving over manual modeling.
Source: Google DeepMind · YouTube · youtube.comPublished · added here