← Back to live feed1 story
Google DeepMind's Gemini 4 Argon model enabled an Apptronik Apollo humanoid robot to learn cornhole through zero-shot learning by observing a single human demonstration video, allowing it to autonomously generate a simulation environment and sink seven consecutive shots in a physical trial without prior task-specific training.
Sign in to suggest
Key sources
- SOURCE@normandipalo“asked the robot itself to explain what it did to us, as you can see at the end of the video. It even came up with the gestures”x.com
- SOURCE@google“sank seven straight shots in a real-world trial”x.com
- SUPPORT@googledeepmind“compresses weeks of manual modeling into just a few hours”x.com
- SUPPORT@chris_j_paxton“changing a lot of the assumptions about how intelligence for robotics is going to work”x.com
- SUPPORT@stocksavvyshay“Google’s massive YouTube library could become one of its biggest advantages in Physical AI”x.com
- SUPPORT@trungtphan“losing cornhole to a humanoid robot would be devastating”x.com
- SOURCEhuggingnewshuggingnews.com