Release
NVIDIA just shipped a robot model that learns new tasks from a single demonstration. The economics of programming robots has changed.
The Brief
- NVIDIA launched GR00T 2 at CoRL in Austin, described as an open world action model for humanoid robots.
- The reported average success rate is 71% across 50 tasks when the robot has seen one demonstration.
- With five demonstrations, the reported success rate rises to 95%.
Why it matters
If these figures hold, teaching a humanoid robot a new task may need far fewer demonstrations than before, which lowers the cost of robot programming. The source is a short excerpt, so the benchmark details are not verified.
Who’s affected
- Robotics developers working on humanoid robots
At CoRL in Austin, NVIDIA launched GR00T 2 - its new open world action model for humanoid robots. The headline number is 71% average success across 50 tasks when the robot has seen exactly one demonstration. Five demonstrations gets you to 95% . One year ago, teaching a robot a…