Gemini Robotics 2 brings us one step closer to physical AGI
The New Stack Meredith Shubel ● Covered by 6 sources
Google DeepMind released Gemini Robotics 2, a system of three AI models enabling robots to perform complex multi-step physical tasks with full-body control and fine motor dexterity. The vision-language-action model controls robot movements, the embodied reasoning model handles planning and self-correction, and a lightweight on-device version runs without internet connectivity. Robots can now adapt skills across different embodiments with as few as 200 examples and coordinate with other robots to complete real-world tasks like fetching items from shelves or tying knots.
Why it matters
This week, Google DeepMind revealed Gemini Robotics 2, an intelligence layer comprising three new models to power more adaptable physical The post Gemini Robotics 2 brings us one step closer to physical AGI appeared first on The New Stack.
Also covered by
- Ars Technica — Google reveals Gemini Robotics 2.0, promising improved dexterity and safety
- MarkTechPost — Google DeepMind Ships Three Physical AI Models For Whole Body Control, Dexterity And Multi Robot Collaboration
- Google DeepMind — Gemini Robotics ER 2: powering robotics with video understanding, task orchestration, and multi-robot collaboration
- The Verge — Google DeepMind’s new AI model can control a robot’s entire body
- Google DeepMind — Gemini Robotics 2 brings whole body intelligence to robots
