Google has revealed Gemini Robotics 2.0, a model aimed at giving robots better whole-body control, according to Ars Technica. The update is framed around improved dexterity and safety for machines that need to perceive, reason and move in the physical world.
Robotics models are different from chatbots because errors can have physical consequences. A system that controls arms, hands or a mobile body must interpret its environment, plan actions and avoid unsafe behavior while dealing with messy real-world conditions.
The promise of whole-body control is that a robot can coordinate more than one movement at a time instead of treating tasks as isolated gestures. That could matter for household, warehouse or lab settings where manipulation requires balance, timing and adaptation.
The limitation is that demonstrations do not automatically translate into reliable deployment. Robots still need hardware, sensors, testing and clear safety constraints. Gemini Robotics 2.0 is another sign that AI labs are moving from screen-based assistants toward systems that act in physical spaces.