Google DeepMind cooks up Gemini robots that pack, plan, and snoop online

Google DeepMind released two advanced artificial intelligence systems for robotics applications. Gemini Robotics 1.5 and Gemini Robotics-ER 1.5 represent a departure from traditional preprogrammed machines. These models enable robots to adapt, reason through problems, and acquire knowledge independently. Carolina Parada leads the robotics division at Google DeepMind.

The dual-system approach separates cognitive functions between models. Gemini Robotics-ER 1.5 handles reasoning tasks and web research, while Gemini Robotics 1.5 executes physical actions. This combination allows machines to plan before acting rather than following single-step commands. Knowledge transfers between different robot designs become possible.

Healthcare and personal assistance represent primary application areas for these enhanced systems. Privacy concerns, safety questions, and reliability issues require extensive testing before widespread deployment. Google DeepMind aims to transform mechanical tools into collaborative human partners.
 

Attachments

  • Google DeepMind cooks up Gemini robots that pack, plan, and snoop online.webp
    Google DeepMind cooks up Gemini robots that pack, plan, and snoop online.webp
    45.3 KB · Views: 63
Top