Google DeepMind's latest work on Gemini 2.0 for robotics presents a remarkable shift in how large multimodal AI models are used to drive real-world automation, introducing two specialized models: Gemini Robotics and Gemini Robotics-ER, which demonstrate the potential of taking a multimodal artificial intelligence model, fine-tuning it, and applying it for robotics. Traditional robots struggle with narrow specialization due to challenges such as lack of generalization, expensive training, and limitations in supervised learning, reinforcement learning, and imitation learning. Gemini Robotics addresses these issues by rethinking how robots are trained and interacting with their environments, using a multimodal model capable of solving dexterous tasks in different environments and supporting different robot embodiments. The model uses Gemini 2.0 as a foundation, integrating physical actions as a new output modality to control robots directly, allowing the robots to adapt and perform complex tasks with minimal human interventions.