Google DeepMind's Gemini large language model enhances robotic capabilities

Google DeepMind's Gemini large language model empowers robots to understand commands and navigate physical spaces with up to 90% reliability. This model enables robots to process video and text inputs, aiding in functionalities like finding desired locations within an environment. By combining Gemini with specific action-generating algorithms, the robot can execute tasks based on instructions and visual cues, showcasing the potential of language models in real-world settings. Researchers see promise in expanding Gemini's applications across different robot types for improved problem-solving abilities.