Google DeepMind says it’ll power robots with AI

In the rapidly advancing realm of artificial intelligence, one major tech company is planning to combine the software and hardware aspects of its AI operations. Google’s DeepMind intends to use its Gemini AI model to operate real-world robots.

Today, Google published a blog post unveiling the Gemini Robots, a new innovation. Equipped with the latest version, Gemini 2.0, these robots function as sophisticated vision-language-action (VLA) systems. In other words, they have the ability to respond physically to tasks assigned to them. Additionally, Google presented Robotics-ER, which are self-aware and capable of autonomously programming themselves.

Google has shared a video showcasing the capabilities of their robots. The video displays a robot performing tasks such as packing lunch, organizing items, and folding paper neatly. In one instance, a person asks the robot to rearrange red dice so that the same number on a green die next to it is matched.

Google is teaming up with Apptronik to construct humanoid robots within this project. They express their aim to introduce practical, AI-driven robotics solutions into our everyday lives.

Read More

2025-03-13 00:27