DeepMind Announces Plan to Utilize AI Models for Physical Robots
Google Introduces New AI Models for Robotics
On Wednesday, Google DeepMind launched two new AI models specifically designed for robotics, both powered by Gemini 2.0, which the company claims is its most advanced AI to date. This marks a significant step towards integrating artificial intelligence with robotics, enabling machines to interact with the physical world more effectively.
The Gemini Models
The new models introduced are named Gemini Robotics and Gemini Robotics-ER (for extended reasoning). These models extend the capabilities of traditional AI, which has typically focused on generating text and images, by providing robots with commands that allow them to perform physical tasks.
Key Features of Gemini Robotics
According to Google, successful AI models in robotics must have three essential qualities:
- Generalization: The ability to adapt to different scenarios.
- Interactivity: Quick understanding and responsive actions to instructions or environmental changes.
- Dexterity: Skills for manipulation resembling what humans can do with their hands and fingers.
Demonstration videos showcased robots from Apptronik, a Texas-based robotics company, performing tasks such as plugging devices into power outlets, filling a lunchbox, and arranging plastic vegetables. These robots interacted with their environment based on spoken commands, showcasing the practical applications of the Gemini models.
Partnership with Apptronik
Google announced a partnership with Apptronik to develop the next generation of humanoid robots utilizing Gemini 2.0. Apptronik has a notable background, having collaborated with industry players like Nvidia and NASA. Recently, Google participated in Apptronik’s $350 million funding round, indicating strong support for their robotics initiatives.
Gemini Robotics-ER
Gemini Robotics-ER is specifically tailored for robotic developers, providing a foundation for them to create and train their own models. This version of Gemini is available to Apptronik and a select group of "trusted testers," including well-known entities in the robotics field like Agile Robots, Boston Dynamics, and Enchanted Tools.
The Larger AI Robotics Landscape
Google is not the only player in the rapidly evolving AI-centric robotics industry. In a notable development last November, OpenAI invested in Physical Intelligence, a startup focused on integrating general-purpose AI into physical applications through large-scale models and algorithms tailored for robots.
OpenAI has also made strategic hires to boost its robotics efforts, including appointing a former Meta executive to oversee its initiatives. Additionally, companies like Tesla are advancing in humanoid robot development with their Optimus robot, further emphasizing the competitive landscape in robotics.
Insights from Google CEO
Google’s CEO, Sundar Pichai, shared insights about the company’s vision for robotics, stating that they view the field as an ideal proving ground for applying AI breakthroughs in tangible ways. He emphasized that the robots developed with Gemini models will leverage multimodal AI to adapt and respond dynamically to their environments.
As Google and other tech giants continue to invest in AI-driven robotics, we can expect significant advancements that could transform how humans and machines interact in everyday life. The integration of these technologies will likely pave the way for more intelligent, responsive robots capable of assisting in various tasks across multiple sectors.
