DeepMind Announces Plans to Utilize AI Models for Enhancing Physical Robots
Google DeepMind Launches New AI Models for Robotics
Introduction to Gemini 2.0
On Wednesday, Google DeepMind introduced two innovative artificial intelligence (AI) models aimed at the robotics sector, powered by their advanced Gemini 2.0 technology. According to Google, this new system represents their "most capable" AI to date and is designed to enable robots to perform physical tasks, moving beyond the traditional outputs of text and images that generative AI typically handles.
Collaboration with Apptronik
To enhance its robotics capabilities, Google has formed a partnership with Apptronik, a robotics company located in Texas. Together, they plan to develop the next generation of humanoid robots utilizing Gemini 2.0. Apptronik has worked with industry leaders such as Nvidia and NASA, and recently concluded a substantial funding round, which saw participation from Google.
In demonstration videos, robots developed by Apptronik and equipped with Gemini capabilities were seen performing various tasks. They successfully plugged items into power strips, filled lunchboxes, manipulated plastic vegetables, and zipped up bags—all in response to straightforward spoken commands.
Key Features of Gemini Robotics
Google has outlined three essential qualities that AI models for robotics must possess to be effective:
- General Adaptability: Robots need to navigate and adjust to different situations.
- Interactivity: Quick comprehension and execution of instructions or responses to environmental changes are crucial.
- Dexterity: Robots must be able to perform tasks traditionally done by humans, like carefully manipulating everyday objects.
Gemini Robotics-ER
Aiming to cater to roboticists, Google has also introduced Gemini Robotics-ER (Extended Reasoning). This model serves as a foundational tool for developers looking to create their own AI solutions for robotics. Trusted testers, including Agile Robots, Agility Robotics, Boston Dynamics, and Enchanted Tools, currently have access to Gemini Robotics-ER, allowing them to tailor AI functionalities to their specific needs.
The Growing AI in Robotics Landscape
Google isn’t the only player in the race to integrate AI into robotics. In November, OpenAI announced an investment in Physical Intelligence, a startup dedicated to developing AI models and algorithms that facilitate general-purpose AI in real-world applications. As the industry grows, OpenAI has also brought on board the former head of Meta’s Orion, who will now guide its robotics and consumer hardware initiatives.
In addition, Tesla has made strides in the humanoid robotics market with the development of its Optimus robot, which indicates a rising trend towards smart, adaptable robotic solutions.
Anticipated Impact on Robotics
Sundar Pichai, the CEO of Google, emphasized the importance of robotics as a testing ground for AI advancements intended for real-world applications. He believes that with the help of Google’s multimodal AI models, robots will be able to make adjustments in real-time and adapt seamlessly to their surroundings.
As this technology progresses, it will likely reshape the way we interact with machines and redefine the role of robotics in everyday life, laying the groundwork for a future where robots offer valuable assistance across various tasks and industries.
