Prime Highlights:
Google DeepMind introduced two advanced AI models, Gemini Robotics and Gemini Robotics-ER, powered by Gemini 2.0, its most capable AI technology to date.
Google has partnered with Texas-based robotics company Apptronik to develop next-generation humanoid robots using Gemini 2.0.
Key Background:
Google DeepMind has introduced two advanced AI models designed to power physical robots, marking a significant step in the integration of artificial intelligence into real-world applications. The new models, Gemini Robotics and Gemini Robotics-ER (Extended Reasoning), are built on Gemini 2.0, which is touted as Google’s most capable AI technology to date. These models extend beyond generative AI capabilities, such as text and image creation, to provide control over robotic actions in the physical world.
In partnership with Apptronik, a Texas-based robotics company known for its work with Nvidia and NASA, Google aims to advance the development of humanoid robots powered by Gemini 2.0. Apptronik, which recently secured a $350 million funding round, is set to utilize these new AI models in the creation of next-generation robots. Demonstration videos showcased the robots performing tasks such as plugging devices into power strips, filling lunchboxes, moving plastic vegetables, and zipping up bags, all in response to spoken commands.
Google highlighted three essential qualities for robotics AI models: generality, interactivity, and dexterity. These traits ensure that the robots can adapt to various situations, understand and respond to instructions, and perform tasks that require fine motor skills, similar to human hand movements. The Gemini Robotics-ER model is particularly aimed at helping roboticists build custom models and is already available to trusted testers, including companies like Boston Dynamics and Agile Robots.
While Google is leading in AI-driven robotics, it faces competition from companies like OpenAI and Tesla, both of which have also invested in AI for robotics development. Google’s CEO, Sundar Pichai, emphasized that robotics serves as a crucial testing ground for applying AI advancements in the physical world, enabling robots to adapt to their environment and respond to dynamic changes. However, the company has yet to announce a specific timeline for bringing this technology to market.
