NVIDIA and Google Announce Collaboration to Advance Physical AI
NVIDIA and Google have announced a collaboration to advance "Physical AI," expanding AI models from purely digital environments to robotics, automation systems, and real-world applications. This partnership combines NVIDIA's strengths in computing power and simulation platforms with Google's expertise in models and data capabilities.
"Physical AI" emphasizes enabling models to understand and act in the real world, including scenarios such as industrial robotics, autonomous driving, and smart manufacturing. The collaboration is seen as an important step in the evolution from generative AI to AI systems with execution capabilities.
Source: Public Information
ABAB AI Insight
This collaboration represents a critical leap for AI from "information processing" to "physical execution." The previous wave of AI primarily transformed digital content such as text, images, and code, while "Physical AI" directly impacts the production side—robots, factories, and logistics systems—indicating that AI is beginning to influence the capital and labor structure of the real world.
NVIDIA's role is changing at this stage, shifting from a GPU supplier to a provider of "AI infrastructure + simulation environments," with its Omniverse platform becoming a key tool for training models of the physical world. Google provides models and data, and the combination forms a closed loop of "computing power + models + environment," which is a necessary condition for the implementation of Physical AI.
A deeper change lies in the rewriting of the production function. Traditional automation relies on fixed rules and mechanical processes, while Physical AI introduces learnable and adaptable systems, enabling machines to handle uncertainty. This will expand the boundaries of automation from structured scenarios to semi-structured and even unstructured environments.
From a historical perspective, this is similar to the second deepening phase of industrialization after electrification: it does not replace a single job type but reconstructs the entire production system. Whoever controls the technology stack and standards of "Physical AI" will be closer to pricing power in the future industrial system. This is also the fundamental reason for the early alliances between tech giants and computing power providers in this field.