NVIDIA unveils Physical and Agentic AI to move robots in the real world

From Chatbots to Physical Action

Artificial intelligence is undergoing a fundamental shift in its primary role. While the public has largely seen AI as a tool for generating text, images, or answering questions, NVIDIA is pushing it toward physical interaction. This evolution marks a transition from digital content creation to real-world execution.

At the AI Day Seoul 2026 event, NVIDIA highlighted two specific paths for this evolution: Physical AI and Agentic AI. Physical AI is designed to help machines perceive their surroundings and take direct actions. This technology allows a robot to understand a physical space and decide how to move through it or manipulate objects within it.

Agentic AI represents the reasoning side of this shift. Instead of just following a single prompt, Agentic AI can plan and carry out multiple steps to achieve a specific goal. As reported by 매일경제, these two technologies work together to allow AI to think, plan, and then physically act in the real world.

Simulating the Real World

Training robots in the real world is often slow and expensive because of the constant need for trial and error. To solve this, NVIDIA introduced the NVIDIA Isaac platform. This system allows developers to test and teach robots how to move in a virtual environment before they are ever deployed in a factory or on a street.

Supporting this is a new development platform called Cosmos. Cosmos is used to generate large amounts of data and videos that mimic the physical characteristics of reality. By using Cosmos, companies can create high-quality training data that helps robots learn how to handle objects and navigate complex environments safely.

NVIDIA also introduced the concept of an "AI Factory." This is a large-scale computing infrastructure designed to handle the heavy lifting of training, inference, and data generation. By combining Cosmos with an AI Factory, companies can generate the massive amounts of data required to train Physical AI at a scale that was previously difficult to achieve.

Multi-Agent Coordination

A major component of Agentic AI is the ability for different AI systems to work together. NVIDIA showcased multi-agent orchestration, which is a method where multiple AIs are assigned different roles to complete a shared objective. For example, one AI might gather and analyze information while another makes the final decision on how to proceed.

To make these systems accessible to businesses, NVIDIA uses a software technology called NIM. NIM consists of microservices that help companies easily deploy and operate AI models in their own environments. By combining NIM with multi-agent orchestration, businesses can build complex systems where different AI components handle specific parts of a workflow.

This approach moves away from a single AI trying to do everything at once. Instead, it creates a team of digital workers that can reason through a problem. When paired with Physical AI, these digital workers can then command physical robots to execute the final steps of a production process or a service task.

Impact on Global Industry

The practical applications for these technologies are expected to be most visible in high-precision industries. Manufacturing, automotive, and semiconductor production are primary targets because they involve complex processes that require extreme accuracy. Physical AI can help these sectors automate production equipment more effectively.

Beyond simple automation, the goal is to improve the actual work accuracy and efficiency of existing robots. Rather than just replacing human labor in repetitive tasks, these technologies aim to help robots handle more sophisticated movements. This could lead to faster production cycles and fewer errors in complex assembly lines.

NVIDIA is positioning itself as more than just a chipmaker to support this shift. The company is expanding into robotics development platforms and AI infrastructure. By providing the tools for both the reasoning (Agentic) and the acting (Physical) parts of AI, they aim to be the backbone of the next generation of industrial robotics.

The Road Ahead for Robotics

The transition to Physical AI suggests that the next era of AI will be defined by how well machines can interact with the physical world. As robots become more capable of understanding their environments, the line between digital intelligence and physical labor will continue to blur. This shift requires a massive amount of data and computing power, which is why NVIDIA is focusing so heavily on infrastructure.

For the average reader, this means that the AI we interact with will soon have