Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
If you have used ChatGPT or twin Before, I had used Great language model. This core technology is what creates the text you see on your screen, and is responsible for the vast majority of AI products. But it may not be the most important AI technology.
The development of artificial intelligence has gone through many stages, and the next stage will be less concerned with creating words and more focused on understanding our natural world. Global models are designed to translate our physical world – such as the laws of physics, object detection and movement – into a digital blueprint that AI can understand.
You likely won’t interact with global models in the same way you would with technology that supports LLM, such as chatbots. However, global models will demonstrate how AI can create realistic videos, guide surgical robots, and enhance the driving capabilities of self-driving vehicles. They are important building blocks in the development of so-called physical AI — technology that not only understands our world, but can take action in it.
Don’t miss any of our unbiased technical content and lab reviews. Add CNET As Google’s preferred source.
A variety of AI leaders have signaled the shift toward building a global model. Yann LeCun, a leading pioneer in the field of artificial intelligence, He left recently His role leading Meta’s AI efforts for a startup focused on building global models. Spatial intelligence — the ability to understand your physical environment — is the next frontier of technological innovation, said Fei-Fei Li, colloquially known as the Godmother of AI.
“Spatial intelligence will transform the way we create and interact with real and virtual worlds, revolutionizing storytelling, creativity, robotics, scientific discovery, and beyond,” she wrote in November. Blog post.
Nvidia CEO Jensen Huang also set aside a portion Consumer Electronics Show 2026 The key word for the company’s efforts is in global models. Building an AI model based on the laws of physics and ground truth starts with the data used for training, Huang said.
Watch this: Every announcement from the Nvidia Live CES 2026 broadcast
AI models of all kinds require huge amounts of data to build and improve their outputs. Typically, AI companies rely on content created by real humans — with or without their consent — which has led to this Major legal confrontations. Global models can be built using human data, including simulations. This data is necessary to build global models capable of thinking and making judgments on the basis of cause and effect.
The Nvidia Cosmos global model uses text, images, and video to understand the physical world.
One area where Nvidia is using global models is self-driving cars. In a live demo, Nvidia demonstrated how its global prototype, Cosmos, uses a car’s sensors to understand its location and the location of every other nearby car on the road to create live video of its surroundings. Developers can use this information to run scenarios, such as car crashes, to see how the vehicle responds and make necessary safety improvements. Synthetic dataor non-human-generated data, can also be useful in conjunction with global models to help predict rare “edge cases.”
As AI continues to be integrated into every part of our online lives, it is essential that it can understand our physical world, rather than continuing to hallucinate and make mistakes. Renewed research and investments from industry leaders in spatial intelligence, global models, and physical AI show that the industry isn’t just building more chatbots — it’s building AI more rooted in our reality, not the other way around.