Artificial intelligence is being revolutionized by agentic systems that are making the technology more pervasive and altering the way that enterprises interact with other enterprises and with users. The next revolution will make AI more available both in computing and in the physical world, said the founder of Nvidia.
“Computing has been fundamentally reshaped as a result of accelerated computing, as a result of artificial intelligence,” said Jensen Huang, founder and CEO of the semiconductor company.
In his keynote at the Consumer Electronics Show, Huang laid out what he called a “blueprint” for the future, that includes a big bet on self-driving cars, a new generation of chips to enable more AI development and a stronger focus on open models that will allow more users to adapt AI for their operational goals.
While large language models continue to get better, the rise of agentic models, which started in 2024 and gained speed in 2025, is only adding to the momentum of AI, said Huang. Users can now train agents to combine different LLMs and functions to complete tasks and build applications that can automate fulfilling those tasks, helping users leverage technology more effectively, he said.
AI models are improving apace, adding reasoning capabilities so they can solve problems and reinforce their learning, said Huang. AI is making the development of applications faster and more accessible, so users can build agents and applications that can work with enterprise and customer service platforms. Huang mentioned companies such as ServiceNow and Snowflake are leveraging AI to accelerate data processing and solve problems.
“The agentic system is the interface. It’s no longer Excel with a bunch of squares that you enter information into. It’s no longer just the command line,” said Huang. “The way you interact with your platform is much more – well, if you will – simple, like you’re interacting with people.”
At CES, Huang introduced Rubin, a new computing platform that promises to speed AI development and shrink its cost to as much as one-tenth for some functions. “The faster you train AI models, the faster you can get the next frontier out to the world,” he said.
Open models and physical AI
Much of the recent AI adoption has been helped along by the availability of open models such as last year’s release of Deep Seek, Huang said. “Open models have really revolutionized artificial intelligence last year. This entire industry is going to be reshaped as a result of that,” he said.
Every six months, a new AI model is emerging and “the number of downloads is growing so fast because startups want to participate in the AI revolution. Large companies want to, researchers want to, students want to. Just about every single country wants to. How is it possible that intelligence, the digital form of intelligence, will leave anyone behind?” asked Huang.
Nvidia is not only opening access to its AI models, said Huang, the company is opening access to the data used to train them, because “only in that way can you truly trust how the models came to be.”
Nvidia is also making a bet on physical AI, systems that can learn the laws to the physical world and react to them. This was a critical development to apply AI to uses such as self-driving cars and industrial AI use.
“There’s no question in my mind now that this is going to be one of the largest robotics industries,” said Huang.
Huang announced the launch of Alpamayo, an AI model to run autonomous driving vehicles. He also announced a partnership with Mercedes-Benz to equip a Mercedes CLA model that will ship this year. Soon, nearly every car on the road will be either an autonomous vehicle—self-driving taxis, rental cars or personal vehicles—or it will have some form of AI navigation built in, said Huang.
“I think we can all agree fully here that this inflection point of going from non autonomous vehicles to autonomous vehicles is probably happening right about this time,” said Huang. “In the next 10 years I’m fairly certain a very large percentage of the world’s cars will be autonomous or highly autonomous.”
