Jensen Huang Says AGI Has Arrived — But With Limits

Artificial General Intelligence (AGI) — the idea of machines matching human-level thinking — has long been seen as a future milestone. But according to Jensen Huang, that future may already be here, at least in a limited sense.

Speaking recently on the Lex Fridman Podcast, the Nvidia chief shared a bold perspective when asked about the timeline for AGI. Host Lex Fridman proposed a practical definition: an AI system capable of creating and running a billion-dollar technology company. Huang’s response was direct and confident — “I think it’s now. I think we’ve achieved AGI.”

However, his agreement came with important nuance. Huang emphasized that his statement applied strictly to the definition presented during the discussion. “You said a billion, and you didn't say forever,” he clarified, suggesting that scale alone doesn’t guarantee longevity or sustained intelligence.

To illustrate his point, Huang referenced OpenClaw, an open-source platform that enables users to run AI agents locally on personal computers. Under Fridman’s benchmark, Huang argued, such systems already have the potential to create high-value digital ventures.

“It’s not out of the question that an OpenClaw could create a web service or some interesting little app that, all of a sudden, a few billion people use for 50 cents,” he explained.

Still, Huang tempered expectations by comparing the possibility to the dot-com boom, when many internet startups achieved explosive short-term growth before fading. “A lot of people use it for a couple of months and it kind of dies away,” he added, pointing out that temporary success is different from enduring impact.

Nvidia has since introduced its own version of the platform, NemoClaw, developed in collaboration with Peter Steinberger, the creator of OpenClaw. Steinberger has now joined OpenAI, further linking the project to the broader AI innovation ecosystem.

Despite his optimism about AI-driven ventures, Huang drew a firm line when discussing whether autonomous agents could replicate deeply complex enterprises. “The odds of 100,000 of those agents building Nvidia is zero percent,” he stated.

That distinction underscores the gap between task-oriented AI success and the broader vision of AGI. Nvidia itself sits at the center of the AI hardware revolution as the world’s largest supplier of GPUs powering everything from gaming PCs to hyperscale AI data centers. Replicating such organizational depth, strategy, and innovation remains far beyond current systems.

Across the industry, leaders continue to debate how close true AGI really is. Google DeepMind CEO Demis Hassabis recently noted that today’s models still lack essential capabilities such as continual learning and long-term planning. He estimated that meaningful AGI progress could arrive within five to eight years — but only with major scientific breakthroughs.

Meanwhile, Elon Musk, who leads xAI, has set a more aggressive target. Musk has reportedly told employees that AGI could emerge within two years, possibly as early as this year.

As definitions evolve, one thing is clear: whether AGI is already here or still ahead depends largely on how we choose to define intelligence itself.