Nvidia CEO Jensen Huang Says He Thinks Artificial General Intelligence Is Here

Last updated: 2026/03/24 at 7:53 AM

News Room Published 24 March 2026

Artificial general intelligence (AGI) has become the ultimate goal for tech CEOs in recent years, and Nvidia’s CEO Jensen Huang claims the industry may have already reached it. However, it all depends on your own definition of AGI.

In an interview on the Lex Fridman podcast, the two discuss the path to AGI, with Fridman defining the term as a tool to “essentially do your job,” specifically referring to Huang’s role as a tech CEO.

He’s suggesting that AGI could be defined by an AI tool launching and running a successful tech brand, with a caveat that the tool would need to make a company worth more than a billion dollars for it to count.

Fridman asks Huang whether he believes it would take five or perhaps 20 years to reach that level of capability. Huang responds, “I think it’s now. I think we’ve achieved AGI.”

Huang points to how he believes a hypothetical situation would be possible where a modern autonomous AI could “create a web service, some interesting little app that all of a sudden a few billion people used for 50 cents, and then it went out of business again shortly after.”

Huang says, “I wouldn’t be surprised if some social thing happened or somebody created a digital influencer, super, super cute, or some social application that, you know, feeds your little Tamagotchi or something like that, and it becomes an out of the blue an instant success. A lot of people use it for a couple of months, and it kind of dies away.”

“Now, the odds of 100,000 of those agents building Nvidia is zero percent.”

Recommended by Our Editors

The problem with AGI is that there isn’t a clear definition of when a tool would have reached artificial general intelligence, and many CEOs and other tech speakers have different views of what it could mean.

Other definitions of AGI, including PCMag’s own, deem it “a machine intelligence that is equal to or greater than that of a human being.” Others include the caveat that it must be equal to or greater than a human being in all cognitive tasks, a condition that Fridman’s definition didn’t specify.