Some AI industry leaders, such as OpenAI CEO Sam Altman, are obsessed with artificial general intelligence, or AGI. While the term has no universally agreed-upon definition, it generally refers to AI that can match or even surpass human cognitive abilities across a large range of tasks.
In a recent interview, Altman shared that OpenAI believes it will achieve AGI by the end of the year.
However, according to Nvidia CEO Jensen Huang, AI has, in some respects, already achieved AGI — and it’s really no big deal.
Nvidia has become the darling of the tech industry in recent years. The company is an invaluable part of the AI industry because it produces many of the high-performance chips that power the technology.
During an earnings call on Aug. 26, Huang was asked about AGI in light of OpenAI’s recent claims.
“For many tasks, we could say that we’ve already achieved AGI,” Huang said. “I think of all of those milestones…they’re kind of senseless at this point.”
Huang said the focus should be on whether AI is “doing productive and useful work.” From a company perspective, Huang added that the concern should be about “generating profitable tokens.”
It’s unclear exactly what Huang meant when saying Nvidia has already achieved AGI. He pointed to AI’s evolution beyond simply responding to prompts, arguing that agents can now reflect on their performance, learn new skills, and improve their future output.
This is not the first time the Nvidia CEO has claimed the company has achieved AGI.
In a March 2026 interview, podcaster Lex Fridman asked whether an AI that could start, build, and run a billion-dollar company would be achievable within the next 20 years.
“I think it’s now. I think we’ve achieved AGI,” Huang replied.
However, Huang then qualified that claim, saying AI might be able to create a billion-dollar viral app but could not build a company like Nvidia. “The odds of 100,000 of those agents building Nvidia,” he added, “is zero percent.”
Source: https://mashable.com/tech/nvidia-ceo-jensen-huang-says-agi-achieved-is-meaningless
