Jensen Huang Says Nvidia Achieved AGI, Again — Not That It Matters

Asked about OpenAI’s pursuit of AGI, Huang said “for many tasks, we could say that we’ve already achieved AGI.” He did not provide a precise definition or benchmark, adding that he thought “all of those milestones and all those, you know, they’re kind of senseless at this point.” The CEO said the real measure of AI is whether it is “doing productive and useful work” and “generating profitable tokens,” with more compute producing more tokens and, inevitably, more profit.

This is not the first time Huang has declared AGI achieved. In March, on the Lex Fridman podcast, he said plainly, “I think we’ve achieved AGI.” After Fridman proposed a definition involving starting a tech company worth more than $1 billion, Huang walked back, saying the odds of 100,000 such agents building Nvidia were zero percent.

AGI has no consensus definition. OpenAI, founded with the explicit goal of building AGI, defines it in its charter as “highly autonomous systems that outperform humans at most economically valuable work.” Altman has acknowledged this is not a measurable standard, calling AGI “not a super useful term.” Separately, OpenAI and Microsoft have reportedly defined AGI as systems that can generate at least $100 billion in profits.

In a recent Time magazine story, OpenAI chief research officer Mark Chen estimated the company was “80% of the way” to AGI, while Altman said that by the end of the year OpenAI would have something he would call AGI.

Other industry leaders have also noted the term’s fuzziness. Anthropic Chief Executive Dario Amodei has called AGI “imprecise” and even a “marketing term,” preferring “powerful AI.” Meta speaks of “personal superintelligence,” Microsoft of “humanist superintelligence,” and Amazon of “useful general intelligence.” Google DeepMind’s Demis Hassabis has said we are in the “foothills of the singularity.” OpenAI cofounder Ilya Sutskever, who once led employees in chants of “feel the AGI,” now runs a company called Safe Superintelligence.

New terminology has not made the concept more tangible. As long as AGI remains poorly defined and carelessly used, declaring it achieved is arbitrary. But with no agreed-upon finish line, the term remains a handy way to hype progress, so industry executives including Huang are likely to keep using it.

Editor’s Summary: Nvidia CEO Jensen Huang said on an earnings call that Nvidia has already achieved AGI for many tasks, but immediately called such milestones meaningless. The report underscores the lack of consensus on what AGI means and notes that Huang had previously made a similar claim in March. Other tech leaders continue to use their own vague terms for superhuman AI.