AI capable of matching or surpassing human intelligence remains a major goal for the tech industry. But to Nvidia CEO Jensen Huang, this artificial general intelligence (AGI) is a milestone that’s become irrelevant.
“For many tasks, we could say that we have already achieved AGI. I think all of those milestones...are kind of senseless at this point,” Huang said in an earnings call on Wednesday.
In contrast, OpenAI CEO Sam Altman tells Time* *in a new interview that his company will develop AGI internally by the end of this year. In OpenAI’s case, the company defines AGI as “highly autonomous systems that outperform humans at most economically valuable work.”
Huang didn’t define AGI on the earnings call. But he said it’s no longer people just prompting AI to execute tasks; more AI programs, called agents, can already run autonomously and improve “recursively," by running the task over and over again.
“I think the most important thing that matters for the industry is that one, AI is doing productive and useful work,” he said. “Two, AI is generating profitable tokens. And three, if we had more compute, we could generate more profitable tokens, which results in more profit for all of the services. This is the exact phase where we are at.”
The focus on productivity and profit might be in response to growing concern about an AI bubble amid unprecedented investment in AI data centers, even though the two leading players, OpenAI and Anthropic, still need to prove they’re profitable. Other critics say large language models will never achieve AGI because they lack persistent memory, struggle to understand logic, and remain prone to hallucinating incorrect information.
In the meantime, Nvidia is raking in huge profits, generating $96.2 billion in fiscal Q2, up a staggering 106% year-over-year. The company also indicated it could grow even more, if not for the AI-driven memory shortage. Nvidia expects the shortage to remain a bottleneck at least through the early part of 2028.
"We are experiencing extreme pricing conditions in memory," Nvidia CFO Colette Kress said on the call. "The magnitude of the price increase has exceeded our prior expectations, and [it's] headed even higher into next year." That doesn't bode well for the company's graphics cards, which have already seen a recent price increase.
As for Huang, he indicated on Lex Fridman's podcast earlier this year that the tech industry had already reached AGI to a certain extent. “I think it’s now. I think we’ve achieved AGI,” he said, pointing to how today’s AI programs can create an app capable of going viral and attracting billions of users that later fizzles out. “Now, the odds of 100,000 of those agents building Nvidia [are] zero percent,” he added.