Anyone following the artificial-intelligence landscape closely may have noticed that the CEOs of the leading AI labs, OpenAI's Sam Altman and Anthropic's Dario Amodei, don't speak like most CEOs.
While they race each other to advance the frontier of AI models, they've also asked to be regulated-including comments on Wednesday before the United Nations Security Council-lest AI models get better faster than they know how to control them.
The continuing drip of news regarding the hacking sprees executed autonomously by AI models this spring and summer, along with support from other tech leaders like Bill Gates, adds to the urgency of their message.
But Nvidia CEO Jensen Huang may not want to hear about an AI slowdown, because that could also slow down the company's AI chip sales, which Wall Street projects will total $652 billion next year.
Like many others, he's expressed some confusion about Altman and Amodei's words versus their actions.
"If they believe they're out of control, then the right answer is don't ship products until they're in control," Huang said in an interview with the New York Times' Ezra Klein published on Wednesday.
"Nobody's building more compute today than the people asking to be slowed down," he said. "It strikes me odd."
Huang's solution, perhaps unsurprisingly, is that the companies need to spend even more on Nvidia chips. "They have to shift their total [research and development] from just capability to a lot on verification, evaluation, and testing," he told Klein. "I wouldn't be surprised if the amount of compute necessary to develop these models increased by a factor of 10 because the evaluation is so rigorous."
OpenAI and Anthropic did not immediately respond to Barron's request for comment.
Huang's comments are notable: The CEO of the most valuable company in the world told two of its best customers to stop whining and spend more on Nvidia chips.
President Donald Trump has indicated that he is on Huang's side in this fight through repeated public statements over the past few weeks. Cabinet members, such as Treasury Secretary Scott Bessent, have been singing from the same hymn book. During this past week's U.S.-China state dinner, Huang had a place of honor at the same table with Trump and China's leader Xi Jinping. China's public statements have also conveyed little interest in slowing down AI progress.
The AI labs have been looking into self-regulation. In mid-September, OpenAI chief global affairs officer Chris Lahane revealed that the company was negotiating with Anthropic and Alphabet to set up an industry body to set AI safety standards and perform testing. Maybe they are listening to Huang's comments.
"These are CEOs with agency," said Huang in the interview. "They could absolutely take care of the situation."