Nvidia's Jensen Huang says AI should go as fast as we can
In an interview with CBS News, Nvidia's chief executive rejects calls to slow AI development - while insisting his company would never ship products that are unsafe.
Nvidia's Jensen Huang has told CBS News that artificial intelligence should be built as quickly as anyone can build it, pushing back on the industry leaders who have spent the past fortnight arguing for a slower pace. Asked whether AI development needs to slow down, Huang said: "We should go as fast as we can irrespective of anybody else."
He drew his one line at shipping unfinished work. "We're going to go as fast as we can, but we would never ever, and never should, ship products before they're ready, deliver products that are unsafe," Huang told CBS News senior business and technology correspondent Jo Ling Kent. "And nobody's expecting us to do that. But everybody's expecting us to succeed and help America be as prosperous as possible." The interview airs on Sunday Morning on 20 September.
Who he is arguing with
The question was not asked in a vacuum. CBS News reports that the debate was spurred in part by Jacob Coxon, an Anthropic researcher who resigned publicly earlier in September, saying AI companies are rushing to innovate at the expense of proper security protocols. Coxon told the network that without intervention the technology could one day pose a threat to humanity.
Anthropic chief executive Dario Amodei went further on the same beat, telling Kent that "for too long the industry lied to people about the fact that this technology had risks" and that the current rate of progress is "a warning sign that we need to slow down" - though he added: "It doesn't mean we need to panic today. It doesn't mean we need to shut it all down." The Trump administration has been generally dismissive of those fears, and a CBS News poll published the same week found Americans want to slow AI development down without stopping it.
Nvidia arrives at the argument from a position of unusual strength: CBS News describes it as the world's most valuable company, with a market capitalisation above $5 trillion, a valuation built on the same build-out that a slowdown would idle.
Our opinion
"As fast as we can, irrespective of anybody else" is the most honest sentence anyone has said about AI in weeks, and Huang probably did not mean it as a confession. Every firm in this race says it will not ship something unsafe; the disagreement is never about the standard, it is about who holds the pen when the deadline arrives. Huang holds that pen, and he is also the man selling the shovels, which makes him the least disinterested voice in the room. The $5 trillion figure makes the point for him: an industry that paused would have a lot of very expensive silicon sitting idle.
That does not make him wrong about everything. Slowing a technology nobody understands is not automatically safer, and the public's own answer in the CBS poll - slow it down, do not stop it - is roughly the position Huang is defending with much louder adjectives. But there is a difference between a considered wager and a race, and "irrespective of anybody else" is the language of the second one. The people who built these systems are the ones resigning over how they are being built; that is not proof the world ends in 2030, but it is a poor moment to declare the argument settled.