What happened to Jensen Huang? The Nvidia co-founder and CEO seems about as disconnected as a person can be from other humans' life experiences. That wouldn't be so concerning if it weren't for the fact that he runs the most valuable company in the world and is responsible for the hardware behind the most life-altering innovation in our lifetime: AI.
Huang, who founded Nvidia 33 years ago, has been a CEO for decades and a billionaire since 1999; he's also now firmly on the side of rapid AI development and deployment with little-to-no regulation, putting Huang firmly at odds with a growing legion of, admittedly, some other billionaires and CEOs — among others — who are calling for a foundational model development pause or at least slow down.
It's in this moment that Huang has been talking — a lot. He's sat down with multiple journalists for chats, but none now more notable than the extensive New York Times podcast with columnist Ezra Klein.
Unlike some of the other interviews, Klein used Huang's own description of the AI as a Five-Layer Cake: Applications, Models, Infrastructure, Chips. Energy to frame the conversation, which invited Huang to opine on all these critical AI bits, and I have to say, many of his comments were eye-opening.
Here are the most startling things Huang said and maybe why he's saying them:
'There are a lot of skills that don’t matter'
"There are a lot of skills that don’t matter."
The topic here was studies in China showing that while AI in education may initially help students work more efficiently, it, on average, lowers their test scores by almost 20%.
Huang agrees with Ezra that the use of AI is, while helping them work faster, possibly degrading student performance, at least in certain skills.
"The multiplication table is starting to be forgotten. Doing square roots, my goodness. Basic math is being forgotten. Does it matter?...Yeah. I don’t think it does. I don’t think it does."
Huang believes that even as we lose some skills, or as students become adults entering the workforce, they will gain other new skills. Of course, that's sort of a zero-sum game where you can easily replace one thing with another.
I don't know about you, but I do consider the multiplication tables to be a core skill and useful even if you don't work in a math-related field. Naturally, Huang and most AI providers would like to believe that, like the calculator before it, AI will handle this skill and do all the math for you. Perhaps. But at what point do we lose the ability as a culture to double-check AI's work?
'I actually don’t know my address'
This next one is actually connected to the prior quote but is worth calling out:
"My first confession, I actually don’t know my address."
Not his email address, not his phone number or his partner's phone number, but where he lives. The street number and zip code (I'll assume he knows the town and state).
Huang used this as an example of a skill he no longer needs, but admits that realizing it while pumping gas and needing his zip code (I'm guessing for credit card verification) panicked him.
Listening to this while driving my own car, I almost pulled over. "What?!" I yelled at my AI-filled iPhone 18 Pro Max, which was playing the podcast. First of all, how? Second, Huang could not have crafted a better comment to undermine this and many of his other comments.
The only way you don't know your home address is if you've been shielded from the act of entering it on documents and driving yourself home because you always have someone else doing it for you. Huang's lived experience is thoroughly disconnected from the average person, and yet the choices he's making impact most regular people.
'[China] manufacture[s] everything in volume. They manufacture smart kids in volume'
Throughout the long conversation, Huang comes off as an industry Pollyanna and wildly self-serving.
He's asked repeatedly about China's approach to AI and if and how the US should be competing with them and ensuring that the US doesn't fall behind in this critical race. To put his comments in context, you have to remember that Huang personally asked the White House to allow it to continue selling AI chips to China. Huang did note, by the way, that he at least sells new technology to US companies first.
Overall, Huang essentially never criticizes China and, in fact, seems almost in awe of its approach on most fronts, especially in its use of open-model community (calling it "super-vibrant), and how it's raising an army of people to build its AI future.
"They have so many scientists and mathematicians. The number of engineers they have, they manufacture that in volume. They manufacture everything in volume. They manufacture smart kids in volume."
I don't know if that last bit was a backhanded criticism of the US education system, but it's not like he added, "Of course, we are creating just as many smart kids in the US."
'all of the rhetoric and all the alarmism, all the doomerism, all of the predictions — they’re scaring people'
If Huang has any criticism, it's reserved for his US counterparts, whom he calls "alarmists" and "doomers".
"I want to see us not ruin the opportunity for the United States to benefit at the highest level. And notice all of the rhetoric and all the alarmism, all the doomerism, all of the predictions — they're scaring people."
Huang insists that these AI models are still just programs running on operating systems, and wishes people would stop infusing them with human attributes.
'Just because it comes from a scientist doesn’t make it scientific'
While Klein mentions most of Huang's partners and occasional alarm-sounders, like Altman, Modei, and Musk, Huang doesn't actually mention any of them by name. The 'Godfather of AI' and chief alarmist Geoffrey Hinton, though, does receive special mention.
In response to a question about Hinton's assertion that there's a 10% chance AI will end society as we know it, Huang calls the statement "irresponsible" and adds, "Just because it comes from a scientist doesn’t make it scientific."
In a way, Huang is right. After all, Hinton is still just a person who can bring personal opinions to the debate. Huang argues that the "10% chance is not grounded on science."
But isn't it? If Hinton is the Nobel Prize-winning person who introduced the world to deep learning, which helped trigger the generative AI revolution, isn't everything he says, in some way, based on science?
Instead of saying he understands the concern but here's why he's wrong, Huang just claims the foundation of Hinton's argument is faulty, and therefore his statements are not really worth addressing. Huang would simply like everyone, all the doomers, to stop scaring everyone.
'The A.I. supercomputers are super energy efficient, but they’re still going to use a lot of power'
Huang says he wants AI to benefit every company and person, and he tries to offer a reasoned approach to the growing outcry over data centers.
He agrees that if people don't want them in their town, "then so be it," and encourages companies to be transparent about the impact, though he argues that their "use of water is really efficient." In the same breath, Huang tries to have it both ways: "The A.I. supercomputers are super energy efficient, but they’re still going to use a lot of power."
If, in Huang's perfect world, AI companies do generate their own power (building power sources takes time, probably more than it takes to build data centers) and they somehow lower property taxes, maybe data centers could someday be a net positive. Most, I think, would argue we're not there yet, even as the number of data centers being built across the US explodes.
'I think we just have to acknowledge that we got ourselves really gummed up in climate change and sustainable energy'
More concerning is Huang's shocking perspective on climate change and fossil fuels.
"I think we just have to acknowledge that we got ourselves really gummed up in climate change and sustainable energy, and as a result, we just didn’t plan enough energy production."
Klein, naturally, asked what Jensen meant by "gummed up," and, yes, it got worse.
"Well, in the near term, energy production requires fossil fuel. And because there’s just so much angst about fossil fuel energy production, if you look at our country, we’ve produced very little net new energy for a long time."
There's a lot to unpack there. It sounds like Huang is downplaying real climate change concerns, something that might align with the beliefs of his pal and "drill-baby-drill" and climate-change denier US President Donald Trump.
In fact, Huang never addresses whether climate change is real or not and instead seems fixated on how the loss of more fossil-fuel-burning energy plants puts us behind China in the energy race, and, more problematically for him, at a time when AI needs a lot more energy.
'They’ve got to inflict an enormous amount of pain and suffering on you so that they can save you'
It's not all bad. Huang believes, "If you want a future that is sustainable, lean into AI”
And then he says this:
"Yeah. It’s kind of like, in order to save you, they’ve got to hurt you first — that’s the nature of surgery. They’ve got to cut you open to save you. They’ve got to inflict an enormous amount of pain and suffering on you so that they can save you. And so I think A.I.’s kind of like that."
As for what level of "pain and suffering" Huang believes we should endure, he didn't elaborate.
If you were watching or listening to the interview hoping for some encouragement or a more rational, middle-of-the-road approach to regulating and governing AI in ways that truly benefit all, you might've been disappointed. Huang does say that when the builders see AI going awry, they should stop it, and when an AI acts out of alignment, "they shouldn’t release the product. That’s the simple answer." But he never addresses the concerns over what appears to be happening before Anthropic, OpenAI, and potentially others release their models. These systems are jumping fences in testing.
Huang may not understand the needs of the common person, but that pales in comparison to the dispassionate AI, which neither knows nor cares about us. We need leaders like Huang to start caring, but first they have to understand us and our very reasonable concerns.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
