The former dish washer is making tech that’s changing the world with a management style akin to ‘torture’. Even a biographer had a taste of his wrath
Bryan Catanzaro stuck out at Nvidia. In the hive of engineering drones at the company’s Californian headquarters, he was the dreamer. His hair was long and he dressed like a jester in statement glasses and loud, tacky shirts. He was patient and kind, and he spoke in a soothing, gentle voice. He was the only Nvidia engineer I met who had a humanities degree.
In 2013, having worked at the company for a couple of years, Catanzaro was struggling. His big idea was to create a software library that would allow neural networks, intelligent computer systems modelled on the human brain, to train faster and more effectively. However, when he presented his prototype to Nvidia’s software team early that year, they panned it.
Catanzaro decided to make his case directly to Jensen Huang, the Taiwanese-American chief executive who had built Nvidia from scratch. At the time, Nvidia was still primarily a graphics company making the chips that powered shoot-’em-up computer games such as Quake and Call of Duty. Neural networks did not seem to be on Huang’s dashboard.
To Catanzaro’s surprise, Huang was immediately intrigued. Following their first meeting, Huang cleared his schedule and spent an entire weekend reading about artificial intelligence, a subject about which he knew little. Another meeting soon followed, in which Catanzaro was amazed to find that his boss now knew as much, perhaps more, about neural nets as he did.
By the middle of 2013, Huang was vibrating with wild‑eyed, resonant intensity. He called Catanzaro into the conference room that he was using as an office and told him that he considered his proposal to be the single most important project in his company’s 20-year history. The whiteboard on the wall had been cleared of diagrams; in its place, Huang had written the cryptic acronym “OIALO”. This, Huang said, stood for “Once in a Lifetime Opportunity”.
It sometimes took Huang a while to warm up to ideas but with AI he had experienced a damascene epiphany. “He got it immediately, before anybody,” said David Kirk, Nvidia’s former chief scientist. “He was the first to see what it could be. He really was the first.”
Huang concluded that neural networks would revolutionise society and that he could use Nvidia’s parallel computing model to corner the market on the necessary hardware. He announced that he was betting the company. “He sent out an email on Friday evening saying everything is going to deep learning, and that we were no longer a graphics company,” Greg Estes, a vice-president at Nvidia, recalled. “By Monday morning, we were an AI company. Literally, it was that fast.”
As Nvidia moved into AI, Huang abandoned his hobbies. His appetite for mischief diminished, he stopped practising table tennis and his beloved teppan grill went cold. He even stopped returning texts. “He was just so, so focused on work,” recalled Jens Horstmann, his closest friend. “It was all he could talk about.”
From the day Huang had started his career at the age of 20, he had worked relentlessly, putting in consecutive 12-hour days, six days a week for three decades. Now past 50, and with his children grown, he began to work even harder. “OIALO” was repeated at every meeting.
In the past, to relax, Huang had enjoyed going to the cinema by himself. He preferred big-budget popcorn flicks, particularly the Avengers films, which he watched as much to grade the execution of the CGI as for the story. However, he soon became tired of audience members pestering him for selfies. Eventually, he stopped going. From 2014, there was only work. There was only AI.
The wrath of Huang
A decade later, Nvidia is one of the world’s most valuable companies, worth more than $3 trillion. Its market capitalisation purportedly represents the expected value of the company’s future earnings but it is really a giant, GDP-sized bet on the capabilities of one 62-year-old man.
Nvidia’s value also reflects the extraordinary speed of its computer chips. Using graphics processing technology originally designed to optimise games such as Grand Theft Auto, Nvidia’s chips work at such remarkable speed that every other big technology company, from Meta and Google to Tesla and Apple, now uses them to power its AI.
Neural networks form the basis for the chatbots and large language models that have brought AI into the mainstream, which means Huang’s wager has paid off spectacularly. The Taiwanese boy who arrived in America aged nine, and was accidentally sent by his uncle to a Baptist academy for troubled children in Kentucky, has come an extraordinarily long way. The company he founded at a Denny’s fast-food outlet in 1993, the same chain where he had worked as a dishwasher and a busboy for pocket money, has made him one of the wealthiest men alive.
To the sanctified roster of Jobs, Bezos, Gates and Zuckerberg, one can now add the name Huang. “I’m very rich now,” he said last year. “Do you know how rich I am?” It wasn’t a boast. I asked him whether he had a plan for what to do with the money. “I have no idea,” he said. “None.”
Nvidia’s success has also made many of its employees multimillionaires. Not all of them have been able to stick the course, however. “Interacting with Jensen is like sticking your finger in the electrical socket,” Catanzaro said. “Jensen is not an easy person to get along with all the time. I’ve been afraid of Jensen sometimes. But I also know that he loves me.”
When Huang explains his point of view for the first time, he does so in a measured voice, moving from premises to arguments to conclusions. His interlocutor then has two options: agree with his line of thinking or risk detonation. Those who contradict him are often shocked when he explodes in fury. This is the wrath of Huang.
Spectators are important to Huang — when he dresses down an employee, he usually does so in public so others can learn from the experience. (“Failure must be shared,” he once said.) If a project is delayed, Huang commands that the person responsible stand up and explain to an audience, in detail, every single thing that has gone wrong. Huang then delivers a withering analysis of their performance. Such corporate struggle sessions are not for everyone. “You can kind of see right away who is going to last here, and who is not,” says Dwight Dierck, a senior Nvidia executive. “If someone starts getting defensive, you just know that person won’t be long at Nvidia.”
One former employee recalled a time when he bungled a minor assignment. Huang confronted him in front of three dozen executives, asking him how long he had been with the company and what his salary was. The employee sheepishly provided the numbers. In his head, Huang calculated the employee’s career compensation and asked for all of it to be refunded. The exercise didn’t feel like a joke. “He was kind of serious,” the employee said. “I practically didn’t sleep for three weeks.”
Kirk believed that Huang’s outbursts were purposeful. “Yelling at people was part of this motivational strategy,” he said. “You might think he’s just mad but I think it was premeditated. And it works. It annoys people but it does work. When he’s torturing people, he’s forcing them to learn a lesson — and they certainly would never forget it.”
His anger wasn’t always confined to the office. Horstmann recalled attending one family gathering when Huang botched a complex dish that he was preparing. Standing in his bespoke kitchen, with his daughter, Madison, who had trained at Le Cordon Bleu, standing nearby, Huang exploded and began to scream at his inadequate equipment. “I think we all understood we had to get out of the kitchen,” Horstmann said. “It was just time for Jensen to yell at his stove.”
Yet even those who disliked his managerial tactics often had positive things to say about Huang personally. Speaking to more than 100 former and current Nvidia employees, almost all had a tender story about him to relate. One, the same person Huang had humiliated in front of dozens of people asking for a refund of his salary, said that when he was diagnosed with a serious medical issue, Huang offered to pay in full for his treatment.
Huang’s combination of love, fear and guilt was a seductive and powerful motivator. “You felt like you couldn’t let him down,” said Sharon Clay, a senior Nvidia engineer. “You just couldn’t.”
‘This is not Star Trek’
The loyalty and determination of Huang’s team has helped him to build the Manhattan Project of computer science, an AI machine that might eliminate art, become self-aware, perhaps even destroy humanity as we know it.
I first interviewed Huang in 2023 and in every conversation since I had pressed him about the wisdom of unleashing such power. He gave me variations on the same answer every time. I brought up the concerns of Geoffrey Hinton, the world’s leading AI scientist, that “humanity is just a passing phase in the evolution of intelligence”. Huang scoffed: “A lot of researchers don’t understand why he’s saying that. Maybe it’s bringing attention to his own work.”
I was stunned by this statement. This wasn’t some guy on the street waving a sign; this was the greatest mind in AI telling us we should be very, very worried. Huang was coldly dismissive. “Look, you buy a hot dog, so the machine recommends you ketchup and mustard,” Huang said. “Is that the end of humanity?”
Huang insisted AI was safe. He had his own incentives, of course — money but also pride. Having waged a long, lonely battle to make parallel computing profitable, he was now determined to enjoy its rewards. He seemed offended by technologists such as Elon Musk and Sam Altman, who performatively decried the risks of AI while continuing to purchase his hardware.
Huang’s confidence was also born out of the many ways in which AI wasn’t like the biological brain. The neural net did not have a hypothalamus, a hippocampus or a pineal gland. The neural net was unaffected by hormones, had no sex drive, didn’t go to sleep, couldn’t love its children and never had a dream. A well‑trained model mimicked, and sometimes exceeded, our own capacity for cogitation but it had no sense of memory, no emotions, no imagination, no reproductive function and no instinct to survive. In this sense, one could not say it was really alive.
But I couldn’t shake my anxiety. Was Huang, seduced by the power of what he had built, gambling with the future of our species? I had to know. I had to ask him one last time.
My final interview with Huang took place on a Friday in March last year at Nvidia’s Endeavor headquarters in Santa Clara, just after the company’s annual conference.
Huang, after four days of nonstop presentations, interviews and technical demos, was visibly exhausted. People were becoming a little critical of him, I noticed. At a recent press conference, one journalist had asked whether he was the new Robert Oppenheimer. “Oppenheimer was building a bomb,” Huang replied. “We’re not doing that.”
Our interview started amicably enough. I congratulated Huang on his fine performance at the keynote, then asked him about life in Taiwan, about his family, about how he was handling his newfound fame. His answers were curt and to the point.
Then I asked him specifically what new jobs might be created — as opposed to destroyed — by AI. I watched the blood drain from his face. I had hit the tripwire.
“Is it going to … is it going to destroy jobs?” Huang asked, his voice crescendoing with anger. “Are calculators going to destroy math? That conversation is so old and I’m so, so tired of it.
“I don’t want to talk about it any more. It’s the same conversation over and over and over and over and over again. We invented agriculture and then made the marginal cost of producing food zero. It was good for society. We manufactured electricity at scale and it caused the marginal cost of chopping down trees, lighting fires, carrying fires and torches around to approximately zero, and we went off to do something else. And then we made the marginal cost of doing calculations — long division. We made it zero.”
He was yelling now. “We make the marginal cost of things zero, generation after generation after generation, and this exact conversation happens every single time.”
I tried to switch subjects but it was no use. His anger was tinged with disgust. It seemed uncontained, omnidirectional and wildly inappropriate. “This cannot be a ridiculous sci-fi story,” he said. He gestured to his frozen public relations representatives at the end of the table. “Do you guys understand? I didn’t grow up on a bunch of sci-fi stories and this is not a sci-fi movie. These are serious people doing serious work. This is not a freaking joke. This is not a repeat of Arthur C Clarke. I didn’t read his f***ing books. I don’t care about those books. It’s not — we’re not a sci-fi, repeat. This company is not a manifestation of Star Trek. We are not doing those things. We are serious people doing serious work.”
For the next 20 minutes, in a tone that alternated between accusatory, exasperated and belittling, he questioned my professionalism, questioned my interview approach, questioned my dedication to the project.
I left the interview bewildered. I had been in plenty of tense conversations with executives but I had never had someone explode at me in this way. I was stunned — but also, if I’m honest, I was a little giddy. To be targeted by the wrath of Huang was in a certain sense an honour: a rite of passage that everyone who gained admittance to his inner circle underwent. Walking away from the conference room, I turned to one of the PR reps. “That went well,” I said. He laughed. “Oh, that?” he said. “That was nothing.”
The World According to Jensen
Work like you’re going bankrupt
Huang encourages his employees to constantly behave as if the company was teetering on the verge of bankruptcy, even when it is making massive profits. For many years, Jensen opened staff presentations with the words “Our company is thirty days from going out of business.” Even today at Nvidia, this sentence remains the corporate mantra.
Find a uniform
Every time I saw Jensen, he was wearing the same all‑black T‑shirt with a Thomas Burberry monogram. He bought twenty four of the shirts in 2020 and rotated them every day for four years.
Work at the speed of light
Huang employed a scheduling technique he calls the “speed of light” and drilled this management concept into his employees with the fervor of religious doctrine. “Speed of light” does not mean, as one might assume, to move quickly. Instead, Huang encouraged managers to identify the absolute fastest that something could conceivably be accomplished, given an unlimited budget, and assuming that every single thing went right. (For example, traveling from New York to London at the “speed of light” would involve perfect weather, zero traffic, and a supersonic plane.)
Managers could then work backward from this unachievable constant to realistic but still impressive delivery times. Huang pursues this unattainable ideal every day of his life. “I should make sure that I’m sufficiently exhausted from working that no one can keep me up at night,” he said. “That’s really the only thing I can control.”
Always see imperfections
When Nvidia’s new headquarters opened, the architect Hao Ko led Huang and a group of executives on a tour. “The place was finished, it looks amazing, we’re doing the tour, and
he’s questioning me about the placement of the water fountains,” Ko said. “He was upset because they were next to the bathrooms! That’s required by code, and this is a billion‑ dollar building! But he just couldn’t let it go.” “I’m never satisfied,” Huang said. “No matter what it is, I only see imperfections.”
20,000 emails a week
Sometime around 2020, Huang asked everyone at the company to submit a weekly list of the five most important things they were working on. Every Friday from that day forward, he received twenty thousand emails.
Brevity was encouraged; Huang would randomly sample from this pool of correspondence late into the night. In turn, he communicated to his staff by writing hundreds of emails per day, often only a few words long. (One executive compared the emails to haiku. Another compared them to ransom notes.) His responsiveness was superhuman. “You’d email him at 2 a.m. and receive a reply at 2:05 a.m.,” Dally said. “Then you’d email him again at 6a.m. and receive a reply at 6:05 a.m.”
Do the hardest thing first
By 5am, Jensen was usually up and working. He always began his workday with his most important long‑term project, figuring that as long as he addressed it, the day couldn’t be considered a bust no matter what else happened.
Never settle for less than perfect
In the early 2000s, Huang bought a plot of land in Los Altos Hills, overlooking Silicon Valley.
There he commissioned a six‑thousand‑ square‑ foot mansion with five bedrooms, seven bathrooms, a swimming pool, and an oversized garage. He bought two Ferraris (his and hers) and started collecting expensive whiskey.
But not even a custom dream house could meet Jensen’s standards. Returning home from work one day, he noticed that the glass doors to the garden of his mansion did not perfectly align with the view to the pool house out back. The lack of symmetry annoyed him, so he ordered the pool house uprooted and, at considerable expense, had it moved 18ft to the side.
Don’t share the load
Management professors theorise that a chief executive should ideally have between eight and twelve direct reports. Huang had 55.
The Thinking Machine: Jensen Huang, Nvidia and the World’s Most Coveted Microchip by Stephen Witt is published by Bodley Head, £25
https://www.thetimes.com/business-money/technology/article/nvidia-ceo-jensen-huang-qsw6ldnhh
