Jensen Huang: 33 Years of NVIDIA - Pain, Suffering, and Corporate Character
What 33 Years of Building NVIDIA Actually Taught Jensen Huang
This isn’t a typical tech CEO interview. Jodi Shelton, who first met Jensen in 1993-94 when NVIDIA was operating out of a former acupuncture parlor in Sunnyvale, gets unusually personal insights from the man who built the company now powering the AI revolution.
On what separates agents from workflows… wait, companies from each other: “An agent has dynamic control flow devised by the LLM at runtime, whereas a workflow is predefined coded graphs” - no wait, that was the other video. Jensen’s version: “Somehow there’s a magic in the corporate culture, the corporate character. How teams come together during adversity.”
On NVIDIA’s actual secret sauce: “Pain and suffering - that’s our secret sauce. Come work with me. That’s my gift.” Not algorithms. Not GPUs. Character forged through hardship. “Getting Grace Blackwell into production almost broke our company’s back, but we wouldn’t let it. That’s 100% character, not intelligence, not hard work.”
On why NVIDIA has 61 CEOs: Jensen has nearly 60 direct reports. “I have 60 people who could be world-class CEOs for many other companies. Every single decision I’ve made, I’ve reasoned through it in front of them. Nvidia has 61 CEOs.”
The Philosophy of Building That Cannot Be Hired
On inventing vs. innovating: “There are countless companies and researchers who have created technology and they say things like ‘I did that before’ or ‘I thought of that.’ All these great inventors didn’t also have great product inventors - people who invent products to carry technology to market, then invent strategies, then invent ecosystems and markets.”
NVIDIA’s edge: inventing technology → inventing products → inventing strategies → inventing ecosystems. Repeatedly.
On why empty chairs are better than wrong hires: “An empty chair is better than a chair filled with the wrong person. The company will keep moving on. Whether it’s a missing CFO or VP of anything - the company will keep moving on.” He interviewed 22 CFOs before hiring Colette Kress. When she asked how long he wanted her to stay: “For as long as we shall live.”
On what he can’t interview for: “They’re all smart. They’re all competent. Find me a CFO and I promise you they’re competent. But what makes the magic of NVIDIA is the chemistry of people together. Mostly it’s corporate character - and that comes from somewhere.”
The Surprisingly Vulnerable CEO
On why he’s a “reluctant CEO”: “I like being inside the company more than outside. I like not giving speeches than giving speeches. I like not giving keynotes at all. But I’m a very enthusiastic Nvidia builder.”
On public speaking anxiety: “Public speaking scares the living daylights out of me. GTC Washington DC is two weeks away and I am deeply anxious. I’ve been deeply anxious for a month. These things wear on me. They’re always in my mind.”
On company meetings being worse than earnings: “During earnings week, people think that earnings is stressful to me. Not even a little bit. The company meeting stresses me out.”
On the Next Five Years of AI
The problem-shrinking thesis: When computers become massively faster, every problem looks smaller. “We took all of the internet data and gave it to a computer - because the world’s internet data looks so small now. These days, all the world’s internet data looks tiny to us.”
His counterintuitive prediction: “Instead of fewer jobs, I feel what’s likely to happen is we’re going to be busier than ever. We’re going to think of more ideas of things we can solve now that we didn’t used to be able to solve. All those things off the table are now on the table.”
On AI closing the technology divide: “One of my favorite things is vibe coding. Anybody can be a software programmer now. AI is going to close the technology divide. People gifted at their craft but who don’t know how to scale with technology now have AI to help scale them.”
The Loveable CEO told him about people making $2-3M/year from businesses built entirely on AI-generated software. “They’re welcomed into the world’s economy, not burdened by technology anymore.”
The Moravec Paradox of Corporate Culture
On why 100% of jobs will change (not 50% lost): “It’s very likely that 100% of people who don’t have jobs today because of AI can make a living. 100% of jobs will change rather than 50% will be lost.”
On working with superintelligent systems: “I’m surrounded by 60 people who are all better at what they do than I am. In a lot of ways, they’re artificial superintelligent relative to me in their field. And I’ve got no trouble working with all of them.”
6 Takeaways from Jensen Huang on Building Enduring Companies
- Character over intelligence - What separates great companies is how teams handle adversity, not how smart individuals are
- Empty chairs > wrong hires - Wait for the right person; the company keeps moving
- 61 CEOs, not 1 - Reason through every decision in front of your team
- Invent the ecosystem - Technology → Product → Strategy → Market creation
- Vulnerability is strength - CEOs are “at the mercy of almost everybody”
- Ignorance is a superpower - “If I had known everything then that I know now, I would never have done it”
What This Means for AI-Powered Organizations
Jensen’s philosophy perfectly mirrors the agent vs. workflow debate: NVIDIA succeeds not because of rigid processes, but because of character that enables dynamic response to challenges. The company that “almost broke its back” shipping Grace Blackwell is the same company that can pivot when circumstances demand.
His prediction that we’ll be “busier than ever” because AI shrinks problems is already playing out. When every hard problem looks simple, you attempt more problems. When every experiment becomes cheap to run, you run every experiment.
The most counterintuitive insight: the man who built the computing infrastructure for AI is most excited about the human elements AI can’t replicate - character, suffering through adversity together, the magic of teams under pressure.