Jul 26, 2026 · 48m · y-combinator
Jensen Huang: The Mindset That Built NVIDIA · Y Combinator
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
At Y Combinator Startup School 2026, NVIDIA founder and CEO Jensen Huang joins Garry Tan for an in-depth fireside chat exploring NVIDIA's early survival stories, strategic pivots into AI, open-source principles, and advice for the next generation of founders. Huang emphasizes first-principles thinking, systems engineering, and personal resilience as the keys to building transformative technology companies.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The partners hold 15.4% of the talking time here. How this is scored →
The partners as informed peer 3.0 Guest teaching 5.8 Guest disagreement 0.4 The partners pushing back 0.0
speaking balance: gold is the partners, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 31:26 Huang rejects the AI job destruction narrative
Huang forcefully dismisses the prevailing media premise that AI eliminates jobs, calling the narrative exactly backwards and contrasting automated tasks with enduring job purpose.
Hardest push from the partners ▶ 13:30 Tan presses on organizational friction in founder modeTan challenges standard Fortune 500 corporate orthodoxy by asking how a leader can micromanage technical weeds without alienating executives and staff.
Biggest teaching moment ▶ 11:00 Huang redefines deep learning as universal function approximationHuang educates the audience on NVIDIA's foundational insight that AlexNet signaled universal function approximation, necessitating the redesign of the entire computing stack 15 years ahead of the industry.
The partners hold their own ▶ 30:03 Tan argues text manipulation is fundamental intelligenceTan demonstrates technical depth by challenging skeptics who dismiss prompt conditioning and markdown file updates, asserting that text representations constitute core intelligence.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The partners as informed peer | Guest teaching | Guest disagreement | The partners pushing back | Why |
|---|---|---|---|---|---|---|
| Event Opening and Garry Tan Introduces Jensen Huang | 1 | 5 | 0 | 0 | Tan opens the event enthusiastically and prompts Huang on the foundational lessons of NVIDIA's early days. Huang explains how starting with the wrong algorithm forced them to learn OpenGL from textbooks purchased at Fry's. | |
| Accelerating Domain-Specific Algorithms Over Mere Hardware | 4 | 6 | 1 | 0 | Tan references backstage conversations about domain-specific algorithmic acceleration. Huang elaborates that building great companies requires focusing on accelerating algorithm domains rather than just designing faster silicon chips. | |
| The Sega Partnership and $5 Million Lifesaver | 2 | 4 | 0 | 0 | Tan invites Huang to share stories about founder hardship and near-death experiences. Huang describes candidly telling Sega's CEO that NVIDIA could not deliver the contract yet asking for the five million dollars needed to survive. | |
| Spotting the AI Revolution in AlexNet | 3 | 7 | 1 | 0 | Tan asks how NVIDIA anticipated the deep learning revolution before anyone else. Huang explains recognizing AlexNet not just as an algorithm, but as a universal function approximator that required reinventing the five-layer computing stack. | |
| First Principles, Curiosity, and Founder Mode Leadership | 4 | 7 | 1 | 0 | Tan asks how Huang manages deep technical engagement without breaking organizational dynamics. Huang explains that founder mode means customizing the corporate vehicle around the founder-driver rather than conforming to conventional corporate management. | |
| Agentic AI Systems and the Controllability Breakthrough | 5 | 8 | 0 | 0 | Tan asks Huang to walk through full-stack AI development from materials to application agents. Huang delivers a detailed breakdown of systems thinking, recursive self-improvement in agents, fine-grained controllability, and thinking models for autonomous driving. | |
| Open Source AI, Sandboxes, and Words as Thoughts | 5 | 5 | 0 | 0 | Tan highlights the importance of open-source agent frameworks and argues that manipulating text constitutes intelligence. Huang validates this premise by emphasizing that words are thoughts and detailing NVIDIA's engineering support for open ecosystems. | |
| Economic Impact: Why AI Automation Creates More Jobs | 2 | 8 | 2 | 0 | Tan inquires about macroeconomic adjustments and potential job displacement from ubiquitous AI. Huang directly refutes the common narrative that automation destroys jobs, illustrating how productivity creates backlogs and net employment growth in software, radiology, and law. | |
| Physical AI and the $100 Billion Robotics Frontier | 3 | 7 | 0 | 0 | Tan asks for NVIDIA's latest projection on commercial robotics. Huang maps out physical AI and world foundation models, predicting autonomous vehicles and robotics will form their next hundred-billion-dollar market within a decade. | |
| Joining X and Celebrating Open Source Foundations | 4 | 6 | 0 | 0 | Tan welcomes Huang's public advocacy on X regarding open-weight models and asks what technical students should study. Huang credits open source for the entire modern computing stack and encourages students to master deep sciences and systems thinking. | |
| Founder Advice: Embracing Resilience and How Hard Can It Be | 3 | 6 | 0 | 0 | Tan asks what core lesson Huang would send back to his early founder self. Huang shares his mental model of overcoming fear with the question 'how hard can it be?' and cultivating daily resilience. | |
| Keynote Conclusion and Farewell Ovation | 0 | 0 | 0 | 0 | Brief wrap-up where Tan thanks Huang and the audience provides a closing ovation. |
Statements from this episode (24)
Huang: NVIDIA invented most major computer graphics breakthroughs over 25 years
“We're the world leader in modern computer graphics. We invented most of the major breakthroughs in the last 25 years.”
Jensen Huang: Starting technology matters less than confronting reality and learning
“And so long as you're able to confront the reality, so long as you are able to learn the technology itself actually doesn't matter.”
Huang: NVIDIA succeeded by augmenting CPUs for specific algorithm domains
“The big idea of the company that was spot on is that it is possible to augment the CPU to solve problems that otherwise are too difficult to solve. And molecular dynamics is one of them. Image processing is one of them. Inverse physics is another one. And so a…”
Huang: Sega contracted NVIDIA to build the console that became Dreamcast
“Sega, ah, had contracted us to build the, ah, game console after Saturn that turned out to have been Dreamcast.”
Huang: NVIDIA went public in 1999 at a $300M valuation
“When NVIDIA went public, our valuation was three hundred million dollars. Three hundred million dollars in 1999.”
Huang: NVIDIA realized AlexNet was a universal method to learn any function
“The breakthrough for us was realizing that AlexNet was not AlexNet. That AlexNet was an approach with deep, deep learning that allows you to learn any function.”
Huang: NVIDIA immediately pivoted into vision, robotics, and self-driving after AlexNet
“Almost right away, we started working on computer vision. Almost right away, we started working on robotics self-driving cars”
Huang: Founders should adapt the organization to themselves, not standard practices
“Whatever it takes to fit the car to you, whatever it takes to fit the organization to you, that's what you ought to do. And the next CEO, whatever the personality is, they can figure it out.”
Huang: 'Founder mode' can scale a company to five trillion dollars
“Founder mode could scale for 34 years. From zero to five trillion.”
Huang: Most software will be built agentically, elevating systems thinking
“In the case of software most software is going to be done agentically anyhow, so you have to be much more able to think abstractly about systems.”
Huang: Controllability is the single biggest breakthrough needed for AI agents
“And so I think controllability is probably the single biggest breakthrough that we need for agents at every single level.”
Huang: NVIDIA must design hardware 5 to 10 years ahead
“We kind of have to live in the future five to 10 years because it takes three or so years just to build a system. It takes a couple of years to ramp it up and you're dealing and you would like them to be able to use the computer for 10 years after.”
Huang: NVIDIA engineers use autonomous coding agents like Claude Code and Cursor
“We've got cloud code autonomously running in sandboxes all over NVIDIA, and that's really fantastic, and some people use codex, some people use cloud code, some people use cursor, some people use cognition, and we let kind of a thousand flowers bloom, let peop…”
Huang: OpenClaw is an AI Linux moment enabling custom model development
“In a lot of ways, open claw to me was very Linux moment to me. And now everybody can build their own AI.”
Huang: NVIDIA offered its full engineering workforce to OpenClaw's creator
“And we contacted Peter and we said, Hey, you know, all of the videos engineers are your engineers. That's what I told Peter. You've got this battleship outside your house. You know, break down the problem as you desire, and we'll contribute as you wish.”
Tan: All competitive alpha lies in building custom, self-improving AI models
“I feel like all the alpha is in building your own AI. I mean, if someone else is using whatever is off the shelf, but you're, you have a thing that can recursively self-improve, and it is, you know, I mean, the mechanic, people are very flippant about markdown…”
Huang: AI automates tasks rather than destroying entire jobs
“The narrative about AI destroying jobs is exactly backwards. AI eliminate tasks. AI automates tasks away. But it doesn't necessarily eliminate jobs. And the reason for that is because the job of a person has a purpose, and that purpose has many tasks. Some of …”
Huang: Radiology jobs grew 20% over recent years despite AI scan automation
“The task of reading radiology scans has been automated, but the number of radiology jobs has increased some 20% in the last several years, even though AI has taken over the whole field.”
Huang: NVIDIA had neural video simulators years before public video models
“Long before the first videos were generated outside that people saw a couple of years earlier inside our labs, we were driving a simulator completely generated by video, and computer, completely generated by neural networks”
Huang: Robotics had its ChatGPT breakthrough moment years ago
“I would say the chat GPT moment of robots happened a couple of years ago already.”
Huang: NVIDIA physical AI and automotive business is near $10 billion
“Our robotics business, autonomous vehicle business, basically physical AI business is probably almost, it's like ten billion dollars, so it's really, really big already.”
Huang: Physical AI will be NVIDIA's next $100B business within 10 years
“Likely this will be one of the largest industries in the world, and it'll take longer than a couple, two, three years. It'll take less than 10, and so this will be our next hundred billion dollar business.”
Huang: Computing technology reset makes this the best time to start companies
“Obviously, this is the greatest time in the last 60 years to start a company. The whole industry has changed. It's a complete reset from a technology perspective. The single most important technology in human history, the computer has been completely reset. An…”