Compiled & Edited by: Deep Tide TechFlow

Program: Axios Behind the Curtain

Guest: Jensen Huang, Co-founder and CEO of NVIDIA

Duration: 70 minutes

Location: NVIDIA Factory in Fort Worth, Texas

Conflict of Interest Statement: Jensen Huang is the CEO and co-founder of NVIDIA, holding approximately 3.5% of the company’s shares, with a personal wealth of about $181 billion. This episode discusses topics related to the expansion of the AI industry, chip demand, and government regulation in China, which are directly related to his personal financial interests. His stance clearly leans towards promoting the expansion of the AI industry and reducing regulation.

Summary

The CEO of the world's most valuable company sits in his newly built Texas factory and addresses nearly every hot controversy in the AI industry. Jensen Huang speaks candidly: The panic on Wall Street regarding China's AI models is a misunderstanding, the doomsday narrative about AI is "nonsense," the chip industry needs to expand 5 to 10 times in the next decade, and a bubble is unlikely to appear within five years. He has also publicly stated that Anthropic's Mythos model should be open to everyone, and OpenAI and Anthropic will be "the most successful IPO in human history." When talking about Trump, he said the president has an excellent memory, recalling chip model names like H20, H200, Blackwell, and Rubin; his only concern is that the government might be misled by "science fiction narratives" into over-regulating.

Key Quotes

"The market misunderstood the impact of DeepSeek and also misunderstood Kimi. Great models lead to more usage, and more usage means selling more Nvidia computers, building more data centers. The starting point is: good models lead to good applications, good applications lead to growth."

"Saying AI will destroy humanity is complete nonsense. Saying AI will eliminate half of American jobs is complete nonsense. All facts and evidence point to the contrary."

"China has more AI researchers than the rest of the world combined. It is destined for China to become extraordinary in this field. Want to stop China? That’s a foolish idea and simply impossible."

"The semiconductor industry needs to expand 5 to 10 times. Today, everything is in short supply: chips are short, memory is short, land, power, and construction workers are also short. This shortage is precisely a good thing, as it gives us time to build infrastructure."

"OpenAI and Anthropic will become the most successful IPOs in human history. How is it possible for a company founded just a few years ago to be worth a trillion dollars? Because AI is useful, and when AI is useful, it can make money."

1. The Wave of China's AI: Wall Street Misunderstood Again

Host Mike Allen opens with the most sensitive topic: The FT reports that China is considering tightening export controls on AI models and semiconductors; as soon as the Chinese model Kimi was released, NVIDIA's stock fell sharply, and semiconductor stocks dropped 18% in a month. What did Wall Street get wrong?

Jensen Huang's answer is straightforward. He says the market misunderstood when DeepSeek was released, and it's happening again now. "Good models lead to more usage, and more usage means selling more Nvidia computers. More data centers need to be built to provide more services, and technology will penetrate into more industries." He repeatedly emphasizes this logical chain: good models lead to good applications, good applications lead to growth, and growth requires more computing power.

Regarding whether models like Kimi should be banned, his answer is equally direct: "Of course they should be used; that’s smart." He explains that after downloading the model, you can fine-tune, enhance, and protect it; the model runs in a so-called "harness," which is in a "sandbox," that sandbox is safe, containing privacy protection, security control, and access control. He likens this to operating systems: Linux is open-source; millions of people around the world review, test, and strengthen it, so we can trust it. The same principle applies to open-source AI models.

He also mentions an easily overlooked point: open models and closed models are not opposed. "The people most likely to upgrade to good models like Anthropic or OpenAI are those who are already using AI." Free models lower the barrier to trying AI; once users get started, they naturally will want better services and will pay for closed model providers when that time comes. So the more open-source models there are, the more opportunities there are for closed models.

When asked about NVIDIA's sales in China, Jensen Huang provided an unexpected answer: "Our sales in China today are approximately zero." He said he has told investors not to expect any revenue from China. If the Chinese government and market welcome them back, that would be "a great honor"; until then, consider revenue from China as zero.

2. Where Doomsayers about AI Went Wrong

This part of the conversation is the most charged. The host asks: Have some of your peers in the tech industry over-exaggerated the risks of AI?

Jensen Huang's response is blunt. "It is acceptable to warn others, and it's better to warn with solutions, but fabricating facts is absolutely unacceptable." He then named several popular AI panic narratives: "Saying AI will end humanity is complete nonsense. Saying AI will eliminate half of American jobs is complete nonsense. All facts and evidence point to the contrary."

He provided several specific cases to support his argument. The number of radiologists has increased by about 20% because AI has automated scan analysis, allowing doctors to see more patients, and the demand for medical attention is so high that more radiologists are needed. The number of legal assistants has increased by about 10%, with similar logic. Manufacturing jobs have grown by approximately 50% in recent years because AI data centers need to be built, and chips need to be produced.

"Intuition and common sense tell you that AI will increase productivity, and increased productivity will create opportunities. Throughout history, technology has made societies increasingly efficient, creating more jobs rather than fewer. If it weren't for this, there should only be 100,000 jobs in America right now."

He also directed criticism at AI leaders. "Doomsayers spend too much time theorizing science fiction endings, perhaps to seem clever." When pressed whether he was talking about certain AI company CEOs, he did not deny it but said: "If you want to alert the world to this technology's incredible capabilities, that goal has already been achieved. We should spend time on how to make technology safe; that is the responsibility of technology leaders."

The host mentioned that Asia's attitude towards AI is markedly different from that of the United States, and Huang was surrounded by fans for autographs in Asia. His explanation is thought-provoking: "Perhaps it's because doomsayers spend too much time fabricating science fiction endings." In Asia, people embrace AI as a tool and opportunity rather than fearing it as a threat.

3. The Chip Industry Needs to Expand 5 to 10 Times; A Bubble is Unlikely Within Five Years

The host posed a sharp question: Every industrial revolution has gone through bubbles; where is this generation's bubble risk?

Jensen Huang's answer is cautious but clear. "A bubble will eventually come, but not today. We are at the very beginning of building." He provided a timeline: it’s unlikely within five years, and it depends on the situation between five to ten years. The reason lies in comprehensive constraints on the supply side. "The industry can build faster now, but there aren't enough chips, not enough memory, not enough land, not enough power, and not enough construction workers. We are constrained in every direction, every aspect."

This constraint is actually a good thing, he said. "This constraint holds the system together and gives us ample time to build infrastructure." Demand is strong, but the ability to translate that demand into productive supercomputers has been delayed due to all these physical limitations, pushing back the point where supply exceeds demand.

More critically, he assesses the overall scale of the semiconductor industry. "I believe the semiconductor industry needs to expand 5 to 10 times." Over what timeframe? "In the next ten years." This means the current chip industry is far too small to support the construction demands of AI infrastructure.

He explained why this semiconductor cycle is different from past ones. "This time is different because it is not demand-driven, not seasonal, not consumer-driven. It is driven by industrial infrastructure." Just like the world needs energy, the internet, roads, and railways, it now needs AI as a layer of intelligent infrastructure built on top of all existing infrastructure. This layer requires chips.

Regarding the risks of customers taking on debt to purchase NVIDIA products, he said he is not very worried. "These companies are excellent, and they generate a lot of cash." He specifically mentioned the initiation of the AI profit flywheel: AI is useful, so AI can make money. Coding agents are extremely profitable; they work useful jobs in high-paying positions, and many companies are willing to pay hundreds of millions of dollars each year to enhance their coding capabilities. "This flywheel is already in motion."

4. Token Economics: The Smarter AI is, the More Valuable It Becomes

This is the most technically deep part of the conversation. The host asks: What makes you confident that tokens will become increasingly profitable?

Jensen Huang's explanation is vivid. "Tokens are a form of embedding. They embed knowledge and intelligence. This number is not static; it’s not like the mathematical constant pi. The intelligence encoded in this number will become smarter over time." As intelligence becomes smarter, it becomes more useful, and the more useful it is, the more valuable it becomes, leading people to be willing to pay more.

He compared the AI industry to the past software industry. The software industry was "light capital," giving software companies high-profit margins. But the software industry of the AI era will become more "heavy capital," because producing modern software requires machines like the supercomputer in front of him to generate intelligence. Every industry will become more capital-intensive, but the payoff is incredible intelligence, productivity, and growth.

"We are laying the groundwork and building infrastructure; this is the largest-scale industrial infrastructure construction in human history."

Regarding the question of whether AI has already peaked, his answer is: "It’s impossible to peak because AI's penetration into society and industries has just begun." He also mentioned that in the past six months alone, $300 billion has been invested in venture capital and startups in the United States, creating new jobs and new companies.

5. Trump, Regulation, and Government Ownership

Jensen Huang's evaluation of Trump is surprisingly specific. "He is smart and remembers everything. Goodness, he really understands numbers." He pointed out that Trump is the only president who remembers NVIDIA chip model names like H20, H200, and Blackwell, and even knows the next-generation product is called Rubin.

He recounted his first meeting with Trump: "He said he wanted to restore America’s manufacturing capabilities, to re-industrialize America, to ensure a secure and resilient supply chain, and to bring semiconductor manufacturing back to America." He noted that the factory in Fort Worth where they sat for the interview directly results from that conversation.

When asked what mistakes the government should avoid, Jensen Huang rarely showed anxiety. "What I worry about is over-regulation, over-correction." He said some companies want the government to help create regulations favorable to themselves. "I believe we should compete openly." He dismissed the analogy of the AI race being like a sprint: "Whoever gets to the finish line first will always win? That’s nonsense. The ultimate victory is measured by whether society uses technology, not by who invented it. We didn’t invent electricity or manufacturing, but America applied it more quickly and passionately, thus becoming today’s America."

When the host pressed: If Trump called for NVIDIA's equity, what would you say? Jensen Huang's response was clever: "There’s no need. The U.S. already has an equity stake in NVIDIA. We paid $10 billion in taxes last year, and we will pay more this year. We create a lot of jobs and tax revenue. And don’t forget, most Americans are in the stock market today; when the stock market rises, everyone benefits."

Regarding whether the Mythos model from Anthropic should be open to everyone (it is currently only open to certain institutions), Jensen Huang's position is very clear: "It should definitely be open to everyone." He said it is Anthropic's responsibility to ensure the technology is safe, just like any software, quickly repairing vulnerabilities discovered. He mentioned the jailbreak incident of Mythos, saying, "Everything is fine; you and I are still here chatting."

Regarding the issue of open model companies distilling closed models and reselling them, his stance is relatively balanced: "It depends on the terms of service. If the service provider is not satisfied, they should contact that company. We have many traditional legal means to deal with it." But he emphasized that AI learning from other sources is a fundamental characteristic of intelligence. "AI must learn from something. The original AI also grabbed all existing knowledge from the internet. The content generated by AI now surpasses that created by humans; in a few years, 99% of online content will be AI-generated. You are constantly distilling the intelligence of other AIs."

6. The Robot ChatGPT Moment and the Trillion Agent Era

The host asks: When will the robot ChatGPT moment arrive?

Jensen Huang gave an unexpected answer: "The robot ChatGPT moment has already arrived." He explains that the ChatGPT moment is not when AI becomes useful; when ChatGPT was released in 2022, it merely felt "interesting and surprising." The truly useful phase has waited four years. Robots are now in that "interesting and surprising" stage. You can tell a robot, "Put the apple in the drawer," and it will reason out the sequence of tasks, including opening the drawer first and then placing the apple. "When you first see a mechanical robot truly do these things, it expands your imagination about the future of robots."

As for when they will become useful, he provided a timeframe of "three to four years" and said, "I wouldn’t be surprised."

Regarding the Agent era, he painted a striking picture. Today, a billion people use computers, but for most of the time, those computers are idle. In the future, everyone will be aided by numerous Agents operating computers around the clock. "We will have 100 billion, a trillion Agents running around the clock. Intelligent Agents, less intelligent Agents, specialized Agents, super Agents, all operating continuously. So the number of computers we need will increase significantly."

7. The Kung Fu and Philosophy of Pain of the CEO

The conversation finally returned to Jensen Huang personally. The host mentioned that he founded NVIDIA at 30, and now the company is valued at $5 trillion but has only 50,000 employees. He said that in ten years, there may only be 75,000 people, "as small as possible."

Jensen Huang defines the job of a CEO as "strategy." "Strategy is to realize the vision of the future as efficiently as possible with limited resources." He said he started this job at 30, making him one of the longest-serving CEOs in tech history. "This is my craft, this is my kung fu."

Regarding the theory that "pain and honing are the keys to greatness," he corrected that it does not relate to a specific painful event but is more like a continuous state. "No great athlete becomes great by chance. It involves a lot of practice when no one is watching, a lot of failures, a lot of pain and honing." He said this honing enhances skill, shapes character, and builds confidence and resilience. "Just like athletes say, at the highest pressure and tension, time seems to slow down. I have the same feeling. It comes from practice, from doing the same thing over and over again."

When asked what he would say to his nine-year-old self who came to America, his answer was an enthusiastic praise of the country: "America is the greatest country in the world, period. Even if we have challenges and differences, it's precisely those challenges and differences that make us great. Through open dialogue, through a dynamic system that allows freedom, free expression, free innovation, free entrepreneurship, and free collaboration." He concluded, "This country was built by immigrants, and it will need great immigrants in the future."

Regarding his habit of not wearing a watch, his explanation is very Jensen Huang: "Because now is the most important time. I refuse to let the schedule manage my life, I refuse to let a watch manage my life. If I am late, someone will tell me. Until then, I am 100% here."

Notes

1. Conflict of Interest: Jensen Huang holds approximately 3.5% of NVIDIA shares (around 860 million shares), with personal wealth estimated at $181 billion (Bloomberg data from June 2026). This episode discusses topics related to AI industry expansion, rising chip demand, and reducing regulation, all of which are directly related to his personal financial interests. His argument that "good models lead to more chip demand" essentially states that Chinese AI models benefit NVIDIA; readers need to judge for themselves.
2. About China sales = "approximately zero": Jensen Huang claims NVIDIA's sales in China are "approximately zero." Based on publicly available financial reports, U.S. export controls have indeed limited NVIDIA's high-end chip sales in China to very low levels, but "zero" is Jensen Huang's phrasing; actual figures should rely on financial reports.
3. About AI job creation data: The increases of 20% in radiologists, 10% in legal assistants, and 50% in manufacturing employment were cited by Jensen Huang and are not sourced in the dialogue. Readers may cross-verify independently.