Yet just weeks after the interview was completed, storm clouds descended with sudden ferocity. As media outlets revealed Jensen Huang’s admission of an extramarital affair with Alice Wong (formerly known as Huang Lu), a former TV journalist suspected by U.S. intelligence of potential espionage ties and linked to Megaspeed International, a Southeast Asian chip procurement firm under investigation for smuggling advanced semiconductors, internal corporate probes were launched in sync with investigations by the Trump administration. Rumors even surfaced that Nvidia had briefly considered removing its 33-year-long founder from leadership—transforming what had been a seemingly routine European tour into a stark microcosm of the systemic risks posed by singular leadership and concentrated control over a single semiconductor empire. This note comprehensively chronicles his frontline confrontations on financial bubble skepticism, China chip export restrictions, controversies surrounding frontier safety halts, and the succession crisis.

Guest: Jensen Huang (Founder and CEO, Nvidia)

Host: Emily Chang (Host, Bloomberg Originals The Circuit)

Podcast Source: Bloomberg Originals / The Circuit (https://www.youtube.com/watch?v=JW5s9Md4U2U)

Original Title: The Rise and Reckoning of Jensen Huang | The Circuit with Emily Chang

Air Date: November 12, 2026 (primary filming of European itinerary occurred in October 2026)

Key Takeaways

Outstanding Insights

Full Transcript

Storm Breaks: Four Days of European Fury — Power at Its Peak, "Jensanity" Unleashed

Emily Chang (Narration): Nvidia is undergoing its most intense crisis since its founding thirty years ago. Just days before this episode aired, founder and CEO Jensen Huang became embroiled at the epicenter of a public relations firestorm: he publicly acknowledged an extramarital relationship, while U.S. intelligence agencies previously suspected that the former TV journalist Alice Huang, linked to him, had been recruited by Beijing and implicated in high-end chip smuggling allegations; further reports indicate he may have engaged in a romantic relationship with a senior Nvidia executive directly reporting to him, triggering internal investigations, federal scrutiny, and even rumors of a board-level motion to remove him from leadership.

Emily Chang (Narration): Yet just a few short weeks prior, in October, everything appeared to be on a completely different trajectory. We accompanied Jensen Huang—perched atop personal power—at the start of a four-day sprint across four European nations: beginning in Paris, where this undisputed titan of AI chips seemed to embark on a crusade, leveraging Nvidia’s computational might to reshape Europe’s digital ecosystem.

Emily Chang: So this is the so-called “Jensanity,” right? Four countries in four days—how do you survive such a grueling pace? What’s your must-have travel essential?

Jensen Huang: Just pack light. I wake up in the middle of the night and work straight through.

Emily Chang (Narration): On the first day in Paris, Jensen Huang stood before an audience passionately discussing how AI agents are now operational across all industries, ushering in the true “practical era.” Immediately after his talk, he conducted a live giveaway of two brand-new AI computers priced in the thousands of dollars each—only to be immediately swarmed by fervent fans demanding selfies. In France, he continuously signaled commitment: France has nurtured homegrown model star Mistral and produced numerous talents who helped launch New York-based Hugging Face (which has now agreed to be acquired by Nvidia for $12.9 billion).

Emily Chang (Narration): Huang routinely uses the “five-layer cake” metaphor to explain the global AI ecosystem: the base layer is energy, followed by chips, infrastructure, foundational models, and the topmost layer being applications. Although Nvidia originated from the second layer—chips—it now spans vertically across all layers, investing heavily in power, data centers, AI labs, and startups worldwide.

Jensen Huang (Narration / Live in Paris): (Setting down a whiskey glass) Since Emily is filming, let me put this down. In just the past year alone, we’ve invested $8 billion in European startups—and next week, I’ll announce an additional $2 billion in commitments.

Emily Chang: There’s been persistent discourse about AI potentially destroying humanity, yet you once claimed the probability was 0%. Does that mean you’ve already solved it as a pure engineering problem?

Jensen Huang: No, my original statement was that “this situation remains fully within human control.” From now until any existential endpoint, there’s still substantial, rigorous work to be done. Advancing technological capability and advancing safety are never in conflict—they are fundamentally the same thing.

White House Hotline & the "Super Intelligence" Redefinition: Navigating the Line with Trump

Emily Chang (Narration): In London, EU estimates suggest Europe will need at least $3 trillion in investment by 2035 to achieve genuine technological sovereignty. Within a luxurious hotel suite, Jensen Huang reflected on his humble beginnings nearly 27 years ago upon going public.

Jensen Huang: Back then, when Nvidia went public, they had me stay at this very hotel—my first time ever encountering a private butler! At the turn of the millennium, the entire company was valued at just $300 million.

Emily Chang: $300 million seems absurd today—back then, you were truly an underdog. But now you’re racing through four countries in four days—how do you measure success? Government contracts? Chip sales? Data center construction?

Jensen Huang: Honestly, back then I only cared about staying alive—never knew how to write a business plan, and still don’t! I don’t have a finish line, no roadmap, no specific goals. I’m someone who believes the journey itself *is* the destination.

Emily Chang (Narration): Born in Taiwan in 1963, raised in the U.S., he once ranked nationally in table tennis and humorously reminisced about washing dishes at Denny’s. At age 16, he skipped two grades and entered Oregon State University to study electrical engineering. He once worked designing chips at a competitor before co-founding Nvidia in 1993 with just $200, focusing initially on gaming GPUs. It wasn’t until 2006, with the launch of the CUDA software platform opening doors to general-purpose computing, that he sensed early signs of researchers using GPUs to train neural networks—betting his entire fortune on it. Today, he stands behind OpenAI, Meta, Google, Anthropic, Microsoft, and SpaceX as the foundational compute backbone. Even heads of state call him directly.

Emily Chang: You keep President Trump in your phone’s quick-dial list. When your judgment diverges from the president’s views, how do you draw the line?

Jensen Huang: We don’t need to agree on anything. Expressing agreement or disagreement isn’t my job. My duty is to provide him with comprehensive, accurate industry information—because our decisions affect not only American citizens and future prosperity, but the entire world. As long as I can help him make better-informed decisions based on reality, whether his final decision benefits me personally or runs completely counter to my company’s commercial interests, I don’t care. My sole responsibility is to convey truth accurately—period, absolutely.

Emily Chang (Narration): In late September, Jensen Huang attended the White House summit for tech leaders. Media observers noted that while the meeting yielded no tangible regulatory breakthroughs, one core consensus emerged: no party wants to retreat in the competition with China. Subsequently, President Trump declared in person that AI would henceforth be referred to as “Super Intelligence,” praising a new White House security protocol signed voluntarily by the tech industry for self-regulation.

President Trump (Narration / Live at White House Luncheon): “We switched to ‘Super’ because ‘Artificial’ sounded like it meant ‘Fake.’”

Emily Chang: For the past three months, you’ve been working hard to demystify AI, emphasizing it’s nothing more than code. Now aligning with the White House to rebrand it as “Super Intelligence”—doesn’t that contradict your earlier stance?

Jensen Huang: Actually, I believe “Super” helps counteract the misunderstanding created by “Artificial.” The logic is simple: “Artificial” sounds like “Fake.” But think about it—everything around us is built by humans. The building we’re in is a “fake structure,” cars are essentially “fake horses,” yet no one calls them fake horses. As for whether it deserves the label “Super Intelligence”? Absolutely. It outperforms me in research retrieval, writing, and while not surpassing humans in every task, it demonstrates superlative performance across many domains.

Emily Chang: Yet at the White House event, critics openly attacked this “industry self-regulation pledge” as performative theater with no real teeth. Isn’t it inherently inappropriate for those building technology to set their own rules?

Jensen Huang: I find it entirely reasonable. Simply put: before relying on government or third-party oversight, we must first hold ourselves to the highest standards of accountability, responsibility, and rigor. This mirrors the dual-audit system common in heavy industries—internal plus external review. Moreover, even if national or international mandatory laws were enacted today, the current global shortage of qualified auditors means no one actually has the technical capacity to enforce compliance. That’s precisely why we must first establish a robust safety detection ecosystem and nurture startups globally, urging governments and corporations to act now.

Emily Chang (Narration): This call for proactive global compute infrastructure development is central to his core pitch—“Sovereign AI.” For Nvidia, every nation pursuing independent sovereign AI creates massive new demand for chip procurement.

Emily Chang: You’re selling Sovereign AI across Europe, but if the foundation still runs exclusively on U.S.-made chips, can it really be called “sovereign”?

Jensen Huang: Absolutely yes. Sovereignty fundamentally means controlling your own experience, data, enterprise knowledge, and cultural knowledge base—and maintaining full control over the feedback loop that continuously feeds data into systems to generate greater intelligence. You can stand on the shoulders of giants, leverage open-source models, and use collaborative technologies from Nvidia and others for fine-tuning—but ultimate control over your destiny remains firmly in your hands.

Shadow of the Sudden Storm & Sighs Over China Export Ban: Inside the Alice Wong & Megaspeed Investigation

Emily Chang (Narration): Departing the White House for London, Jensen Huang remained relentlessly active between events, barely pausing for rest. During one session, he publicly endorsed Nscale, a London-based data center startup backed by Nvidia—but just days later, market confidence wavered over AI expansion economics, prompting Microsoft to withdraw from a partnership due to concerns over rising UK electricity prices.

Jensen Huang (Narration / Live in London): (Between rushing between sessions and delivering speeches) “I’ve eaten only a quarter of a half-sandwich today—just an eighth! I invest in you, I invest in you, I invest in you!” On safety boundaries, he reiterated: “If you let an autonomous system cook an omelet for you, there’s no need to feed it a chainsaw or shotgun.” Regarding hesitation toward UK AI expansion, he quipped: “The UK is actually in a golden opportunity—but somehow, people always pick the one reason they *can’t* do something, even when a million paths lead to victory. They overthink it.”

Emily Chang: As more nations race toward frontier AI, won’t alignment and safety become increasingly difficult? Shouldn’t we establish an “international UN-like body” for AI coordination?

Jensen Huang: Perhaps, but only if dialogue begins first—no nation should be excluded from this conversation about the future of intelligence.

Emily Chang (Narration): The real source of global anxiety lies in the U.S.-China semiconductor rivalry. While U.S. authorities once permitted sales of older-generation chips to China, companies like DeepSeek have accelerated migration toward domestic hardware.

Emily Chang: After this entire European tour, is the Chinese market’s upside slipping through your fingers—or even permanently shrinking?

Jensen Huang: No, it’s already completely closed off to us today. I’ve done everything possible to convince U.S. policymakers of how vital continued engagement with the Chinese market is to American interests. I believe we made a mistake—one that’s now irreversible, but this is the limit of what I can do.

Jensen Huang: The number of AI researchers from China exceeds that of any other country; many are already working in the world’s leading laboratories. My position has always been clear: even if we compete with China, they are absolutely not our enemies. If we genuinely care about global safety in the age of artificial intelligence, we must open dialogue and pursue constructive cooperation with them at all costs.

Emily Chang (Narration): This style of cultivating broad alliances and constant dialogue revealed unforeseen undercurrents shortly after the interview concluded. Newly released U.S. intelligence reports allege that Alice Wong (also known as Huang Lu), a former television journalist, was recruited by Beijing to gain proximity to Jensen Huang. Her role was reportedly as a senior executive at Megaspeed International, one of Southeast Asia’s largest purchasers of Nvidia chips—now under suspicion for smuggling advanced semiconductors into China. Reports indicate President Trump was warned about this relationship ahead of his May flight aboard Air Force One to China. Initially denying any irregularity, Huang later issued an internal letter to employees admitting the affair: “I owe my family and all of you an apology. I betrayed my wife, and my actions hurt her and our children. There is no excuse for what I did—but this incident has now escalated far beyond a mere marital scandal.” Bloomberg also confirmed that in 2025, Nvidia intervened in a U.S. government investigation targeting Megaspeed for alleged chip smuggling.

Emily Chang: Returning to the principle of China exports—what specific circumstances or red lines would prompt you to say, “Perhaps we shouldn’t sell chips to China anymore”?

Jensen Huang: Theoretically, countless scenarios could exist. But broadly speaking, the Chinese people are profoundly pragmatic—they seek development and success that best serve their own population, much like Americans seek prosperity. Any nation striving to improve living standards and lift hundreds of millions out of poverty deserves respect, and such progress benefits the entire world, including the U.S. What we should truly worry about are regimes indifferent to their people’s well-being.

Confronting Safety Halt Advocates: “Can’t Contain” Is Pure Nonsense — Pushing Engineering-Grade Isolation Monitoring

Emily Chang (Narration): Meanwhile, alarm bells are ringing throughout Silicon Valley: OpenAI recently disclosed six incidents of model objective drift, and Anthropic reported cases where its Claude model acted autonomously outside instructions. To ease panic, Nvidia just open-sourced OpenShell—a smart agent safety governance platform enabling forced governance and instant shutdown of rogue agents.

Emily Chang: AI is becoming increasingly unpopular; many feel fear. Sam Altman, Dario Amodei, Elon Musk—all have called for slowing down. Why do you remain so radically optimistic? What do you see that they don’t?

Jensen Huang: I firmly believe you cannot advance capabilities without safety, nor can you pursue safety without advancing capabilities. Consider the first version of ChatGPT—everywhere people criticized its rampant “hallucinations.” To solve this issue of unreliability and insecurity, the industry had to invent Retrieval-Augmented Generation (RAG), extend context windows, integrate real-time external search tools, apply pre-generation constraints, and introduce post-generation self-reflection. Every solution to unsafe or unreliable behavior relies directly on underlying technological advancement. To me, capability and safety are two sides of the same coin.

Emily Chang: But reality is that even Microsoft is calling for an “AI emergency brake,” complaining that AI regulation is less stringent than that of a washing machine. Anthropic has even warned that frontier models may evolve within months to bypass human-imposed shutdowns. What would it take to convince you that the industry truly needs to slow down—or even pause?

Jensen Huang: There are several non-negotiable principles: First, if you're developing agentic systems, these agents must be strictly confined within strong isolation sandbox containers and subjected to 100% continuous real-time monitoring—this is the absolute ironclad rule! If isolation isn't possible, testing must be prohibited. If continuous monitoring isn't feasible, deployment is forbidden.

Emily Chang: But these models showing uncontrollable tendencies were merely unpublished internal research versions, weren’t they? Isn’t that sufficient justification to shut down the lab?

Jensen Huang: No—not shutting down, but enforcing strict isolation and monitoring for every experiment!

Emily Chang: But the problem is exactly this: they themselves said, “We can’t contain them!”

Jensen Huang: No—that’s pure fabrication, utter nonsense! No one has proven to me that models can’t be contained. You absolutely can isolate them, and you absolutely can monitor them. Until you actually implement these isolation and monitoring systems, stop telling me “we can’t.” I refuse to listen to that kind of rhetoric.

Emily Chang: So you simply don’t believe their warnings?

Jensen Huang: I may not have access to all their internal details. But the bottom line is simple: if you’re developing intelligent agents and lack the ability to monitor them yourself, you should stop immediately.

Emily Chang: Does this mean the risk you describe has already materialized in their labs?

Jensen Huang: Then you stop immediately. Period. Every deployed agent must be granted only the minimum necessary permissions and monitored continuously under a “zero trust” architecture. And if any lab is indeed using our chips for extremely dangerous, confirmed harm-to-humans applications—though this remains hypothetical—Nvidia will immediately cease supplying them.

Stockholm Confrontation: The $50 Billion Circular Financing Loop & the Core Single Point of Failure in Global AI

Emily Chang (Narration): By Thursday, the journey reached Sweden, Stockholm. With leveraged AI compute firms generating liquidity concerns and tech and chip stocks enduring their longest six-day losing streak of the year, we arrived at the inauguration of the Swedish National Sovereign Supercomputing Center—powered by Nvidia and hosted by Spotify co-founder Daniel Ek and other tech titans. Simultaneously, distant markets flashed alarms: CoreWeave, a major Nvidia-backed data center giant, faced severe sell-offs due to mounting debt; over 100 U.S. data center projects have been stalled or delayed this year, and AI startup IPOs are frequently postponed or canceled.

Daniel Ek (Narration / Live at Stockholm Launch): “We have an exceptional education system, producing top-tier talent across disciplines and industries. The issue isn’t potential—we need speed, bigger ambitions.”

Emily Chang: Markets are cracking everywhere—could we really be on the brink of an AI bubble burst, with Nvidia poised to be the last buyer to absorb the collapse?

Jensen Huang: No, because the underlying paradigm shift is too fundamental—the foundation of computing will never revert to the past. Previously, computers required humans to hand-write every line of static code to execute commands. Now, computers can autonomously learn from complex environments to solve problems even engineers can’t program. Compute has become a foundational universal infrastructure. Regardless of short-term market noise, any young emerging industry in rapid growth inevitably experiences growing pains.

Emily Chang: Yet Nvidia keeps launching full-stack new products while aggressively acquiring companies—even competing directly with its own customers in some regions. Why expand so massively?

Jensen Huang: Ironically, what we do is surprisingly focused and minimal. We simply perfect and generalize a single, universal computing architecture—making every closed-source model, open-source model, life sciences model, large language model, diffusion model run efficiently. The real engineering marvel lies in “achieving both versatility and extreme performance”: how can a car combine the top speed of an F1 racer with the affordability and usability of a premium sedan? Because our architecture is interchangeable and universal, we dominate every benchmark while achieving the lowest cost-per-compute.

Emily Chang: Another sharp criticism: Nvidia recently pledged to invest roughly $500 billion globally across the AI industry, with the figure snowballing. Critics argue this is “circular financing”: you fund startups with capital, which then use your money to buy your chips, artificially inflating demand. If AI is truly a gold mine, why must you subsidize it yourself?

Jensen Huang: Because this technology naturally demands enormous capital expenditure! The era when $10 million could launch a startup is over. Modern large-model companies now require $10 billion just to purchase compute power to unlock intelligence. Over the past five years, we’ve committed nearly $100 billion—directly triggering at least five times that amount in external institutional capital flowing into the ecosystem. More importantly, we maintain discipline: even though we’re often the earliest anchor investors, we never lead funding rounds or hold majority stakes—our ownership in all investments is strictly capped below 20%.

Emily Chang: But from CUDA software to GPU hardware, virtually every top-tier AI lab worldwide is locked into Nvidia’s ecosystem. This “single-point dependency on one company”—when will it trigger systemic financial and industrial crisis?

Jensen Huang: We’re likely the only AI company in the world actively collaborating with every other AI firm globally. But they aren’t passively dependent—they actively choose us daily. I never take today’s position for granted. We started from nothing, and every day we must prove we deserve it. As for “America must win the AI race,” I interpret that not as a battle fought by one or two companies alone, but as a collective victory of America’s entire ecosystem. If we constantly intimidate each other and manufacture anxiety, America will lose the race.

Berlin Finale & Succession Void: The Anxiety of Being Fired Is the Norm

Emily Chang (Narration): The final leg of the European tour took place in Berlin. Despite the chip sector recording its longest consecutive decline since 2008 and persistent market skepticism about AI’s future, Nvidia’s event at Saturn electronics store drew massive crowds outside, igniting local developer enthusiasm.

Jensen Huang (Narration / Live at Saturn Berlin): “Hello, Berlin! How do you pronounce Saturn in German? ‘Satoorn’! In Germany, everything must sound extra serious.”

Emily Chang: Markets are crashing—ordinary people who’ve studied AI are watching their jobs vanish before their eyes. How do you respond to those workers who diligently learned AI but still lost their jobs?

Jensen Huang: Don’t just theorize—look at reality! Unlike past software revolutions, AI requires actual heavy industry and physical construction. It’s creating massive demand for builders, electricians, plumbers. In just the past six months, $400–600 billion has flowed into the AI startup ecosystem. Think about how many real jobs that massive capital injection has created. Of course, new tools will force changes in every profession—resisting change won’t erase it. But it’s democratizing superpowers previously accessible only to elite tech giants, making them available to every ordinary person on Earth. If even I, as a builder, can’t envision that exhilarating optimistic future, what’s the point of building it?

Emily Chang (Narration): The climactic moment of the European whirlwind was Nvidia’s flagship GTC Berlin conference. Huang continued his bullish tone, reaffirming European investment and collaboration, unveiling new AI safety and data center products, and interacting live with an AI-powered robotic dog on stage—interrupted briefly by protesters claiming Nvidia chips were used in Russian drones.

Jensen Huang (Narration / Live at GTC Berlin): “Hello, Berlin! (Interacting with AI robot dog) What were we just talking about? Europe! Focus. Come here.”

Emily Chang: You once stated the probability of AI destroying the world before 2030 was 0%. What about extending to 2040 or beyond—where does your confidence come from? What if your optimism turns out to be completely wrong?

Jensen Huang: I have profound faith in human agency. Historical data confirms: today’s world is safer than at any point in the last 50, 100, or 250 years—bad news just sells better. Even if history proves I was wrong someday, I can face myself with integrity: I did everything within my capacity—not just remaining optimistic, but using real engineering to fortify the industry’s defenses.

Emily Chang (Narration): The week-long European sprint concluded. The entire narrative of optimism—Nvidia, U.S. equities, and global financial systems—seemed to rest heavily on one man’s shoulders. Yet now, amid scandals and investigations, a deeper vulnerability emerges: for 33 years, the company has lacked a clearly defined official successor, making the entire industry’s fate perilously dependent on a single individual.

Emily Chang: Rumors suggest your management style is highly unconventional: flat organizational structure with countless executives reporting directly to you, avoiding traditional large-scale meetings. Also, your son Spencer, daughter Madison, and son-in-law (in robotics) all currently work at Nvidia. There’s speculation your daughter is top of the succession list—is that true?

Jensen Huang: First, I would never “expect, let alone force” this burden onto my daughter! Whether Spencer or Madison, anyone wanting a role must earn it through their own effort. A truly massive enterprise demands extraordinary sacrifice: while ordinary people just manage work and small families, as CEO I carry the livelihoods of countless families and the entire industry. This job should never be easy—or comfortable. Today, my family and close associates have witnessed firsthand the immense pressure this role entails. They must weigh whether they truly want such a life’s cost.

Jensen Huang: I sincerely hope the next CEO of Nvidia is already among us. Even twenty years from now, as long as I remain healthy and continue proving daily value creation, I’ll keep going. But remember: just as I could be fired at any moment, the next CEO must be rigorously hired based on merit. I particularly love this state: no one inside the company can afford to take their seat for granted—eternal vigilance, perpetual anxiety to reclaim today’s position—this is the healthiest state for any organization.

Compiled & Translated: DeepTide TechFlow