Jensen Huang's Carefully Built Nvidia Story Is Getting Harder to Control

Jensen Huang has always treated the story of Nvidia as part of the product. That's not a knock. Few executives in tech have been as deliberate about how their company gets described, and for most of the past decade that deliberateness worked. The question now is what happens to a carefully managed narrative when the facts around it get messier.

A prep call that looked like a rehearsal

In July 2024, I flew to Denver to interview Huang on stage at SIGGRAPH, the annual gathering of computer graphics researchers and practitioners. I had already spent a long sit-down with him earlier that year for a WIRED feature, so I assumed this would be a follow-up in the same vein. The prep call suggested otherwise. Madison Huang, the CEO's daughter, ran the call. She walked me through the slides Jensen planned to present and suggested moments where I might set him up to talk about them.

Prep calls with communications teams are routine. Handlers usually try to steer journalists toward approved topics so that an executive doesn't get blindsided. But being handed a slide deck and a list of cue points is a different thing. It turns the conversation into a shared production, and the journalist's job becomes keeping the questions from being absorbed into the presentation.

I raised the issue with my editors and with Nvidia. The compromise was that I'd ask the questions I had planned, including ones about the environmental footprint of data centers, and Huang would refer to his slides where he wanted to. The session didn't go smoothly. It's on YouTube, and you can judge it yourself. Huang was domineering in places, sidestepped some questions, and then, right after the session ended, held an impromptu Q&A with a far more receptive crowd. Meta's Mark Zuckerberg was next on stage for a warm fireside chat.

I've done plenty of adversarial interviews and don't lose much sleep over them. What stayed with me was how clearly Huang expected to set the terms. He didn't just want to control the frame. He assumed he would, and for a long time that assumption held up.

The founder story as a product

Huang's narrative is well-built, and much of its appeal comes from how familiar its parts are. He's the immigrant engineer who started with little, the underdog who ate at Denny's during the early years, the husband who chased his college sweetheart, and the executive who worked Sundays and lay awake listening for competitors' hoofbeats. He brought his children into the business and even put an abstract version of the Nvidia logo on his arm. Each detail reinforces the same idea: this company is an extension of one person's conviction.

Storytelling isn't incidental to that approach. In a conversation earlier in 2024, Huang pointed to Adobe CEO Shantanu Narayen as someone he admired, noting that Narayen had wanted to be a journalist because he loved telling stories. Huang went further, saying that strategy-setting and culture-building are both forms of storytelling. He also said Nvidia was never really sold on a pitch deck. It was sold on a story.

That framing has real strategic value. Companies that sell hardware on long time horizons, especially in AI, depend on investors, customers, and engineers believing in a direction before the results are visible. A founder who can make the direction feel inevitable has an advantage. The risk is that the story starts to carry more weight than the operating details underneath it.

A flat structure with a single center

Nvidia's management structure is another piece of the Huang lore. The company has long been described as flat, with a few dozen senior executives reporting to him. Figures floating around the press vary, and people close to the company have told me the number of direct reports is nearer to 35 than 60. Huang has argued that the setup lets Nvidia move faster, which matters when the underlying technology changes this quickly.

Flat structures have obvious advantages. Decisions don't wait in queues, and information reaches the top without much filtering. They also concentrate attention in one person. Huang has been known to sit in on most significant meetings and to track sourcing, supply chains, and how chips get allocated. That's an impressive level of engagement, and it's also a single point of dependence. When one person is expected to know everything, every gap in knowledge becomes harder to explain, and every surprise reflects on the center.

That tension is where the story gets harder to tell. A former employee I spoke with put it bluntly: you can't claim to have known every moving part of the business and then plead ignorance about where the chips ended up. Whether or not that critique holds in any specific case, it speaks to a structural problem. The more a company's identity depends on its founder's omniscience, the more damage a single credibility gap does.

What happens when the plot gets complicated

Huang's story was always going to face tests. Nvidia sits at the center of the US–China AI rivalry, where export rules and chip sales are shaped by forces no single executive controls. Huang's answer to that pressure, as I wrote at the time, was to lean on the same storytelling skills that had built the company, including pressing the case in Washington to loosen export controls so Nvidia could sell some chips into China. Those efforts were a reminder that narrative is a tool for shaping outcomes, not just for describing them.

That's the core issue for anyone watching Nvidia closely. A founder-centered narrative works well when the company's trajectory is clean. It gets harder to sustain when the business is entangled in geopolitics, supply constraints, and questions about how its products are used. In those conditions, the story has to account for facts it didn't write.

Huang built the most valuable publicly traded company in history, and that achievement doesn't depend on the mythology. What's changing is the balance between the myth and the mechanics. For developers, researchers, and technology buyers, the practical takeaway is to evaluate Nvidia the way you'd evaluate any supplier at this scale: by its roadmap, its supply commitments, and how it handles the constraints it can't narrate away. The founder's version of events is one input among many, and it's becoming a less reliable guide than it used to be.