The Man Who Built a Trillion-Dollar Company Still Goes to Bed Worried

Jensen Huang's real secret isn't vision. It's what he does with the weight.

Jensen Huang has built one of the most valuable companies in human history. Nvidia's chips power the AI revolution. His personal net worth is measured in tens of billions. He is, by almost any external metric, one of the most successful people alive.

And he goes to bed worried.

Not metaphorically. He describes it directly: the weight of knowing that hundreds of thousands of people — teachers, policemen, ordinary investors who bought stock years ago and are now millionaires — have tied their financial futures to the company he runs. The upstream suppliers. The downstream partners. The ecosystem of businesses that depend on Nvidia continuing to function at the level it has been functioning.

He carries all of that. He doesn't pretend it away. He doesn't tell himself it isn't real.

What he does instead is something specific, and it's worth paying close attention to.

The System

When something worries him, he decomposes it.

Not vaguely. He means this as a precise practice: take the thing that's generating anxiety, break it into its actual component parts, identify what can be acted on, figure out who needs to know, tell them, and then — only then — release it.

The logic is simple but the execution is harder than it sounds. Most people, when something worries them, either suppress it entirely or ruminate on it endlessly. Both are costly. Suppression means the problem doesn't get addressed. Rumination means the problem gets addressed repeatedly, in the same unproductive loop, consuming energy without generating resolution.

Huang's approach is to convert the anxiety into a task list. What exactly is the threat? What would need to happen to address it? Who is capable of doing something about it? Tell that person. Now it's their problem too — or more precisely, it's no longer sitting entirely inside your head.

He describes it as sharing the burden. Not offloading it, not making it someone else's to carry alone, but distributing it. Getting it off your chest and into the system where it can actually be worked on.

After that, he says: what else can you do?

It's a genuine question, not a rhetorical one. If you've broken the problem down, identified the actions, delegated what needs to be delegated, and taken the steps within your control — what remains? Only worry. And worry about a problem you've already addressed is just noise.

Systematic Forgetting

He draws an analogy to machine learning that's more useful than it might first appear.

One of the core mechanisms of AI training is selective forgetting — the system has to learn which information to retain and which to discard, or it gets overwhelmed with data it can't generalize from. You can't learn everything. You can't keep everything. The ability to forget, deliberately and systematically, is part of what makes learning possible.

He applies this to his own psychological life. The setbacks, the embarrassments, the failures — he talks about forgetting them not as denial but as a deliberate practice. The last point is behind you. The mistake was made. The analysis was done. Now it's over.

What keeps him moving, he says, is being pulled by the light of what's next — the next problem, the next opportunity, the next version of the future he's building toward. Great athletes work this way, he notes. They worry about the next point, not the last one. The capacity to let the past be past — not by pretending it didn't happen, but by genuinely moving your attention forward — is a skill, not a character trait you either have or don't.

How Hard Could It Be?

There's something else in what he describes that's worth sitting with.

He's said publicly that if he'd known in advance how difficult building Nvidia would be, he wouldn't have started. He frames this not as a cautionary note but as an insight about how hard things actually get built.

You go in thinking it's manageable. Then you discover it isn't. Then you discover it's much harder than that. Then things happen you couldn't have anticipated, setbacks that surprise you, humiliations you didn't see coming. And you keep going not because you foresaw all of that and decided it was worth it, but because by the time you know how hard it actually is, you're already in it, already past the point where stopping makes sense.

His default question when encountering something daunting is: how hard could it be?

Not as false confidence. As a psychological mode — the deliberate choice to approach new things with a fresh mind rather than pre-loading them with simulated disasters. If you spend too much time imagining all the ways something could go wrong before you start, you build a version of the difficulty that might be worse than the reality, and you carry that simulation as dead weight before you've taken a single step.

Go in fresh. Deal with what actually happens. Forget what you can't change. Keep going.

Reasoning Out Loud

One more thing he describes that's unusual for someone at his level.

He reasons out loud, in front of his team. Not presenting conclusions. Walking through the actual steps that got him there. Showing his work.

The conventional assumption is that leaders should project certainty — that showing your reasoning exposes you to challenge and undermines authority. Huang's experience is the opposite. When you show your reasoning, people can engage with the steps rather than just accepting or rejecting the outcome. They can say: I think that step is wrong. And then you can reason forward together, which gets you to better conclusions faster than any individual could reach alone.

He calls it collective path searching. It requires a specific tolerance: the willingness to be wrong in front of people and not experience it as catastrophic. He describes this as something that got easier over time, not harder — years of being wrong in meetings, having ideas challenged and sometimes dismantled, and discovering that the process doesn't destroy you. It just produces better thinking.

His first job, he mentions in passing, was cleaning toilets.

He built the most important chip company on Earth.

The gap between those two points wasn't filled with certainty. It was filled with the practice he describes — decomposing, delegating, forgetting, moving forward, staying curious, being willing to be wrong.

What This Actually Is

This isn't a story about genius. Huang is genuinely brilliant, but the specific things he describes — the decomposition practice, the systematic forgetting, the willingness to reason publicly and be corrected — these are learnable. They're not traits you either have or don't. They're responses to difficulty that can be developed.

The interesting question, if you're someone who carries weight the way most people carry it — silently, internally, in the loop of unprocessed anxiety that recycles without resolution — is whether you've ever actually tried the alternative.

Not suppression. Not rumination. Decomposition. Delegation. Forwarding.

The thing you're worried about: what exactly is it? What are the actual components? What action would address each one? Who needs to know? Tell them.

After that, what else can you do?

If the answer is nothing, then the worry is just noise. And you can, as Huang describes, choose to let it go — not because it doesn't matter, but because carrying it further serves no purpose.

That's not stoicism in the abstract. It's a specific practice. The man running a trillion-dollar company does it every day, before he goes to sleep.

There's probably something in that.