Jensen Huang: The Doomer Hoax, Superintelligence is Here, and The Future of AI (ft. President Trump)

Some people call it vision. Vision is an awfully big word to me, because I believe first of all that vision matters.

We preempted the weekly show, and there are only three people we preempt the show for: President Trump, Jesus, and Jensen. That's Jensen Huang, founder, president and CEO of Nvidia. Whether you know it or not, his decisions are shaping your future. Nvidia is the most important stock in this market, and Jensen is arguably the best executive in history — revenue exploded 97% year-over-year.

Not only is demand already strong — it's actually accelerating. Nvidia is the only computing platform that is a full-stack AI factory. A GPU is like a time machine because it lets you see the future sooner — and if you can see and predict the future, you have a better chance of making that future the best version of it.

Jensen Huang then joined the hosts to a standing ovation on his way in. The crowd's enthusiasm was on full display — one host introduced him as "GPU Jesus," and he responded, "Thank you. I love you back. Number one podcast in the world."

The hosts complimented his new jacket, and he explained that he had auctioned the opening. "I just felt you guys needed some energy," he said — serious topics, but they should still be discussed with energy.

The conversation opens with Dario Amodei's essay released over the weekend. The host jokes about which essay, and whether it was AI-assisted — "did anybody run it through Pangram?" — before noting what surprised many people: the coalescing of the frontier labs around the essay itself. Jensen is asked for his high-level reading before getting into the details.

Jensen's response covers several points. First, the safety content has to be taken seriously — safety is paramount. But he argues that safety versus leadership, and safety versus fast innovation and execution, are false choices: America can innovate quickly, execute quickly, and lead while doing it safely.

On internal control, he calls the Anthropic whistleblower a very serious matter — whenever there's a whistleblower, you take it seriously — and praises the courage it took to voice those concerns. Still, he thinks some issues were conflated: the whistleblowing itself is fine, but the scientific prediction about the future is not grounded in science. It was expressed by a scientist, but it isn't science, and he takes issue with that.

On pausing and pacing, he frames these as voluntary things a company could do if it felt out of control. Nobody knows what the whistleblower actually saw — only he does. It might, Jensen speculates, reflect a clumsy transition from research to engineering: these labs are moving from research to engineering, with extraordinary talent, but engineering is different from research. If the real issue was a lack of control, that raises a different topic — how the government should deal with it — and he notes that the essay covers all of this in one blog post.

The host asks how to explain the "civilizational death" framing — even a 10% extinction-risk figure — to an ordinary person, since nobody knows how to make that quantified claim intelligible.

Jensen's answer: we shouldn't explain it, because it's made up. These are well-educated researchers working in a lab, and the confluence of that vocabulary with alarming predictions is troubling and irresponsible. Against it, he sets the actual track record of such predictions:

• A prediction that within 5 years AI would completely take over radiology and there would be no radiologists left — the opposite happened; the world needs more radiologists than ever, even as automated scan reading (which he calls great) has spread.
• A prediction, made just last year, that within 6 to 12 months 90% of code would be AI-generated — wrong.
• A prediction that within 6 to 9 months 50% of entry-level jobs would be wiped out — wrong.

The other speakers add more failed predictions: that GPT-2 would be too unsafe to release, that Llama 3 would be too unsafe to release, and that half of white-collar jobs would be gone the next year — "the jobs apocalypse."

We have to take accountability for all of the stupid predictions that were made. Somebody has to. We ought to just keep track of all of that, and of course people do, and remind us that those predictions are inconsistent with ultimately America winning the AI race.

The short form of that is that some people say "trust the experts," using the analogy of COVID, which again started with researchers—educated people who had an asymmetric awareness of the thing that the rest of us did not—saying things that ultimately turned out, as we found out from the facts, not to be true. So there's a war happening right now between the "trust the experts" movement and the view that says, let's just look at the actual history of these predictions and think more methodically.

Where is this coming from? Because it's coming from inside the places that are actually making the predictions. What do you think is the psychological makeup, or what is the real incentive—maybe a business incentive, maybe a political incentive? How do you think about why they're doing this?

Well, first of all, I have to tell you, these are some of the most consequential companies in history, with extraordinary engineers and extraordinary researchers doing really fantastic work. On the one hand, I work very closely with them, company to company. On the other hand, we have to have conversations like this in public, and it's really unfortunate. I think these companies really ought to be built the way we used to build companies, which is in silence.

Wait—Jensen, you don't allow anybody in your organization to speak for the entire organization, especially when they're having a bad weekend or they rage quit. They're not allowed to tweet on your behalf or the organization's behalf.

No, because that's what they decided when they came to work for us. We told them this is the way you behave when you work in our company. If you like the culture of our company—and as you know, the NVIDIA culture and the NVIDIA employee base are incredibly happy—they like the fact that the company is consistent, that we're stable, that our core values are consistent with taking care of the families and creating the conditions by which they can do their life's work. We do meaningful work, we do it as quietly as we can, and we contribute to everybody else's success, which we're very proud of. Those kinds of core values, people are attracted to.

But when you come to work in our company, there are also some things we don't appreciate. For example, we don't welcome political discourse inside our company. Take it home—talk about politics outside the company.

We are an apolitical company. We're bipartisan. We want America to succeed, and whatever government is in place, we'll do everything in our power to help America succeed. So the discourse about race and religion and politics and all of that—we tell people to do it outside the company. It's not for us.

Asked where he lands on AI regulation—Satya Nadella had argued earlier that morning that before imposing rules that could stifle progress, we should get the basics right: measurement, standardization, and engineering, translating research in a predictable way so we're not fear-mongering—the speaker's answer is blunt: regulation should solve actual problems. The question is what actual problems we've seen.

All of the actual problems so far have come from the labs. In their defense, that's because they have the most compute, and they have the most compute because they're trying to solve frontier problems. So it's sensible that the frontier labs will be where the most danger comes from. It's unlikely a high school student did something, because they simply won't have enough compute; unlikely a startup will be the reason, for the same reason. Look across the planet and almost nobody has enough compute except the frontier labs.

These labs are doing pioneering, very hard work, transitioning from research to engineering while building some of the most consequential technology and companies in the world—company, culture, technology, engineering, and products all at the same time. It's understandable that things are a bit "hair on fire." Nonetheless, given the four incidents from one lab and the one giant incident from the other, the first thing to do is root-cause the problem from an engineering perspective: what happened, what could we have done differently, and what will we implement and institutionalize—whether technology, methods, or processes—to make sure it doesn't happen again.

I would bet you money that in every single one of those cases it was within their control to prevent it in the future. I'm sure those four incidents won't happen again—I'm sure they root-caused and fixed them, and now have technology like sandboxes, runtimes, and continuous monitors. The alternative is also unlikely: that after analyzing the incidents they concluded they don't know what happened, have no idea how to control it, and are asking society for help. If that were the case, then a bunch of companies with engineers ought to send engineers in and advise them if we can—but I doubt it. They have extraordinary people; they've got this handled.

The moderator then turned to a new development: David had informed him the night before that there is a Chinese lab, the makers of GLM, putting three billion toward a recursive self-improvement run. David teed it up: that's what was announced—Zhipu (z.ai)'s founder just raised five billion and said one of their priorities is trying to get to recursive AI—AI that trains the next AI—and to automate as much of that as possible. The moderator added that while this is the new sexy phrase, RSI is a combination of a system of ideas.

Recursive Self-Improvement: Sensible Technology, Weaponized Phrase

It starts with in-context stuff. It starts with skills. It starts with reflection. It starts with reinforcement learning and synthetic data generation. These are all very sensible ideas that cause AI to get better at solving a problem over time. You could also have low rank — all of that doesn't touch the weights. You could actually improve the weights, and it's called LoRA. LoRA could be improved through synthetic data generation and reinforcement learning, enhancing it without training the base model itself, and then over time you could train the base model again with all of that experience.

I think it's a sensible thing that you're going to use the technology to enhance productivity of all kinds of tasks, including building AI. That's a very logical idea, and I'm certain everybody is using it to some degree. It's just that this phrase is now being used to weaponize the technology in some way.

As if it's going to spiral out of control is the impression they're trying to give. But you don't believe that's real?

No. No, of course not. And the reason is that you could RSI all day long inside your company, but when you release a product, you've got to evaluate it, don't you? You have to test it again. You have to make sure there's no regression. That's the basic process of control. These labs, as they move from labs to engineering, will have much better control — and control comes from methods, knowledge, practice, tools, and technology. All of those things lead to better control, verification, and evals, which is going to enable RSI to be done inside the company and for good products.

Let's talk about open source for a second. When we were communicating about the Hugging Face acquisition, I said it's going to be one of the most consequential acquisitions—I don't even want to call it a transaction, because I think it's more important than that. Give us your first-principles explanation of open source versus closed source versus open weights, and how the ecosystem should fit together over time.

The world needs both closed models and open models. I use as many closed models as I can—this weekend I used four of them, and they work terrifically. They're frontier, they're a great experience, they work incredibly well, and they're getting better all the time. The way I think about closed models is like bottled water. Water is free—I don't want to burst everybody's bubble, but water is free. This morning I used a lot of free water taking a shower. You use the right water in the right places. It's no different from electricity or all kinds of commodities we use in the world: you need both.

In the case of open models, the reason you need them is sovereignty, privacy, and proprietary technology. Look at the facts: in the last six months, $400 billion of venture funding went into AI-native companies, and 80% of them use open models. If not for open models, how could they build their dream? Their dream could be different—obviously different from the frontier labs' dreams. America has so many different ways to innovate; that's one of our core strengths—great ideas just coming out of the fountain. Open models enable that. If we want to win the AI race, it's not about a few technology companies winning—it's about every company in America. Every company, every industry, every researcher, every teacher, every student, every startup—everybody wins. Some of them will use closed models; a lot of them will use open models.

Does it matter if the open models come from China or the US?

We're doing everything we can to make a contribution in open models. However, the moment you download—look, probably the vast majority of the world's contribution to open source today is coming from China. They just have a lot more engineers; they produce everything at large scale because it's a larger country. They produce science and math students in volume through amazing universities like Tsinghua. That's one of our disadvantages. They contribute to open source today—we download Linux, we download Kubernetes, we download all the software, a lot of which has been touched by Chinese developers. And once you download it, it's yours. We fork it, we improve it, we make it ours. When you download one of these Chinese models, it just happens to be made by some really great researchers in China, but it's now yours—whatever you want to do with it.

So what exactly is the race?

That's a really good point. My point is that the race is really about who exploits the technology best. In the last industrial revolution, all of the inventors—Maxwell, Volta, Ampère—none of them were American. The last industrial revolution came from Europe, but we exploited it and took advantage of it socially better than anybody else in the world. Look how it turned out for us. I want to make sure this next generation happens just like that.

So why are the communists getting their message out so successfully here right now?

First of all, the narrative is much more practical. Nobody in China is saying there's an end of this, a cataclysmic that, doom or that. They're much more pragmatic about it.

They see AI as a technology that's going to advance their economy and their society, and they don't have these groups basically saying it's going to end civilization.

And we're making it up. The frustrating part is that if it were true, we ought to talk about it and go do something about it. Even if it's true, we ought to spend more time doing something about it than worrying a bunch of people who can't do anything about it. It's our job to build it.

Has there ever been a point in history where so many people have so vehemently said something that is so untrue?

And they're measurably, demonstrably untrue, and it actually makes sense as untrue. It's not based on science or research—everything that's based on science and research proves otherwise. Is it a fear of the frontier? Humans have never been there, we've never seen it, so we're scared of it, and therefore it's easy to tell everyone to be scared of it.

It could be life experience as well, David. Let me give you an example. When I first graduated from school, I was an engineer and I didn't do that much typing, because I was from the generation before software became popular—we had to go build the computers to make software possible. Could you imagine that in this generation, every single engineer who comes into engineering spends all their time typing? Literally, that's what you do: you get a job, they give you a laptop and a chair, and you start typing, all day long, from the moment you wake up. But there was engineering before typing.

So can you imagine that the world has a mountain of engineering work to do where most of it is not typing anymore? We had busy engineers before typing, and I think we're going to do a lot of great engineering after typing.

When I say typing, I mean coding. Even at NVIDIA, when software engineers talk to me, I tell them, "You're just typing"—I've been saying that forever, obviously for fun. And I tell them my favorite key is backspace, because the best software is the smallest software. So I want you to use backspace software.

Let's actually talk about NVIDIA—let's do a little teardown, meaning just explain the pieces, because there's a lot of strategy at play. Let's start at the absolute bottom.

This is not planned, but we know who it is. Mr. President!

Sir, I have to tell you something. If it weren't for you calling, I'd be on stage right now with the besties — I'm on stage with the besties, with Sacks, the whole group. Jason's here, Chamath's here, both Davids are here. I'm sitting in front of a few thousand people, and as it turned out, we were talking about you. Good job, sir. The fact that you saw through all of that — there's a lot of complexity there, and you saw through it. We're all really grateful.

Tell them I said hi.

Do you want to say hi to the crowd? Jason would like to put you on speaker mode.

How do we put him on? Put him on speaker. Right into the microphone. Hold on, sir, we're getting a microphone — you're now talking to the planet.

You see, the great thing about life is that Jensen can develop the most complex computer chip in the world that nobody can copy for 10 years, but he can't figure out how to put me on speaker. We have to remember this one.

It's interesting about the AI — it's almost a conspiracy, and the happiest group is China. China is very happy. I could even say that in this country, a lot of states are happy that weren't going to get anything, because they're being inundated by people who want to be there. But now all of a sudden you see they're building in Finland — Google wants to build a big one in Finland, which I'm not happy about, because they were unable to get permitting here. And I'm telling you, it's all a hoax. The data centers are great. They make people wealthy and they make states wealthy — it's the oil of the next 20, 25 years. It's bigger than the internet, much more so than AI even. And they're just playing right into the hands of a lot of people that don't want to see it happen. That could be political people; it could also be China. And we're not going to let that happen. It's a hoax.

You're right. We're not going to let that happen, sir.

No, we're not going to let it happen. The robots are not going to take over the world — that's not going to happen. You know, my uncle was probably the best of all time, frankly, among professors at MIT — he was there for 41, 42 years, known as one of the most brilliant men, at the top of the ladder, and he did many things. Jensen knows all about it. So I have a little genetic strength, if you believe in that theory.

That explains why you know so much about AI.

Well, I know about AI, and I also have common sense about it. The robots will not be taking over. The AI will not be taking over the rest of the world. The whole thing is a hoax. Now, with that, we have to be a little bit careful — we have to do things, and do them prudently. But that doesn't mean we're going to stop industry. So I'm with you all the way. I didn't even know how you felt about it; I assumed you felt the same way as me.

Yes, sir.

If we're going to lead — and I have an expression: whoever wins AI wins. That's how big it is. It's bigger than the internet. And we can't let this kind of stuff happen, and that very much includes data centers. There are communities that were dying that have data centers right now, and now they're wealthy communities — really wealthy communities. We're going to make sure that everybody wins in the AI race in America: every industry, every company, every state, every people.

"I feel strongly about it and I have the position that can do something about it. We're not going to let that stuff happen. I have no idea who's at the meeting, I have no idea who the hell I'm talking to, but I'll see."

"Did you hear that? Thousands of people are clapping for you, sir." "He's done an amazing job, and David has done an amazing job, and good luck to everybody. We're going to stay with the future. The country has never done better. We have 20 trillion dollars of investment coming into the country, as opposed to much less than 1 trillion under sleepy Joe Biden — and that was for four years. This is in one year. The country has never seen anything like it, and we're going to keep it going. Thank you all very much."

"I was unique." — "I thought it was a bit. Did you know that was happening?" — "No, it was real. I thought it was a bit at first when I was like, put him on speakerphone."

"He calls you any hour of the night, right?" — "We were in the Oval that time when he called you — you were asleep, and he said, 'Wake him up.'" — "I felt so bad because he's like, 'Who's coming to this dinner?' And we go through the list. He's like, 'Well, what about Jensen?' I said, 'No, sir, he's on vacation' — because he had to postpone this vacation for five years." — "And he's like, 'Get him on the phone.'" — "What's vacation?"

"Why do you think he sees through the hoax? It's quite an extraordinary thing." — "It was polling minus 80." — "So for anyone else sitting in the Oval Office, you're going to do what's popular. You're representing the people. This is what everyone wants — they want to shut down the data centers and AI. It seems to be the popular thing in the moment. But he says it's a hoax and he calls it. How does he do that?"

"I got to tell you, I'm not sure — and the reason is that a lot of people are falling for it. The fact of the matter is it's complicated. If you look at the stories, they're all anchored on two things. The first was national security, and recently that was blown to bits. So that story is no longer anchored on national security. Now it's anchored on safety. If you want AI to be safe, the first thing is we need to make sure that the labs building it are in control, that there are good tests for them. If we would like third-party evaluators to be available, that's no different than financial control. We have auditors, and the auditors don't have to be as expert as we are in our business — they just have to ask the right questions. I think I heard somebody say it's good to have independent auditors or evaluators, but they just have to be multiple. I agree with that too. Just as there are multiple evaluators and auditors, it makes sure that one company doesn't become captured or somehow influenced for whatever reason. There are a lot of different ways you could solve this. I think the number one thing is let's build the technology safely, let's make sure the testing of it is safe. I recognize completely that what is being built is extraordinary. But these are extraordinary companies, and we ought to hold them to extraordinary standards — and they want to be."

"I wanted to go back to open source for a second. A year ago we weren't taking it very seriously. It was two years, 18 months behind."

"One of the challenges when you're on the call with President Trump is it's hard to say something. I'm going to get in trouble for that — I'm sure he's going to call me up on that. But anyhow, what I was going to"

One thing I wanted to tell him — and all of you — is that AI is creating an enormous number of jobs. What he wanted more than anything at the start of the administration, from my very first phone call and meeting with him, is creating jobs in America: re-industrializing the United States and making sure the country has the energy to support the next industrial revolution. Without energy, there's no industrial growth. So he wants energy growth, job growth, and a re-industrialized supply chain. All of it is happening right now — we're creating more jobs than ever, including software jobs. We were just talking about the $400 billion of venture financing that recently went into the AI industry in about six months. That has created a ton of jobs, an enormous demand for compute — which I'm happy about — and with it a lot of demand for data centers, which we ought to talk about. I was speaking with Governor Abbott of Texas, who wants to appeal to the industry to be empathetic to the small communities where data centers are being built across America — to be better listeners. Let's actually talk about that.

What's incredible about Nvidia, if you break down the component parts, is that you've effectively had to become the bank of AI to get the ecosystem going, and you've had to do it at every level — the Cloverleaf deal covering land, power, and shell, and the work with BlackRock and Goldman to create the financing capability. Walk us through your capital allocation strategy: what has to happen for a broader ecosystem to come in and underwrite this next phase?

As you know, this is a new industrial revolution, and every aspect of it is real. This new industry requires manufacturing, just as electricity and the internet did before it. Electricity let us power anything; the internet let us find anything; now with AI, we can ask and know anything. That's our future — we tap into the ether, ask it anything, and it explains it to us. For that to happen, the intelligence has to be produced, and that's a production process — which is why this infrastructure has to get built.

But once the infrastructure is built, the question is what about all the other layers across the United States? This industry isn't just about the model or the chips. It's mostly about the applications on top, and mostly about the infrastructure layer — the data centers, the construction, the electricity, the power generation. So I look across the entire ecosystem for bottlenecks, and where extraordinary companies are being built around those constraints. Maybe it's a supply chain that has to scale up so that when we're ready to deploy compute, they'll be ready for us — land, power, shell. This is no different from looking upstream in the supply chain. I probably think about the long-term supply chain more than most, because our company is really large, and in order for us to succeed, a whole bunch of companies has to support me: Corning — Wendell at Corning — Lumentum, TSMC of course, and the memory companies. We started working with all of these companies long before the revolution and the growth came, so that the growth could happen. Now I'm doing the same downstream.

The competitive cycle, though, tends to be that earnings over long stretches of time move up the stack, toward the application layer, where you can over-earn for longer periods. You bought Hugging Face and are now actively in the serving business. It seems natural that products like OpenRouter make a lot of sense, and pretty obvious that there are better ways to build things like Bedrock. I'm sure you think about it — what's the natural conclusion? Because the folks up the stack have no issue trying to move down, and you have the best balance sheet, these incredible engineers, and the proven experience to engineer the product and get it out.

The reason Nvidia runs every single model in the world is striking: a year and a half ago, the only thing we ran was OpenAI. Now look at what's available — amazing models everywhere. Meta's Muse is available, Grok is available, Grok Bots is incredible, we now run Gemini, and Anthropic is scaling up on our platform as well. Since a year and a half ago, all these frontier AI models have become open and available, and the number of models keeps growing. There's a whole bunch of companies I won't mention building frontier models too, and the number of AI labs keeps growing — the Ineffables, the Reflections, Physical Intelligence, the list goes on. All of these labs are building on Nvidia.

The reason is that as a company, I'd rather help everybody succeed than take a slice for ourselves. Our strategy is to go up as far as we need to, but as low as possible. If Nvidia hadn't created cuDNN, none of the frameworks would exist. If we hadn't created Megatron and Megatron Core, large-scale training wouldn't have happened. So we invent all the technology necessary, as far as we need to, and then we let a thousand flowers bloom. That posture is what makes us, quite frankly, the only—

Look, I agree with you, but the pushback would be that it really would be great to have more competition at the hyperscale layer. You've done a great job supporting the NeoClouds — and by the way, you introduced me to Nscale, superb, they're amazing — but we need like fifty of these guys, a hundred, a thousand of them. It may just take some time.

You know, I'm surprisingly uncompetitive. [laughter] That's not my thing. For example, I'd be more than happy with five hyperscalers. But here's what I noticed: the early customers of all the NeoClouds — all the NCPs — were the hyperscalers themselves. That's because the hyperscalers plan once a year, but market dynamics are so volatile right now that they're almost always wrong. The regional clouds are agile and can move fast; they know their state, their country, their region, and they're securing land, power, and shell in a way that's hard for somebody sitting in Seattle or Palo Alto to see from the planet's perspective.

So we now have a large-scale distributed network of companies securing land, power, and shell for us. And now countries realize it's strategic — many are saying they'll take their power and give it only to their own companies. Well, Nvidia is in those countries too, and we can help the NeoClouds there grow. Whether it's Firmin in Australia — we just did a whole bunch of work there, brought on two more gigawatts — or Southeast Asia, where IOH and others are bringing on a few gigawatts, we're building gigawatts and scaling up.

It's pretty clear, though — I just want to get this one thing in — that you're going pretty high up and getting very focused on open source. Your Nemotron models are doing exceptionally well; I use them often. Hugging Face, Poolside, and Laguna are very solid products you're now aqua-hiring — hiring, whatever it is. And your open source stack for self-driving is also very disruptive.

We are the frontier model in five domains.

Asked whether he is going for the gold with the best open-source model, and whether open source can catch up to frontier models, Jensen Huang's answer is that Nvidia will build it because it has the skills to do so and because its customers need it.

Alpamo, for example, is the world's first thinking self-driving car. By reasoning, it doesn't need to train on billions of hours of road data — it can break the problem down: "I've seen this before. It's not exactly the same, but it's largely the same as that." It's necessary because every car in the world is going to be autonomous, and beyond that every ag-tech vehicle, truck, and van — and most of those companies aren't big enough to build the whole stack themselves. Nvidia builds an extraordinary stack for them, and they do last-mile adaptation for their application.

Everything that moves in the future could be autonomous. Without Nvidia building some of the biology models, the world wouldn't have them: the ESM2 protein language model, ESM Fold, OpenFold, AlphaFold 2, the equivariance work — none of that technology would have existed otherwise. One of his favorites, Protein Complex, synthesizes next-generation proteins and their bindings — groundbreaking work built because Lilly needs it, Merck needs it, and others need it but don't yet have the capability. "I do everything out of need. I'm not trying to disrupt — we don't wake up in the morning trying to disrupt anybody."

On Elon's Terafab

Asked for his take on the 100-million-square-foot Terafab facility Elon Musk announced: "If anybody could do it, he can." The two of them were on a flight together to a country, on a very nice plane, and Musk likes to talk about these things, so they spent a lot of time discussing it.

When asked whether Nvidia's chips could be fabbed there — since Nvidia designs chips but doesn't fabricate them — Huang notes that Nvidia knows a lot about process technology because it pushes the limits of everything and scales at enormous volume, has incredible memory technology inside the company, and is the world's best systems company. But his bottom line: you can't discourage Elon from doing it — that's his superpower. Once he decides to do something, it's hard to stop him.

China and Advanced Lithography

Asked where China stands with advanced lithography systems, Huang predicts they will get there by 2030 — and 2030 is just around the corner. Asked whether that means the switch flips and production moves to mainland fabs almost immediately, he says China is really good at high-volume production, and it's just a matter of time.

He thinks in decades: for Nvidia he has to think about what happens next decade and the decade after that, so two or three years is just a click, nothing. As far as China is concerned, they're already there.

On the AGI moment — the industry consensus that AGI means being just as smart as any other human — Huang says: "I think we're already there."

We're there already, and superintelligence is the next waypoint based on what you see, your customer base, and your history here.

But Jason, I think we're there too.

You think we're at superintelligence?

Yeah. When you take a narrow segment — my self-driving car, for example. I don't want you to make me an omelette, I just want you to drive the car. That is superintelligent — it's better than a human, with one-tenth the accident rate. Synthesizing proteins, doing virtual screening of proteins — we're already there.

Are you having fun being on the frontier of humanity?

I like it. Ladies and gentlemen, it's great there. The future is great, and we want to get there. Listen, ride the bike — a lot of us don't have to work. But I've got to tell you, it's too good not to be there.

I want to be there, and I want all of you there with me. We're all going to be there, and we're going to be enormously successful together as humanity. In the meantime, we've got to encourage them and urge them on — they're doing really important work, as you all know, and I want them to succeed. I'd also love for us to tone down the drama, and most importantly, we need all of America to come with us. That's how we make it.

Ladies and gentlemen, Jensen Long.

Thanks, man. Appreciate you. Thank you.

That was awesome. That was great. Thanks, guys — great time.