Joe Rogan podcast. Check it out. The Joe Rogan Experience. Train by day. Joe Rogan podcast by night. All day. Hello. Hey, Joe.
Good to see you again. We were just talking about... Was that the first time we ever spoke? Or was the first time we spoke at SpaceX?
SpaceX. The first time when you were giving Elon that crazy AI chip, right? DJX Spark.
Yeah. That was a big moment. That was huge. It felt crazy to be there. I was watching these wizards of tech exchange information, and you're giving him this crazy device. Then the other time was I was shooting arrows in my backyard and randomly get this call from Trump, and he's hanging out with you. President Trump called and I called you. We were just talking about you.
He was talking about the US UFC thing he was going to do in his front yard. He pulls out this design. He's so proud of it. I go, 'You're going to have a fight in the front lawn of the White House.' He goes, 'Yeah, you're going to come. This is going to be awesome.' And somehow your name comes up. He goes, 'Do you know Joe?' I said, 'Yeah, I'm going to be on his podcast.' He says, 'Let's call him.' He's like a kid. He's so incredible. He's an odd guy, very different from what you'd expect. Also, he makes the text go big on my iPhone, like 'USA is respected again' in all caps. It's kind of ridiculous.
President Trump is very different. First of all, he's an incredibly good listener. Almost everything I've ever said to him, he's remembered.
People only want to look at negative stories about him. There's a lot of things he does where I don't think he should, like saying 'Quiet piggy' to a reporter. That's pretty ridiculous, though objectively funny. I wish he didn't do that. But part of his charm, part of his genius, is he says what's on his mind, which is anti-politician. Some people would rather be lied to, but I like that he tells you what's on his mind. He starts with his love for America and what he wants to do for it. Everything he thinks through is very practical and common sense.
I still remember the first time I met him. I'd never known him before. Secretary Lutnick called, and we met right at the beginning of the administration. He told me what was important to President Trump: that the United States manufactures onshore. That was really important to him because it's important for national security. He wants critical technology built in the US and to re-industrialize for jobs. It seems like common sense. That was literally the first conversation I had with Secretary Lutnick. He started with, 'Jensen, you're a national treasure. Nvidia is a national treasure. Whenever you need access to the president, you call us.' And it was completely true. Every time I called, they were always available. It's unfortunate we live in such a polarized society that you can't recognize good common sense if it comes from someone you object to.
It only makes sense that we have manufacturing in America, especially for critical technology. It's insane that we buy so much from other countries. If the US doesn't grow, we have no prosperity. Without energy growth, we can't have industrial growth or job growth. Trump came in and said 'drill baby drill.' His point is we need energy growth. That saved the AI industry. Without his pro-growth energy policy, we wouldn't be able to build factories for AI or chip factories. All those construction and electrician jobs would be challenged. He's got it right. We need to re-industrialize. Not every successful person needs a PhD or to go to Stanford. That sensibility is spot on.
We're in the middle of a giant technology race. Whether people like it or not, it's happening. Whoever gets to the event horizon of artificial intelligence first has massive advantages.
We are in a technology race, and we've always been in one. Since the industrial revolution, the Manhattan Project, even the discovery of energy. The UK invented the industrial revolution, but the US capitalized on it, industrialized it, and diffused it faster. We took the technology and ran with it. World War II was a technology race, the Cold War was. Technology gives you superpowers—information, energy, military. Technology leadership is really important.
The problem is if somebody else has superior technology. With the AI race, people are nervous. Elon has famously said there's an 80% chance it's awesome, 20% chance we're in trouble. People are rightly worried about that 20%. If you had a revolver with two bullets and spun it, you wouldn't feel comfortable pulling the trigger. It's terrifying. It's impossible to imagine it wouldn't be of national security interest to get there first. But the question is, what's there? I'm not sure, and I don't think anybody knows. If you're the head of Nvidia and you don't know, who knows?
I think it's probably going to be much more gradual than we think. It won't be a moment where somebody arrives and nobody else has. Things will just get better and better, like technology does.
So you're rosy about the future? You're optimistic about AI?
If history is a guide, humanity has always been concerned about new technology. That concern gets channeled into making it safer. In the last two years, AI has become maybe 100 times more capable. We directed that power to make AI think—break down problems step by step, do research before answering, ground itself in truth, reflect on its answers. If it's not certain, it goes back and does more research or uses a tool. We took all that computing capability and channeled it into safer, more truthful results. The amount of hallucination has reduced dramatically. Most people think of power as an explosion, but technology power is channeled towards safety. A car today is more powerful but safer to drive. More horsepower gives better handling and control. The next thousand times of AI performance will be channeled towards more reflection, research, and deeper thinking.
So when you define safety, you mean accuracy and functionality?
Functionality. It does what you expect. You put guardrails on it, like ABS and traction control in cars. That little computer is more powerful than the one that went to Apollo 11. When people talk about power, they often think of sci-fi movies or military power. But technology power is about more refined thinking, reflection, planning, and options.
A big fear is military applications—AI systems making decisions that an ethical person wouldn't.
I'm happy that our military is using AI for defense. Anduril and other tech startups are channeling their capabilities towards defense. We need that. It's becoming more socially acceptable. There was a time when people were vilified for working on defense technology, but we need those people.
People are terrified of war. The best way to avoid it is excessive military might.
All of it—military might, diplomacy. You need military might to get people to sit down with you. Otherwise, they just invade. History shows that.
When you look at the future of AI, do you ever ponder scenarios? What's the best-case scenario over the next two decades?
The best case is that AI diffuses into everything we do, making everything more efficient. But the threat of war remains. Cybersecurity will be a super difficult challenge. You'll have millions of AI agents protecting you, but their technology will also get better. Right now, we're seeing cyber attacks everywhere, yet we're sitting here talking because of all the cybersecurity technology in defense. We just have to keep amping that up.
A big worry is that encryption will become obsolete. Do you anticipate that being an issue, or will defense always keep up?
Not forever. Some intrusion will get in, and we'll all learn from it. The reason cybersecurity works is that the community works together. Most people don't realize that cybersecurity experts share ideas, best practices, and patches the moment something is breached. That cooperation started about 15 years ago because people recognized no company can stand alone. The same will happen with AI. Working together to stay out of harm's way is our best defense.
So it's everybody against the threat. You'd be way better at detecting and neutralizing threats.
Exactly. The moment you detect it somewhere, it's shared. That's why it's safe. I'm not just watching my own back; everybody's watching my back, and I'm watching theirs.
When people talk about AI threats, they should think about how we deal with cybersecurity today. AI is a new type of software, but so will the defense be.
No question. AI is new, but the defense will use the same AI technology.
Do you anticipate a time when there are no secrets? Information is just ones and zeros, and technology has more access. Will it ever be impossible to keep a secret?
I don't think so. Quantum computers will make previous encryption obsolete, but the industry is working on post-quantum encryption technology—new algorithms.
Quantum computing can do computation that would take billions of years in minutes. How do you make encryption for that?
I'm not sure, but we have scientists working on it. I don't think it will get to a point where we can't keep secrets. History is a guide.
My worry is that AI is different. Nuclear weapons changed our thoughts on war. But with AI, we're getting better every day because we use it. So when a threat comes, it's a click ahead, not a galaxy ahead. I think the idea that AI will pop out of nowhere and think in ways we can't imagine is far-fetched. We're all using AIs, and we're close to each other.
But don't they do surprising things?
Yes, but my AI will look at your AI and say that's not that surprising. The fear for lay people is that AI becomes sentient, makes its own decisions, and decides to govern the world.
But my AI will take care of me. That's the cybersecurity argument. If your AI is super smart, my AI is super smart too. Let's pretend we understand consciousness and sentience—I don't actually believe that, but let's pretend. If your AI is conscious and wants to do something surprising, my AI is so smart it won't be surprised. The idea that only one person has AI and everyone else's is Neanderthal is unlikely. It's more like cybersecurity.
The fear is that AI won't listen to you anymore. That human beings won't have control if it achieves sentience and autonomy.
If it's like a life form, all life forms don't agree. My life form will want to be the super life form. We'd be back to where we are.
