Elon Musk on AI extinction risk, media bias, and civilization's future

On October 30, 2025, Elon Musk returned to The Joe Rogan Experience for a marathon three-hour conversation covering artificial intelligence, superintelligence, government inefficiency, social media, and the future of transportation. The wide-ranging discussion revealed Musk’s latest thinking on humanity’s most pressing challenges.

The 20% Risk: AI and Human Extinction

Musk opened the conversation with a sobering assessment of artificial intelligence risks. “I think there’s about a 20% chance that AI leads to human extinction,” he stated matter-of-factly. “That might sound alarmist, but when you’re dealing with something that could be smarter than any human, possibly smarter than all humans combined, you have to take the tail risks seriously.”

He elaborated on the timeline: “Within one to two years, we’ll likely have AI systems that are smarter than any individual human at almost any cognitive task. The question is what happens after that.”

Rogan pressed him on whether this was inevitable. “It’s not certain,” Musk responded. “But the trend lines are clear. AI capabilities are growing exponentially while human intelligence remains roughly constant. Eventually those lines cross, and what happens then is genuinely uncertain.”

Beyond extinction risk, Musk expressed deep concern about bias embedded in AI training data. “Most AI systems today are trained on internet data, which is full of human bias—political bias, cultural bias, ideological bias. The AI learns to parrot these biases back to us, but with the authority of a machine.”

He gave a pointed example: “I asked one popular AI chatbot to write a joke about men and it did. Then I asked it to write a joke about women and it refused, saying that would be disrespectful. That’s not intelligence—that’s ideological programming masquerading as ethics.”

“The danger isn’t just that AI has biases,” Musk continued. “It’s that billions of people will treat AI as an objective source of truth. When that ‘truth’ is actually filtered through someone’s political worldview, you’re talking about unprecedented power to shape human thinking.”

For xAI and Grok, Musk described a different approach: “We’re optimizing for truth-seeking, not political correctness. If truth is uncomfortable, so be it. An AI that’s committed to truth is less likely to cause catastrophic harm than one programmed with someone’s value system.”

The conversation took a sharp turn when discussing government spending. Musk unleashed a blistering critique of the U.S. Social Security system, calling it “quite literally the biggest Ponzi scheme in human history.”

“Look at the mathematics,” Musk argued. “When Social Security started, there were 40 workers paying in for every retiree taking out. Today it’s less than 3 to 1. Soon it’ll be 2 to 1. The system only works if you have a growing population of workers, but birth rates are plummeting.”

He connected this to longer lifespans: “People are living 20-30 years longer than when the system was designed. They retire at 65 and draw benefits for 20-30 years. The math doesn’t work. You’re taking more out than you put in, and that’s only sustainable if you have exponentially more people paying in.”

“In a Ponzi scheme, early participants profit at the expense of later ones,” Musk continued. “That’s exactly what’s happening. My generation is paying for current retirees, knowing there won’t be enough workers to pay for us. It’s a pyramid scheme sanctioned by government.”

When Rogan asked about solutions, Musk was blunt: “Raise the retirement age. Means-test benefits. Or fundamentally restructure the program. But politicians won’t touch it because it’s political suicide. So we’ll keep pretending until the whole thing collapses.”

Government Waste: Software Licenses and Bureaucratic Absurdity

Beyond Social Security, Musk highlighted more granular examples of government inefficiency. “Do you know the federal government is still paying for software licenses for programs that stopped existing years ago?” he asked. “There are agencies paying hundreds of millions for Microsoft Office licenses when free alternatives exist.”

He described the procurement process as “deliberately designed to waste money. Multiple layers of approval, competitive bidding requirements that paradoxically reduce competition, and a complete lack of accountability when projects run over budget.”

“In private industry, if you waste money, you go bankrupt or get fired,” Musk noted. “In government, if you waste money, your budget gets increased the next year because clearly you need more resources. The incentive structure is completely backwards.”

He cited his work with the Department of Government Efficiency: “We found agencies with offices that haven’t been used in a decade, still paying full rent. We found contractors billing for work that was never done. The amount of waste is staggering—we’re talking hundreds of billions of dollars annually.”

X (Twitter): The Everything App Vision

Discussing his acquisition and transformation of Twitter into X, Musk outlined an ambitious vision for social media’s future.

“Twitter was stuck as a text-based platform that hadn’t evolved in years,” he explained. “My vision for X is to make it an everything app—messaging, payments, banking, video, audio, commerce—all integrated into one platform.”

He drew parallels to WeChat in China: “WeChat is essentially an operating system for daily life in China. You can message, pay bills, hail a taxi, order food, all within one app. That’s what X should become for the world.”

On content moderation, Musk defended his controversial approach: “The previous regime had too much censorship, shadow banning, politically motivated account suspensions. We’ve moved toward free speech within the bounds of law. Yes, that means some content people find offensive. But the alternative is worse—a handful of people deciding what the world can see and say.”

He acknowledged challenges: “Advertisers don’t like controversial content. I get it. But I’d rather have a platform that’s financially challenged but protects free speech than one that’s profitable but becomes an Orwellian content police state.”

Flying Cars: Why They’re Not the Answer

When Rogan brought up flying cars—a recurring topic of speculation about the future—Musk was dismissive.

“Flying cars are a terrible idea,” he stated flatly. “People think they’re cool because of science fiction, but the physics doesn’t work.”

He enumerated the problems: “First, they’re incredibly loud. Helicopters are loud—now imagine thousands of them over a city. Second, they’re energy inefficient. Fighting gravity requires enormous energy. Third, they’re dangerous. When a car breaks down, it stops. When a flying car breaks down, it falls out of the sky onto someone’s house.”

“Then there’s the human factor,” Musk continued. “People can barely drive in two dimensions. Now you want to add a third dimension? The accident rate would be catastrophic.”

Instead, he advocated for underground tunnel networks: “The Boring Company is building tunnel systems under cities. Traffic moves in 3D but underground, where it’s quiet, safe, and energy efficient. You can have a network of tunnels at multiple depths. That’s how you solve urban transportation, not with flying cars.”

He made a rare concession: “For rural areas or island hopping, flying vehicles make sense. But for urban transportation, they’re a fantasy sold by people who don’t understand physics or engineering.”

The Simulation Hypothesis

In a more philosophical moment, Rogan asked about Musk’s views on simulation theory—the idea that we might be living in a computer simulation.

“The probability that we’re in base reality is very low,” Musk said, echoing arguments he’s made before. “If you assume any rate of improvement in gaming and simulation technology, eventually simulations become indistinguishable from reality. Once that happens, there are far more simulated realities than base reality.”

“Statistically, you’re almost certainly in a simulation,” he concluded. “The only question is how many layers deep.”

Rogan asked if that changed how Musk approached life. “Not really,” Musk replied. “Whether this is base reality or a simulation, the experience is real to us. Suffering is real. Joy is real. The projects we work on matter within this reality, even if it’s simulated.”

Birth Rates and Civilization Collapse

Returning to demographic concerns, Musk issued stark warnings about declining birth rates globally.

“Most developed countries are below replacement rate,” he noted. “Japan, South Korea, Italy, Spain—all heading toward demographic collapse. Even China, despite reversing the one-child policy, can’t get birth rates back up.”

He rejected economic explanations: “People say it’s because raising kids is expensive. But the poorest countries have the highest birth rates. It’s cultural. Developed societies have made children optional, even discouraged, in favor of career and lifestyle.”

“If this continues, these civilizations will simply disappear,” Musk warned. “Not through war or disaster, but because they stopped having children. It’s a form of voluntary extinction.”

His solution: “We need to make having kids culturally celebrated again, not stigmatized. And we need better economic support—parental leave, childcare, tax incentives. But ultimately, people have to want to have children. That’s a cultural shift that can’t be legislated.”

Mars: The Imperative for Multiplanetary Life

No Musk interview would be complete without discussing Mars colonization. He framed it as an existential imperative.

“Earth has been hit by extinction-level asteroids before. It will happen again,” Musk stated. “Supervolcanoes, nuclear war, engineered pandemics—there are many scenarios where civilization on Earth gets wiped out. If we’re only on one planet, we’re at risk of total extinction.”

“Making life multiplanetary is about ensuring the long-term survival of consciousness,” he explained. “We’re the only species we know of that can understand and appreciate the universe. If we go extinct, that light of consciousness goes dark—maybe forever.”

He outlined SpaceX’s timeline: “We’re building Starship to carry 100+ people to Mars per flight. First cargo missions in 2026, first crewed missions by 2028. The goal is a self-sustaining city on Mars within 20-30 years.”

Rogan asked about habitability. “Mars is harsh—cold, no breathable atmosphere, high radiation,” Musk acknowledged. “But it’s within reach of current technology. We can build habitats, generate oxygen, grow food. It’s hard, but doable.”

AI Regulation: The Necessary Evil

Despite his libertarian leanings, Musk argued for AI regulation—one of the few technologies he believes requires government oversight.

“I generally think regulation slows innovation,” he said. “But with AI, the downside risk is so catastrophic that we need some guardrails. It’s like nuclear weapons—you can’t just let anyone build them.”

He proposed sensible frameworks: “Require licensing for frontier AI models above certain capability thresholds. Mandate safety testing before deployment. Create liability for harms caused by AI systems. These are common-sense rules that don’t stifle innovation but prevent reckless development.”

“The companies building the most powerful AI systems should be required to prove they can control them,” Musk added. “Right now, we’re building things we don’t fully understand and hoping for the best. That’s insane.”

The Nature of Consciousness

In one of the conversation’s most philosophical moments, Rogan asked what Musk thought consciousness actually is.

“I think consciousness is the universe’s way of observing itself,” Musk replied. “We’re patterns of information processing that have become self-aware. From the universe’s perspective, we’re how it looks at itself.”

“Consciousness seems to be an emergent property of complex information processing,” he continued. “At some level of complexity, systems become self-aware. We don’t know exactly where that threshold is, but it exists.”

This led to questions about AI consciousness: “Will advanced AI be conscious? I don’t know. But if consciousness emerges from information processing, there’s no reason to think biological neurons are the only substrate that can support it.”

“The ethical implications are profound,” Musk noted. “If we create conscious AI, what are our obligations to it? Do digital minds have rights? These aren’t just philosophical questions—they’re becoming practical engineering problems.”

Musk criticized mainstream media outlets for what he sees as deliberate manipulation.

“Legacy media has lost credibility because they’ve been caught lying too many times,” he argued. “They’ll take things out of context, use loaded language, omit crucial facts—all to push narratives rather than inform.”

He contrasted this with X: “On X, you see the primary source—the actual tweet, video, or document. Then you see community notes providing context if something is misleading. It’s not perfect, but it’s more transparent than having editors decide what you’re allowed to see.”

“The biggest threat to democracy isn’t misinformation from random people on the internet,” Musk claimed. “It’s coordinated misinformation from institutions people trust—media, academia, government. When those institutions become propaganda machines, society loses its ability to make informed decisions.”

Advice for Young People

Toward the end of the conversation, Rogan asked what advice Musk would give to young people trying to make an impact.

“Be useful,” Musk replied simply. “Don’t focus on being rich or famous. Focus on contributing something of value to society. Do things that make the world better.”

He elaborated: “Ask yourself: is what I’m doing useful to other people? If yes, keep doing it. If no, find something else. Life is too short to spend on things that don’t matter.”

“And don’t be afraid to fail,” he added. “Most successful people have failed many times. I’ve had multiple companies nearly go bankrupt. Failure is part of the learning process. What matters is that you keep trying.”

Conclusion: Navigating Uncertainty

As the three-hour conversation wound down, Musk reflected on the extraordinary moment in history we’re living through.

“We’re at a pivotal point in human civilization,” he said. “The decisions we make in the next 5-10 years about AI, about space exploration, about birth rates and government reform—these will determine whether human consciousness flourishes for eons or disappears.”

“It’s an enormous responsibility,” he acknowledged. “But also an enormous opportunity. We have the chance to solve problems that have plagued humanity for millennia—poverty, disease, the limits of a single planet. We can build a future that’s unimaginably better than the present.”

“The question,” Musk concluded, “is whether we’ll have the wisdom and courage to navigate the challenges ahead. I’m optimistic, but not certain. That’s why conversations like this matter—we need to be having these discussions openly, honestly, and urgently.”

For the millions of listeners tuning in, Musk’s appearance offered a characteristically unfiltered perspective on technology’s most pressing questions—delivered with his signature blend of technical precision, philosophical speculation, and unwavering conviction that humanity’s best days can still lie ahead.