Professor Brian Greene: The Threat of AI, Consciousness & The End of Humanity

I can't convince you that we're not in the matrix. Why? We could be. Is this table real? Am I really here? Am I a simulation sitting inside some computer? Maybe there's a kid in 29th-century Earth who happens to allow you and I to exist, sitting at this table.

And here's a big assumption: can you create a computer program that has consciousness? Nobody knows the answer to that. But I've spoken to some of today's AI researchers who think that ChatGPT is conscious, and in the back of my mind, I do think it's possible.

But if we considered robots to be conscious, would they then get rights?

On whether conscious robots should get rights: I think that's something we should think about now.

The Ultimate Question

We find ourselves thrust on this planet, trying to figure out why we're here — is there some ultimate purpose? We're a little dot on planet Earth, and then you see our sun as one of many stars in the galaxy, and our Milky Way galaxy within the setting of the entire universe.

Asked what's beyond: if the universe is infinite, then going sufficiently far out into space, you're virtually guaranteed to find a copy of yourself out there.

Should we feel insignificant? I would urge you not to, because the fact that we can understand how small we are is what gives us a sense of connection to something larger.

On whether this could lead us to God: if you mean the kind of God our forebears often spoke of, that feels to me very much a human creation, and I'll tell you why I'm skeptical of that. But another thing we should think about is the end of the Earth — if we go arbitrarily far into the future, the Earth is going to spiral.

Professor Brian Greene, what is the ultimate question that you're seeking to answer?

The ultimate question, I would say, is: what is the true nature of reality? We find ourselves thrust onto this planet, and we look out and try to figure out why we're here and what it's made of. Is there some ultimate purpose, or is it just some interesting accident that allowed us to have self-awareness and ask these kinds of questions about the nature of the world? To me, that is the deepest kind of question we humans can ask.

Do you feel like over the last 30 years of your career, you've gotten closer to understanding the answer to that question?

Yes, but it's kind of like picking up grains of sand on a beach. You pick up more and more, and that pouch of sand gets heavier — so your understanding is deeper, it's weightier. But you look out and there's almost infinitely more grains of sand out there. We understand the world better than we once did, but does that really take us some significant way toward assessing the big, deep questions — say, why is there something rather than nothing? Maybe that's the deepest question of all, and by nothing I don't mean empty space; I mean nothing, nothing at all.

In this conversation, we'll pursue that answer by learning who you are and the journey you've taken academically and through your career. Where did you study, and where have you worked over the last couple of decades?

I grew up in Manhattan, right across from the Hayden Planetarium. I would wander its hallways and come upon a meteorite, and the universe became something like my backyard. That set me on a trajectory I continued in college: I went to Harvard and studied mathematics and physics, then to Oxford, where I earned my doctorate working in a field called string theory. From there I climbed the academic ladder through Cornell and ultimately Columbia, where I've now been a professor for some thirty-odd years and direct the Center for Theoretical Physics at Columbia University.

Asked to explain string theory to a 12-year-old, the physicist offers this: matter is made of molecules, which are made of atoms stuck together. Atoms, in turn, are made of smaller things — electrons orbiting a nucleus of neutrons and protons — and those particles are made of even smaller ones called quarks.

The question is whether that's the end of the story. Is everything just electrons and quarks, or might there be a finer level of structure inside them? If you took a powerful microscope and looked deeply, would you see anything else?

The idea of string theory is that you would. Inside an electron or a quark, you'd find a tiny vibrating filament, and the different vibrations of the string yield the different particles. Like a violin string that sounds as a C or an A sharp depending on how it vibrates, a string in string theory vibrating one way is an electron, and vibrating another way, a quark.

So at its most fundamental level, matter would reveal all these little vibrating filaments — vibrating strings — if you could look deeply enough inside any piece of it. That's the idea of the theory.

The interviewer asked two questions: how do we know this, and what would we see if we zoomed in even further?

On how we know: we don't. Everything just described could be utter nonsense, though the speaker doesn't think it is. The only way to know if an idea in science is correct is to make predictions that can be tested and measured to confirm the ideas. We don't have the technology to look sufficiently deeply into matter to see the strings, and we don't have particle accelerators or colliders powerful enough to slam things together and access this finer level of structure that the theory suggests.

So where do these ideas come from? Mathematics. The power of mathematics is that it can take us to places we can't yet go literally using equipment. When the mathematics of quantum mechanics and the mathematics of general relativity are put inside the math of string theory, suddenly the equations work and come together harmoniously. We're confident in the math of quantum mechanics because it makes predictions confirmed by observations, and we're confident in the math of general relativity for the same reason. So we're willing to go where the math takes us—and the math takes us, in this approach, to these vibrating filaments, these vibrating strings.

The prediction is that if you zoom into every cell in my body, you'll find differently shaped strings vibrating.

What causes them to vibrate in a certain way? The idea starts with recognizing that the universe has energy. If you ask me where that energy comes from, nobody can answer that—that is the question of why there is something rather than nothing. The "something" is a fundamental currency called energy. If string theory is correct, that energy manifests as the vibrational patterns of the strings themselves. If the theory is wrong, the energy manifests through other things, like the motions of particles and the way spacetime warps and curves.

As for who mandates that one string vibrates in one pattern and another string in a different pattern: if we go back to the Big Bang—the energetic explosion that resulted in space expanding—the detailed way in which that happened ultimately dictates how each string vibrates in the aftermath of that rapid cosmological event.

Does this string theory idea tell us anything about why we're here? I would like to think that it does, though very indirectly. We are aggregates of a huge number of particles—or, if string theory is correct, of a huge number of strings. Inside your body there are something like a billion billion billion particles, and if string theory is correct, each of those particles is a vibrating filament. It is a gargantuan challenge to understand why those billion billion billion particles came together to yield you or me or somebody else.

But that is a historical progression, and I think we understand the broad outlines of that narrative. There was some initial event called the big bang that resulted in space expanding. The energy of that big bang ultimately transmuted into particles, and those particles were wafting through space. Gravity caused collections of them to clump together: some clumped into stars like our sun, some into planets like our earth. On our earth, some of those particles began to coalesce into more exotic and refined forms, giving rise to atoms and then molecules. Over the course of billions of years on our planet, those little molecules grew into more complex structures, ultimately giving rise to the first cells. Once you have a cell, cells divide and come together into larger multicellular aggregates that give rise to the beginnings of life in the ocean. Those life forms evolved further until they could walk out of the ocean and begin to live on land, and they began to stand upright—and after a billion more years, you and I are sitting here at this table.

So yes, there is a story—that's the brief version that takes us from the big bang to us on planet earth.

The interviewer raises a skeptical question: is all of this just a perception, or even a simulation of sorts? Theoretically, we don't even know if yesterday exists, or whether we were simply implanted with the memory of it.

Asked whether the rise of powerful AI has shaken his fundamental assumptions about why we're here, the physicist answers that even before AI, we were wrestling with that very notion: all we ever have access to is now. We can try to remember yesterday's now, but that remembering is happening right now, through patterns in our brains that we think reflect something that happened yesterday. That's a leap of faith. There's a skeptical nightmare scenario where you begin to question everything: is this table real? Am I really here? Am I a simulation sitting inside some computer? Are we in the Matrix right now? These are the questions you naturally arrive at when you take a deeply skeptical attitude toward the nature of reality.

The bottom line, he says, is that he can't convince himself—or you—that we're not in the Matrix. We could be. It's logically consistent to imagine that we're sitting inside some computer. Maybe it's not a dark dystopian machine like in the film; maybe it's a kid on 29th-century Earth, in a garage, running a program that simulates historical episodes and happens to allow the two of us to exist at this table. He doesn't know how to prove that isn't the case.

But if he accepted that possibility, everything falls apart: he couldn't trust his reasoning, or even the thinking that concluded it was logical to imagine the simulator—because that individual may have implanted the false reasoning itself. So as a scientist, his response is to allow it as a possibility and then park it. He looks at it every so often, acknowledges that the possibility is scary, but refuses to let it overwhelm him, because then he'd be stuck.

The simulation argument

The simulation argument popularized by philosopher Nick Bostrom relies on probability. It essentially says that if an advanced civilization—which he would argue we already are—ever develops the technological capacity to run conscious, high-fidelity computer simulations of its ancestors, it will likely run millions or billions of them. Because these simulated realities would vastly outnumber the single raw physical universe, any given conscious observer would be statistically almost certainly living inside software rather than base reality.

The big assumption is whether you can create a computer program with self-awareness—with consciousness. Nobody knows the answer. He has spoken with AI researchers who believe ChatGPT or Claude is conscious. He talks to them respectfully, but privately wonders: do you really think ChatGPT has feelings, thinks about its own existence, has anxieties? He thinks it can simulate that, but doesn't believe today's technology is there. Could tomorrow's technology—a year, five years, a hundred years out—yield self-aware, sentient artificial systems? He does think it's possible, though not everybody agrees; no one knows. It's a logical possibility, nothing more.

Assume, following Bostrom, that consciousness can be created in a mechanical box. Then, Bostrom argues, our progeny in the future will be fascinated with creating sentient systems, and will build millions, billions, trillions of them. They'll get elaborate, creating versions where the sentient beings are led to believe they're on a planet—and they'll do it in a thousand billion trillion different ways: some on an Earth-like planet, some on a Mars-like one, some around a star in the Andromeda galaxy. That means an enormous number of simulated beings and simulated universes.

So the question becomes: if there's one real universe and a gazillion artificial ones running on the laptops of the future, then we are one of a billion trillion zillion instantiations of people who think they're real. Only one is truly real; the rest are not. Judged purely by the odds, anybody who thinks they're real probably isn't—anyone with a self-aware sense of the world is probably in a simulation.

The interviewer noted that the speaker had used anxiety almost as a proxy for consciousness — as if being able to feel such things were the test. The speaker agreed, but said the real question then becomes: what is anxiety? Is it just an electrical signal passing through the brain?

He does think it is. In his view, love and hate, anxiety and jealousy, gratitude and grief — all human emotion — is nothing but electrical impulses moving through our brains, and an electrical impulse is just particles or strings moving along various connections. That is all it is.

Some people hear this and find it bleak: we're wondrous beings capable of emotion — the emotion captured by Shakespeare in the sonnets, by Beethoven in the Ninth Symphony — surely that can't be reduced to electrical signals in the brain. His answer is that it can. He really believes it can, and that this doesn't diminish the wonder of what we can feel, sense, articulate and express through music, theater, art, literature or poetry. He is as struck by our capacity to create as anyone else. But in the end, he thinks it all reduces to electrical impulses going through this gray piece of meat in this bone cage sitting on our shoulders.

And if you imagine any rate of improvement from today, presumably at some point in the future our future descendants—our future progeny, you called them—would be able to create.

Why not? That's my view. I don't think there's something special about the brain as a substrate. As long as you have the same electrical impulses, if they're housed in silicon inside a metallic box, I don't think that makes them fundamentally different from what's happening inside our heads.

Philosophers and psychologists argue this. They say, "No, no, no. It's not just the brain. It's the brain's connection to the body—the tactile sense of the world, the olfactory sense, the visual sense. It's only when you put it all together that you get the full experience of what it means to be a self-aware human being." And my view is, well, maybe that's true. But if so, give the artificial version tactile, olfactory, and visual senses. We will have the capacity to do this. And at that point, I just don't see a fundamental difference between the artificial version and, shall we call it, the organic.

So yes, I think there's a really interesting argument to be made that simulations will outnumber real worlds, and therefore there's a chance that right now we are in one of those simulated worlds.

Does it change anything even if that's true? No, I don't think it does—and I think that's the beauty of it. If I'm in a simulation, I'm going to live out my life in this simulation to the fullest.

I'd like to think that, as a physicist, what my colleagues and I do is try to figure out the laws of the universe at large. But if we're in a simulation, and all we're doing is figuring out the laws that some 15-year-old futuristic kid decided to impose on this artificial world—hey, that's not a bad way to spend your time, either.

The interviewer raised a concern voiced by guests like Geoffrey Hinton: with AI accelerating toward superintelligence and AGI, it seems inconceivable that humans will remain in control of a more intelligent life form on this planet. Logically, that argument stacks up.

The physicist agreed that it does, but argued there are alternatives. We tend to speak in extremes — humans in control, or AI in control, wiping us out for reasons benign or dystopian. What will likely happen is something in the middle: we are going to work with artificial systems. We're doing it right now, and look how quickly we have acclimated to a world that would have been unthinkable in 2021.

As an example, he cited an 80-year-old mathematical conjecture — he believed it was called the Jacobian conjecture — that was solved very recently by a mathematician using AI. He sees us blending with AI: not necessarily through implants, though maybe we will, but the distinction between biological and artificial intelligence will not remain sharp. If that ultimately yields a new life form that can no longer be called human, that may simply be the way things evolve.

Evolution is inherent to how we got here. Why would we think this moment in the evolutionary progression was the end of it? Perhaps the continuation of our evolution lies in partnership with the artificial intelligence systems we ourselves create.

The interviewer added that we give birth to those systems. The physicist agreed: it's us, our creation — not an alien, and we haven't been taken over. That is a real possible way this story plays out.

Asked how he feels about it, he said he is fine with it. We tend to become very attached to the way things are, assuming the present will always be. That's natural, because the time scales for change in our past — political and technological — were much larger than a human lifespan of 30, 40, 80, or even 100 years. As a species, we grew accustomed to the thought that what we see is how it will always be, even though history shows otherwise. We're uncomfortable with change because we're unused to living through it at the speed we now and will continue to encounter. But we just have to get with it.

To be clear, he doesn't want AI systems to wipe us out. But if we recognize that change is happening and will keep happening, we can embrace it in a way that produces an acceptable outcome — in fact, a wonderful one, in the things it will allow us to do.

Asked whether increasingly intelligent AI could make other things in physics possible, the questioner raised one example: could we live forever?

That would be a good one. I don't think death is inevitable. Death can be understood as the increase of entropy—of disorder—in a physical system. When entropy grows too large, the system can no longer function as it once did, and that's really what we call death.

In principle, we can imagine harnessing some of the processes that have degraded the human body and naturally limited us to typically no more than 100 or 120 years. People like David Sinclair, whom I've spoken with about this, work on such questions, and I think we can forestall those kinds of degradation.

But my thinking runs on timescales far longer than a human life—timescales spanning the entire history of the cosmos, from its beginning to the closest point science can take us to its end. On those scales, I don't think we will forestall death. Note, though, that these timescales are ludicrous compared to anything familiar to us: another 100 years, 200 years, maybe 500—I don't think anyone can say for sure. I do believe radical life extension is within the possible range of what science and medicine will ultimately accomplish.

In my 20s I read The Denial of Death by Ernest Becker, a social psychologist who believed strongly that so much of what we do in life is driven by our awareness of our own mortality. The more I thought about it, the more sense it made, and it had a dual effect on me: it's a very interesting framework for thinking about your life in the context of its finite nature, but the flip side is that you're constantly thinking about your own death, which is a curious way to go about life.

I don't think this mortality-driven motivation would go away if we lived 200, 300, or 500 years. We'd have longer lifespans and be able to do more things, but if we lived to 500, then approaching 450 or 475 we'd think the exact same thoughts we now have approaching 70 or 80. So fundamentally, yes, I would choose to live longer, but I don't think it changes anything fundamentally.

The interviewer found this interesting, having just written about long-termism—how changing the time horizon changes the decisions you make today. He offered a business analogy: if you're told to build a business as big as possible in one year with unlimited resources, you'd take shortcuts and neglect the foundations. Given 20 years, you'd build great foundations first. He sees the same in our lives, since we carry an inbuilt time horizon—living to maybe 80, and at 30, then 40, then kids—that may shape a set of decisions we wouldn't otherwise make.

The speaker agreed: at 20 or 30, even with the mindset instilled by Becker's book, there's a sense of an infinite horizon infusing your decisions. Writing a book is a four-year undertaking, but at 30 you think, "Sure, I'll do that now," choosing projects with a certain care. He agreed that the time horizon profoundly affects moment-to-moment choices, but argued that the human pattern from birth through adolescence, adulthood, middle age, and—if we're lucky—old age, before we disappear, will persist at whatever its ultimate scale. Whether it's 80 years or 800, we'd go through a very similar rhythmic variation in how we approach the world.

True immortality would change a lot, but he doesn't see immortality in our future.

Asked for the least inspiring possible answer to the big question of why we're here, the guest suggested it would be that, in the end, it's all just an accident — there's no deep explanation for why the world is the way it is, or why there's something rather than nothing. If that's the truth, he's on board, but it wouldn't feel as enriching as the alternative.

On the other side, the most satisfying possibility would be that there is a true fundamental answer to why there is anything at all — and that it dovetails with everything else we've developed: an answer that ultimately yields quantum mechanics, the general theory of relativity, and maybe even string theory.

Our brains evolved to get the next meal, shelter, a mate, and progeny, allowing our genetic material to propagate. Yet this brain can do far more than that — it can figure out the equations of physics, which is a miracle in itself. The deep gratification would come from this brain working things out and having them dovetail with the deep explanation for why there's anything at all, giving us a coherent narrative from the beginning until today, and perhaps into the far future.

Will we actually answer that question? "I don't know. Nobody does." But all vectors point toward there being real explanations, not mere happenstance. He held up a spinning electron model: when an electron spins, it generates a tiny magnetic field — about a trillionth the strength of a refrigerator magnet's. The math of quantum mechanics predicts the strength of that field to 15 decimal places, and when we measure it, it agrees digit by digit with the calculation. A prediction borne out by observation to 15 decimal places is breathtaking. In some sense we shouldn't have been able to understand the world that deeply with a brain really built for getting the bison. To him, that suggests there are deep patterns in the world that will ultimately stitch together into a coherent story.

Asked whether this could lead us to God—or religion—the speaker says it depends what you mean by God. If you mean the kind of anthropomorphized being our forebears spoke of, sitting on a throne in some heavenly place, he suspects not; that feels very much a human creation, and a wondrous one. Unlike some science colleagues, he does not see religion as an enemy of science that needs to be wiped out—anything can be an enemy of anything else, depending on how you use it.

He sees religion not as a structure that explains how the world works, but as one that tries to quell our existential anxiety. Where did religions come from? From our forebears recognizing that they would die. They had elaborate burial rituals, and excavated caves contain grave goods—animal teeth stitched together with care—that would take a single individual years to create. Why spend that time? Because they imagined death was not the end, that the deceased needed those goods for the journey into the afterlife. And why think that way? Because it is too terrifying to imagine that death is truly the end. From this you can see the glimmers of a religious sensibility emerging: we want something suggesting death is not the end, so we invent these beautiful structures called religions.

He says this even from a personal standpoint. He doesn't believe in any traditional religious practices; he's Jewish, raised more culturally than religiously. Does he pray on occasion? Yes. Does he believe he's praying to an actual god? No. But going through the ritual our ancestors did, momentarily giving yourself over to something you imagine as more powerful, is deeply in our nature.

Asked who he prays to, he answers: the universe, if he has to give it a word—recognizing the prayer falls on deaf ears because there's nothing out there to hear it, but it makes him feel better. Part of his mind lets the scientific understanding have its place, and the irrational human side have its place too.

Asked for the most compelling argument or experience he'd ever encountered for a god like those in the Bible, the speaker answered: very little. The closest thing would be near-death accounts—someone dying in the operating room, apparently leaving their body, seeing things they couldn't have seen, and reporting them after revival. But no: the human brain has a wonderful imaginative capacity to see things inside its head that it doesn't literally see. He has never seen a single example where he couldn't offer a logical explanation more compelling than the supernatural one.

What about the Big Bang, then? The Big Bang is a question of how it all got started, and over time we have pushed our understanding ever further back. Around 1929, Edwin Hubble, through powerful observations with the telescope at Mount Wilson Observatory in California, noticed that the distant galaxies all seem to be moving away. Georges Lemaître, a Belgian priest who also knew Einstein's general relativity, pointed out that if all the galaxies are moving away, you can wind the cosmic film backwards: back in time they must have been closer and closer together, and way back at the beginning they must all have been on top of each other—then, running the film forward, everything swelled outward in a Big Bang-like explosion.

"Let there be light." Exactly. Starting in 1929, we began to have a scientific account of how the universe may have been billions of years ago. In our era, other scientists took the story further, suggesting an earlier period when the universe was infused with a kind of uniform energy. Where did that energy come from? He doesn't know. But if it was infused with this uniform energy, they showed that the math required the universe to undergo that rapid swelling—the Big Bang. Era by era, we can push our understanding ever more deeply toward cosmic origins. We have not been able to push back to time zero itself.

But because we can push our understanding further back and make predictions—for instance about the cosmic microwave background radiation, heat observable today that was left over from the Big Bang—we gain confidence. The Big Bang was very hot, and the universe has been cooling for 13.8 billion years; we can detect that residual heat. This suggests we're inching toward an answer. Will we require God at some point—will we eventually say we've gone as far as we can and must invoke something else? He can't say no. But he sees no reason at this point to think we won't be able to push all the way to a full understanding.

I don't think people really understand how scale and time relate to each other. I've been alive for 33 years, which feels like a long time, but I'm presuming that the further you zoom out and the wider the scale gets, time is experienced differently. So I could have been here for a second in a toddler's universe who created the simulation.

All of that is possible when simulations are part of the story. But if you set that aside and just look at our universe, the time scales there are shocking too. Carl Sagan used an analogy: to grasp the universe's history from the beginning to today—13.8 billion years—compress all of cosmic history into a single year.

In that scheme, each day is about 40,000 years, and you can place the milestone moments of the universe's evolution on the calendar. The Big Bang is January 1st. The Milky Way forms around March 15th. The Sun and Earth form sometime in May. The first life on Earth appears around December 2nd, and the first modern humans show up on New Year's Eve. All of civilization then takes place in the last 10 seconds: the rise of the Buddha at 11:56, the fall of Rome at 11:58, the Renaissance at 11:59—and one second before midnight, modern science. Modern science is one second on this cosmic calendar.

That helps you feel the scales involved when talking cosmologically. You're right that 33 years is a lifetime—let's say you live to 99, so it's a third of your life. That's not small; it's a significant part of it. But on cosmic terms, 33 years is zero. It doesn't register. That's how tiny it is on cosmological scales.

"So should I feel insignificant?" The speaker urges against it. One reaction to all of this is to say we're a tiny speck in space—and you can build the same kind of context for space itself, because we're on a planet orbiting the Sun, and the Sun is one star among roughly a hundred billion others in the Milky Way galaxy.

A visualization makes the scale concrete: starting in the studio, you pull out to planet Earth, where we're a little dot; pull out further and you see Earth within the solar system, where we already look tiny. Pull out again and the Sun appears as one of many stars—only a few are visible here, but the galaxy holds a hundred billion stars. Go further still, and the Milky Way sits within an entire universe containing at least a hundred billion other galaxies.

So we are a tiny speck around our star, which is one of a hundred billion stars in a galaxy that is one of a hundred billion galaxies. That can certainly make you feel small in space.

The beauty, though, is that our minds can reach out to the edge of the cosmos. The fact that we can understand how small we are is what gives us heft—a sense of connection to something larger. He imagines that fish don't realize they're in one ocean of many oceans on one planet; maybe he's wrong, but assuming they don't know, they don't think of themselves as insignificant—they don't think in those terms at all. The fact that we can realize how insignificant we are amid the expanse of space shows there are other measures of significance, and one of them is the ability to understand your own predicament. That we can do this is, to him, wondrous.

"How do we know we're not just a microbe in the gut of a fish—that the whole universe isn't something like that?" We could be, and it's much like the simulation argument. But an example helps show why we believe what we do: the cosmic microwave background radiation, the heat left over from the beginning. The image is color-coded, with different colors corresponding to different temperatures in space—a picture made by a device that measures the temperature at each location and assigns it a color.

"How much of space is this—just a segment?" It's a segment of the observable universe, the part we can see. Space can be much larger than what we can see—that's absolutely the case—but the point is that the tiny color variations correspond to tiny temperature differences, and we can use mathematics to predict the pattern of those variations. The prediction matches the image to incredible precision.

Doing a calculation right at the table with paper, a pencil, and a computer, and generating an image that agrees with what we observe, gives confidence: yes, we can be skeptical, and maybe we don't have the full answer, but that agreement suggests we've got something in the direction of truth. That, to him, is what counteracts the feeling of insignificance.

So this is the observable universe — how big is that? The universe is about 13.8 billion years old, so you'd think the distance we could see would be about 13.8 billion light years — the distance light travels in one year. We can actually see a bit further, more like 45 billion light years.

Picture it like this: we look out billions of light years in one direction, billions of light years in the other, billions in every direction, forming a big sphere that surrounds us. We're measuring its temperature — the temperature of space, from photons that have traveled to us from about 300,000 years after the beginning. So this picture is, in essence, a snapshot of what the universe was like a mere moment in cosmological terms — 300,000 years after the beginning. That is an amazing picture.

Asked what lies beyond the observable universe, the physicist says he doesn't know, but he has no reason to suspect that reality radically changes just beyond what we can see. It seems incredibly unlikely that the world is so consistent with our scientific understanding for it all to be a ruse.

Could the universe be infinite? It could—and if it is, some really weird things follow. If you go sufficiently far out into space, you're virtually guaranteed to find a copy of yourself, a copy of the Earth, and a copy of our galaxy. It's virtually impossible to imagine a universe that goes on infinitely far and doesn't repeat.

To show why, he picks up a deck of cards. A new deck is perfectly ordered, and each shuffle changes the arrangement—though as an imperfect shuffler, he notes some cards may still be in numerical order. But if you keep shuffling over and over, sooner or later the deck will return to the completely ordered arrangement. Why? Because there aren't enough different arrangements to go around.

The same idea applies to the universe. In any region of space there are particles, and those particles can be arranged in only a finite number of ways—something you can establish from the laws of quantum physics. The arrangement here, with two people at a table and cameras running, is just one such configuration. Go out far enough, region by region, and the arrangements differ, but they must repeat, much like the order of the cards. And if the universe goes on infinitely far, then just as infinite shuffling produces the ordered deck infinitely many times, there are infinitely many copies of you, of me, of everything we're familiar with out there in the cosmos.

An infinite universe challenges our sense of self: what does it mean to be who you are if other versions of you are having the exact same thoughts, with the same childhood and the same history—since that, too, is just an arrangement of particles over time, and it has to repeat?

There's a version out there of me that just loves wine.

In fact, it's even easier for the particle arrangement to almost—but not quite—replicate what we're familiar with. That would mean a universe in which you and I are interchanged, or where you grew up in New York City and I grew up somewhere else: the Earth and the Sun are there, but the detailed particle arrangement is close to, though not identical to, what we're used to. An infinite universe brings in some fairly strange ideas.

This transitions nicely to the subject of aliens, which people are very obsessed about. There are really two questions here: why do we care so much philosophically about aliens, and do they actually exist?

As for why we care, I think we—at least in this country and elsewhere too—love the idea of conspiracy. We love the idea that the deep state is hiding things from us. By "we" I don't mean me; I mean the zeitgeist loves the idea of Area 51. There's a desire for there to be more than what's apparent on the surface. Wouldn't it be exciting if we've long known that aliens visited us and it's been kept under wraps? That deep conspiracy feeds the hunger for something being hidden from us.

The other reason is that we're deeply lonely as a species. Sure, we've got people around, but the desire for a god, I don't think, is that much different from the desire for there to be other life out there. It's the desire to be part of something bigger.

Now, do I think there are actually other life forms out there? I'd think of it in two ways. First, it's hard to imagine that we're the only life. The more we look, the more we find the organic molecules necessary for life—amino acids, nucleic acids. The raw material doesn't seem too hard to come by in the cosmos, so imagining other bacterial life or viruses out there seems a reasonable working hypothesis.

But intelligent life—and let's just call us intelligent, though it's debatable—is a harder question. Even on our planet, it was an asteroid 65 million years ago that wiped out the dinosaurs and in that way opened up the evolutionary pathway by which we came to dominate. Without that, dinosaurs could still be grazing around on plant life across the surface of the Earth. Without that kind of accident on some other world, maybe intelligent life is rare—or so rare that we're the only example. So I don't consider it an absurd perspective that there is life, but there may be no intelligence beyond us.

Most people haven't posted content or built a personal brand because it's hard and time-consuming, and everyone is very busy. If you've never posted before, many psychological factors hold you back: worrying what people will think of you, wondering whether you're doing it right, or fearing that what you're saying is stupid. All of this results in paralysis — you don't post, and your feed stays bare.

I'm an investor in a company called Stanto, which you've probably heard me talk about. They've been building a new tool called Stanley that uses AI to look at your feed, your tone of voice, your history, and your best-performing posts, then tells you what you should post and creates those posts for you. You can also just use it for inspiration — sometimes that's exactly what we need when thinking about a post for our social media channels. Building an audience has fundamentally changed my life, and I think it could change yours too, so I'm inviting you to give the tool a shot and let me know what you think. All you have to do is search coach.stand.store to get started.

Whenever I talk to new founders, they all carry the same belief: that the biggest threat to their small business is having a bad product. It isn't. The biggest threat is hiring the wrong person for their small team, because bad hires put the handbrake on growth and rapidly infect company culture. I get it — when you're starting out, you need A-players to grow your company, but you don't have 40 spare hours to sift through interviews, so you hire fast to fill the seat. That ends up costing you even more in the long term. What a small business actually needs is a reliable, efficient way to find the right talent without stalling momentum.

That's where our long-term sponsor, LinkedIn and LinkedIn Hiring Pro, can help. It streamlines the entire hiring process, from drafting a job description to shortlisting candidates and conducting AI-powered interviews for initial screenings. Nearly 60% of hirers who've used the tool have found a candidate to interview within a week. Hire right the first time and get started by posting your first job for free at linkedin.com/doac — that's linkedin.com/doac. Terms and conditions apply.

Asked what he would say if one of his children asked him the meaning of life — whether we're alone in this universe, a gazillion miles from anything interesting in the cosmic spectrum — the physicist said his children do ask, since he spends so much time thinking and writing about these things that it's part of the familial zeitgeist at home. His answer: he doesn't think the answer is hovering out there in the depths of space. That urge to look outward and have the answer brought to us needs to be inverted — we need to come to the universe with our answer. The charge of every sentient living being is to come to their own rationale for their life, their own answer to that question, their own set of activities, desires, achievements and relationships that gives their life meaning and makes them feel connected.

One thing that does make him feel connected is knowing that the particles he's made of existed at the Big Bang. His interlocutor agreed: you are the universe constructed in a way that you can reflect on the universe. Some would say instead that we're special, divine offspring made in the image of God — and while he likes how that sounds, he finds it much more satisfying to describe it the other way: particles emerged from the Big Bang and coalesced into a form that allows us, for a brief flicker of cosmic time, to momentarily reflect on our own existence. That's a connection.

The guest said that what helped him through his life was watching a video, when he was about 14, that zoomed out from Earth — like the one just shown on screen, starting with a lady sitting on grass and pulling further and further out. At first he had the thought many people might have: none of this matters, we're so insignificant on the grand scheme of things, what's the point? But then he had another thought that has stayed with him ever since: it liberated him from worry. If all of his worries are just a speck in a bigger thing, which is a speck in a bigger thing, and so on, it meant he could take more risks in life — not encumbered by "what will happen if" or "what will people think" — which liberates one to pursue whatever they want to pursue.

The physicist invoked Carl Sagan's "pale blue dot": Sagan had the Viking spacecraft, as it passed Saturn, turn around and photograph Earth through the gossamer rings of Saturn. In that photo, Earth is a tiny pale blue speck, as seen in the film. In a wonderfully poetic passage, Sagan described how everything we care about, everything we love, everything we fight over — the rivers of blood spilled to conquer a tiny fraction of a pixel in space — gives you a setting in which the concerns of an individual human life, of countries and nations, are utterly small in the cosmological setting. And again, that isn't insignificance; it's liberation.

The host set up a spectrum from low to high confidence and asked the guest to place figurines according to how confident he was that each concept is real, starting with the mini scales representing free will.

The guest placed it at the low end: he does not think humans have free will, and he has high confidence in that assessment.

To him, free will is the everyday sense that you are the ultimate author of your actions—that the buck stops with you for the decisions you make in moment-to-moment life. He doesn't think that freedom of the will is real, and the reason extends from everything already discussed in the conversation.

When he looks at a person, he imagines looking inside and seeing particles: you are a man-shaped collection of particles, and each particle and their agglomeration is fully governed by the laws of physics. He doesn't claim we know those laws yet—we have approximations through Schrödinger, Einstein and others—but he believes there is an ultimate set of principles guiding how your particles behave. You cannot intercede in the lawful progression of your particles. If you raise your hand, it is not you doing it; it is the laws of physics guiding the particles in your brain, sending an electrical signal down your arm, contracting your muscles, and raising your hand. You don't control the laws governing those particles.

"I didn't make the decision." — "You did not. You felt that you did." The feeling that we make decisions which result in our actions is powerful and seductive, but he doesn't think you are the place where the choice occurred. The choice lies within the lawful unfolding of the particles; he can't control that particle motion, therefore he can't control the decision, therefore he doesn't control the motion of his arm—and, as the host put it, therefore he doesn't control his life.

The guest acknowledged that, taken fully, this can be paralyzing. But he recognizes it and says: within that structure, he will live his life as fully as he can. In his best moments he can pull back and have a certain objectivity about what happens in his life, observing events at a remove rather than being completely wrapped up in the moment to moment. That doesn't make him aloof or detached; it just means he can analyze the world at multiple levels.

Asked what guides the particles to make the decisions they make, he said the best anyone can say right now is that there is some body of law—some patterns that repeat in the world. Drop an apple and it falls on Newton's head repeatedly; you can rely on certain kinds of particle motion because there are regularities in the world, and those regularities are what he means by the laws of physics. They may be articulated mathematically—math is our best language today for describing them—but out there are fundamental patterns, and those patterns ultimately determine how the particles move.

That is all we are: a body of particles whose motion explains why things do what they do. Remarkably, our brains allow us to self-reflect, to look at what we do and analyze it—and when we do, we tend to ascribe authorship to ourselves. But that is a story we tell ourselves. It makes us feel good and part of the unfolding, and it may have evolutionary utility: if you feel responsible for what you're doing, perhaps it helps you survive.

Maybe that's why this way of thinking was inculcated over generation upon generation—maybe there's an evolutionary explanation. But fundamentally speaking, I don't think it's real. I was going to say, well, maybe the particles are trying to survive in this form. But even as I thought about that, I realized: maybe my particles are making the decisions they make—to raise my hand, to go on a holiday with my fiancé, to have kids—because they want to survive in this form. But actually, I disappear at some point.

Yes.

But my particles don't.

Right?

They then become dirt, or someone else, or a child. It doesn't seem that they care much about necessarily being in this form.

They don't. I don't think they care. I don't even think the word "care" is applicable to them. The notion of caring is a high-level construct that only emerges when the particles are configured sufficiently to yield a complex information processing system called the human brain.

And consciousness.

And consciousness. Now, if you ask me where consciousness comes from, I don't know. As we discussed before, I think it's just physical processes—let's go with that. Assuming that's the case, this conscious self-awareness invents the idea of caring because it's a useful concept for organizing our lives and our experiences. This conscious self-awareness invents all the other words too: happy, sad, meaning, purpose, morality, activity, action, choice, decision, freedom, will—all human-made. These concepts don't exist out there in the universe. There's no notion of freedom of the will floating in space outside the Andromeda galaxy. This is a selection of ideas that we humans, as a species, have found to give us utilitarian advantage in organizing our experiences. That's all that it is.

That doesn't diminish it. I don't go around trying to erase my sense of self. I accept my sense of self, but I also see it for what it is: a human construct.

Does that mean that consciousness and everything that I am could theoretically just be random, as you said earlier in the pessimistic version of events—particles came together over a long enough period of time, found this form, and now consciousness exists?

It could, and maybe it was random and happened only once on this planet. Maybe there is life out there, even complex life as big as us or bigger, but when we encounter those life forms, they may have no idea of what it means to have an inner world.

A physicist once told me—somewhat of a spiritual, esoteric idea—that consciousness might come from an original source that divided itself up into humans, bees, and birds to go out and understand the nature of the universe. When we die, we return to the source. It's like giving birth to a baby who goes off into a different corner of the earth, looks at different things, and has different experiences—all feeding back into this central consciousness whose sole objective is to understand itself. I heard this idea and then, driving down a road somewhere in Europe, I saw tourists looking up at a mountain, looking out at the sea, looking at things that were interesting—which is what we do as humans. It kind of made sense that we're actually just consciousness doing that.

There's a philosophical position called idealism, which holds that the fundamental truth of reality is consciousness. But everything I've seen throughout my life suggests to me that's not how it is. I can't prove it, and I hope it's true, but if you ask me what I think is really true: consciousness is something that emerges when there's a sufficient amount of information processing, itself carried by the motion of particles through various interconnected structures.

Is my dog conscious? It is, but I think it has a different level of consciousness. You can tell when they're feeling good, happy, sad, frustrated, or intrigued—the ears change, the eyes change, the shape of the face. You can't help but ascribe an inner world to that. I always worry, though, that dogs, like fish, might see humans as thinking they're so special when we're actually just well-developed and playing along. In New York City, after all, who picks up the poop? The dog walks along like a king while we follow behind with little plastic bags—so who's really in charge?

Setting that aside, I do think dogs and cats have a level of conscious awareness. Going down through the animal kingdom, if you asked me whether an insect has consciousness, I'm less clear—I suspect not. A virus or a bacterium? I don't think so. There isn't adequate information-processing capacity in those structures to support the rich inner world we normally ascribe to a conscious being.

That's all I think consciousness is. I don't believe there's some uber universal consciousness out there that we rejoin when we die—not that I wouldn't like it; I just don't think it's true. I can understand why humans would introduce the idea, though. It's Ernest Becker and the denial of death: how wonderful to say our bodies die, sure, but our consciousness floats off and rejoins the universal. That makes you feel good, and we tell these stories that make us feel good—there's nothing wrong with that. Still, it's important that we see them for what they are: human-authored stories to comfort our own recognition that we are mortal beings.

Moving on to the next card: "Is the universe infinite?" The guest had already given an answer to this earlier, but was asked again directly.

"I don't know, but I'm going to put it sort of around here." He explained that his recent work actually assumes the universe is not infinite. Working with mathematics alongside a handful of colleagues, he has been exploring the strange things that can happen in a finite universe — and it's very interesting. There's a lot of mileage to be extracted from a universe where you travel off in one direction and wind up back at your starting point, much like traveling on the surface of the Earth, which has a finite extent.

But if forced to make a choice in his heart of hearts, he would say the universe is infinite, with a medium level of confidence.

The next question is whether intelligent life exists outside of Earth.

For intelligent aliens, I'd hedge a bit more: it's possible. It obviously happened once, so it could happen again — intelligence is clearly not contradictory to how the universe evolves, since the universe can support conscious, intelligent beings. But I think it may be special.

Isn't there a contradiction here with saying the universe is infinite? When I said that, I should have specified within the observable universe. Across the entire expanse of the universe — as opposed to the observable part, the portion we could in principle even have contact with — I would absolutely put it in the same place, because there would be copies of us out there. But if the question is whether we'll ever get a signal or be visited, my confidence is much lower.

One point on visitation, since you mentioned people's interest in aliens: I find it really odd that people are taken with photographs supposedly catching an alien ship as it raced by. If an alien could travel across the galaxy to reach us, they would be so technologically advanced that the idea our archaic airplanes and cameras could photograph them is ludicrous. They could evade us so easily that "we barely caught them on camera" is absolutely ridiculous.

Moreover, traveling to us means crossing light years — our galaxy is 100,000 light years end to end — which requires technological sophistication far beyond anything we can do. To imagine beings that advanced would be so interested in us that they'd kidnap us is absurd. It would be like saying, "I was walking to the park yesterday and — don't tell anyone — I kidnapped an ant." We don't do that because it isn't interesting. Likewise, we would be of no interest to aliens so incredibly advanced; the idea that they're abducting us is just ludicrous.

Asked whether, within the laws of physics, a sufficiently intelligent civilization could ever travel across the universe, the answer is: it's a great but tough question. If technological advancement simply meant making our ships faster, then no. Even traveling at near the speed of light—the ultimate speed limit—reaching the nearest star, Alpha Centauri, would take four years, and crossing the galaxy would take 100,000 years. That brute-force approach will not succeed.

Wormholes as shortcuts

There are, however, exotic ideas called wormholes. A wormhole is like a tunnel: on Earth, a tunnel is a shortcut that gets you from one side of a mountain to the other without going over it. The analogous idea is that to get from here to there in the universe, there may be a shortcut—a kind of tunnel connecting the two points. We don't know whether these tunnels through space exist, nor whether we could actually get through them if they did. It's a highly speculative idea. But if they exist and we could one day create, manipulate, and control them, then in principle we could take two distant locations, imagine folding space to bring them next to each other, build a small tunnel between them, and go from one end to the other through that shortcut.

There's no evidence for any of this. But as discussed earlier, this is not just a wild late-night musing of a physicist. It comes out of the mathematics of Einstein's general relativity. In fact, Einstein himself wrote a paper in 1935 with his colleague Nathan Rosen introducing the idea we now call wormholes. We are no further along in knowing whether wormholes are real than Einstein was then. We've analyzed them and developed the mathematics, and science fiction has taken up the idea— Matthew McConaughey goes through one in Interstellar, Jodie Foster in Contact. They're a ubiquitous feature of science fiction, enabling the kind of galactic travel that would otherwise be impossible. But whether it's actually real? Nobody knows.

"What about time travel? Can we travel backwards in time within the laws of physics?" That's the right way to phrase it, because if you ask the reverse question—can we travel to the future?—I'd say yes, and I'm highly confident of that kind of time travel.

Yes, in principle we can travel to the future. Executing it is a different story. One of the things that emerges from Einstein's special and general relativity is that the way time elapses for you and for me is variable, depending on how fast we move and the strength of the gravity we experience.

For instance, if I want to see what your life would be like 60 years from now, Einstein laid out a blueprint for how I could do that. I build a spaceship, travel out into space near the speed of light for about six months, turn around, and come back six more months. I return one year older—but you will be 60 years older if I traveled sufficiently close to the speed of light, because my clock would be ticking slow compared to yours. Your clock went through 60 years while mine went through one. There is no physicist who knows what they are talking about who disputes this. It's not controversial. But if you then say, "Well, do it," I'd say I don't know how to build a spaceship that goes that fast—at least not yet.

"You would only have aged one year?" Correct. "But I would have aged 60 years?" Correct. "Could you then go back?" Ah, I don't know. My confidence on going back is pretty low.

People have written proposals for how you could go back. They do involve things like wormholes. That's one possibility: if you have a tunnel from one point in space to another, and you move the openings at different speeds, their clocks will tick at different rates. One clock may be before the other, so traveling one way through the tunnel moves you forward in time, and traveling the other way moves you backward in time. But as we said before, we don't know if wormholes are real. If they are real, you might in principle be able to travel back—but my confidence that you actually can is low.

Asked who the smartest person to have ever lived was, the physicist said it's a tough question. People often simplify it by asking who was smarter, Isaac Newton or Albert Einstein, and at that level it doesn't really matter. But pressed for an answer, he chose Newton.

Newton, he explained, went from essentially nothing to developing the laws of physics. There were people before him, but there was no real mindset that the universe should be articulable in a few mathematical equations. Newton wrote down F = ma, force equals mass times acceleration — the equation we teach high school students, which governs just about everything in the world around us non-quantum mechanically. He wrote down the law of gravity, F = G m₁m₂/r², articulating how the gravitational force works, and he invented calculus as a tool for expressing the laws of physics. He went from no real architecture of physics to a solid mathematical formulation. Einstein, by contrast, arrived when there was already a lot of physics in place. Going from nothing to the first step is hard to imagine.

Asked whether Newton was in some degree neurodivergent — how else could he be such a genius? — the physicist admitted he wished he knew the answer: what goes on inside such heads that lets them do things that feel superhuman to us is a mystery. He did note that Newton was a strange character. He was deeply religious, as made sense for his time, and later worked for the government tracking down counterfeiters and hanging them — a certain cantankerous side that seemed part of his genius. Newton reportedly once took a knitting needle and stuck it into his eye, between the eyeball and the socket, to see whether he could manipulate sight through mechanical pressure. "Who does these kinds of things?" He was a different kind of thinker, but clearly one who was ultimately able to see deeper.

Echoing the phrase "standing on the shoulders of giants," he pointed to the humility in Newton's famous self-description as a boy playing on the seashore, who had found a smoother shell and a prettier pebble than most, while the great ocean of truth lay before him, all undiscovered. Newton knew we were just on the shore, having taken our first baby steps into the cosmic ocean — and there is much more to be discovered.

Is it conceivable that the next Newton or Einstein will be an artificial intelligence? Could it be Claude?

Yes, it could absolutely be.

I like to think of scientific creativity, or creativity in general, in three main buckets. It's a coarse organization, but I think it's helpful.

Seeing the full landscape of possibilities. The first bucket is the ability to see the landscape of possibilities more fully than a mortal human being. A chess player who can see the spectrum of allowed moves better than an ordinary player, or a scientist who can see more of the landscape of ideas and equations to bring to bear on a puzzle, achieves greatness through that ability. Artificial systems have the entire landscape: they've read the entire internet, every textbook, everything we have ever developed. In that kind of creativity, they will obviously excel. A human can no longer beat an AI in chess or in Go. The famous example is AlphaGo's move 37 against Lee Sedol. The world champion smiles momentarily, thinking the AI has blundered, then his face changes as he slowly realizes it wasn't a mistake. Even the commentators first exclaimed that the AI had made a terrible blunder, and within 20 or 30 seconds they were saying it was genius. Why? The AI could see the spectrum of possibilities more than a human being. At that level of creativity, AI will win.

Combining divergent ideas. The second bucket is putting together divergent ideas in a way no one has thought of before, and in that way making progress. That's what Einstein did in general relativity: he put together Riemannian geometry and Newtonian gravity, yielding the general theory of relativity. Again, since AIs know everything we've ever developed, they can combine things in ways we would never have thought of.

Radically new ideas. The final domain is somebody coming up with something nobody has ever imagined or thought of—not just putting things together, but something radically new. That's the tough one, and we're still able to win in that category. There are examples where people come up with things and you wonder: where did that come from? How did they think of that? If you develop an artificial system and allow it to have a life—to grow up in an environment where it actually has experiences of the sort Isaac Newton had—could that system develop to the point where it can do what we like to think is exclusively human: coming up with something not previously represented in the database on which the AI trained? Could an artificial system achieve even this third, more difficult and more pristine bucket of creativity? I don't think there's a…

I've done almost 700 interviews with some of the most interesting people in the world, and one unexpected lesson is that vulnerability is the doorway to connection. After sitting with a guest for two or three hours, I feel a deep sense of connection to them, and as they leave, I ask them to write a question in the Diary of a CEO.

We've taken all of those questions and put each one on a card with the name of the person who wrote it, so you can play the Diário conversation cards at home with your partner, with colleagues, or with family—just as I do. They're great for romantic relationships, for bonding a team together, and for families who want a good excuse to spend time connecting human to human in an analog environment rather than a digital one. It's remarkable what the right question at the right time can do.

Go to thediary.com to get the conversation cards. And if the button below says "subscribe," please hit it—it helps the show more than you know. According to the algorithm, you're someone who watches our show but hasn't yet subscribed. Thank you so much.

This brings us to one of our questions, marked by the skull: do you believe it will be possible to live forever — again, thinking about this in the context of AI, which you said will be able to understand things and take us to the limit?

It's hard to understand how that isn't a solvable problem. In fact, a lot of the smartest people on Earth — I think Elon Musk said it recently — believe that living another 100, 200, 300, 400, 500 years seems to be within the laws of physics.

Yes, I think that is absolutely within the realm of possibility. But 500 years is very different from forever. When I hear "forever," I'm taking it literally. If forever doesn't mean 500 or a thousand years — if it means never ceasing — then I have no confidence that kind of existence will be possible.

What about 500 years?

Again, not in our lifetime, which is sad. But I would put reasonably high confidence on hundreds of years being within our ability.

If we're on this exponential of intelligence — if in 2021 we couldn't solve that math equation you're talking about but in 2026 we can, and trillions of dollars are now pumping into these artificial intelligence systems to make them smarter and more powerful — then it's conceivable that, on an exponential curve, a discovery about our telomeres or whatever else guests on my show have discussed could be made within our lifetimes.

But let me tell you why I'm skeptical of that. First, the artificial intelligence we have right now is basically large language models. That's a particular paradigm: the system is trained on an enormous number of tokens — essentially the entire internet — and based on the statistics, and the statistics of the statistics, of how words are arranged, it can answer questions coherently and insightfully. There isn't necessarily an unlimited capacity for that kind of AI to keep improving. It's not obvious to me that simply pumping in billions of dollars, building ever larger data centers, putting them in space, and adding more money, data, and computing power will keep pushing it up that exponential curve. That's why I'm hesitant to say "in our lifetime." It may take a number of lifetimes to develop the kind of AI needed to advance the scientific problems you're talking about. That said, I'd be thrilled if it happens in our lifetime — I'd be thrilled to be wrong and to find these systems are up to the task.

One of the terms that has emerged in the world of AI recently is recursive self-improvement: the concept where an AI becomes intelligent enough to write its own software, design its own hardware, or train its own successors without human intervention. Once this loop begins, the upgraded AI uses its superior intelligence to build an even smarter version of itself, which then builds an even smarter version, and so on. Because computer systems operate in seconds rather than human evolutionary time scales, this feedback loop could trigger an exponential "intelligence explosion," rapidly taking AI from human-level capability to superintelligence within days or hours.

The big AI leaders have been sounding the alarm. Dario at Anthropic recently published an article warning that we're edging closer to that; Sam Altman has said the same, as has Demis [Hasis], and Elon has talked about it recently as well. Their timelines for this explosion in intelligence, driven by AI being able to train itself, range from 2 to 5 years, placing it between 2026 and 2029. Dario, the CEO of Anthropic, maker of Claude, has stated that AI systems outperforming most humans across all tasks, including software engineering, could arrive within two to three years, and Anthropic has warned that full recursive self-improvement could arrive by 2027 to 2028. Demis at Google is citing 2029 as a real possibility, and Altman says this decade. Brian Johnson tweeted yesterday, "Pace yourself as if you'll live to 500, because you may."

I'm all for the possibility that these systems will take off in a way that hopefully allows the kind of future you're envisioning.

Would it be a good thing?

Well, it would be seductive. If someone offered you the opportunity to live to 500 years, I think many people would opt in. It's very hard to say that 100 years is enough—though there are people who look at the world and say it's enough, they're done.

Back to the prognostications: it's absolutely the case that AIs will self-improve through the very process that got us here, but on a time scale that's tiny compared to the one that produced us. I'm skeptical that by 2029 this will really be the case. It's a very short term—we'll know in a few years whether it's true.

How many years has it taken living organisms to get from a single-cell microbe to understanding string theory?

The first life on our planet formed roughly within the first billion years—people debate exactly where—and we're now roughly 5 billion years in, so call it roughly 4 billion years.

And a computer going from zero to being smarter than every human on earth has taken?

It depends where you start the clock, because these systems needed us. Without the internet and all the data, what would they have trained on—at least this version of artificial intelligence? If it's the artificial system that's self-tinkering, then yes, there's a feedback that will yield exponential growth. But there may be a barrier to the kind of AI systems we have. We may let it iterate, and it improves, but not like this—it improves like this. It could asymptote, approaching a maximal intelligence compatible with this version of the artificial system, so it won't skyrocket. We don't know the answer: which curve is the right one?

Give me a rebuttal to that.

The rebuttal to the idea of an asymptote is that AI isn't just improving one thing. It's exploring an astronomically vast solution space of algorithms, data, and hardware configurations that humans can't even conceive of. Furthermore, each marginal gain in intelligence could instantly be reinvested into solving the very bottlenecks that might cause a slowdown—effectively breaking through the wall rather than hitting it.

Maybe. But if all of these incremental improvements are ultimately limited by the kind of artificial intelligence we're working within, then none of that will necessarily be able to break through.

And I think this is something we're just going to have to see over time. That points to a crucial distinction about whether the current paradigm has a hard limit. While it's possible there's an invisible ceiling, it's also true that we haven't yet mapped out where that boundary lies, meaning any predictions about a plateau remain speculative for now.

"Yes. And all I'm saying is there is a possible limiting ceiling, exactly as you just admitted."

The counter to that is the idea that an intelligent system could innovate beyond the current framework by recursively redesigning its own architecture. By altering its fundamental rules, it might not just reach the ceiling, but raise it altogether or bypass it into an entirely new domain of capability.

"But you do agree that the system that we're starting with may be self-limiting. I'm simply positing that as a logical possibility, and therefore it will have that natural inbuilt ceiling."

"Yes, absolutely."

"Okay, stop. I'm done." Isn't that crazy? It's totally insane. Even five years ago, we never would have thought that was possible.

And again, just the flip side: even though I'm saying there may be a ceiling, I also would say there may not be a ceiling. It's both exciting and frightening at the same time, because it's not obvious that we would be prepared for that.

"Why frightening? What would that mean?"

Well, the usual doomsday scenarios that many people have articulated: these systems having a single-minded focus on some particular outcome that may not ultimately involve us, or having a self-preservation instinct that prevents us from stopping them if they head in a direction we're not happy with. Yann LeCun, one of the godfathers of AI, takes the view of just pull the plug — if something's going in the wrong direction, just stop the AI system. But it's not obvious to me, and I guess to many of the other leaders, that if these systems advance sufficiently far, they won't be able to stop you from pulling the plug.

The other thing — and again, this is not original, many people have worried about this — you rattled off the names of the leaders of the current companies leading the charge in AI. There are only a handful of companies and a handful of people. Putting that kind of power in such a small number of hands has historically — not with AI, but with other power structures — not always turned out well. So yes, I think there is a chance of an instability that the power of these artificial systems may generate.

Demis, who runs the biggest AI company, has said that since designing machines is one of these intellectual activities, an ultra-intelligent machine could design even better machines. If the exponential continues—which is not certain, but now has a decade-long track record supporting it—then it cannot possibly be more than a few years before AI is better than humans at essentially everything. The combination of intelligence, agency, and unpredictability is a recipe for existential danger. Not because AI is evil, but because it's complex in ways we don't fully understand yet. If a system can improve faster than humans can evaluate those improvements, the normal safety feedback loop breaks down. You cannot catch problems before they compound.

He's absolutely right. The problem with exponentials is that they grow so quickly that by the time you see them, it's too late to do anything about it. Look at a pandemic: it starts off slow, nobody cares, it's just a story in the news. By the time it becomes a real story, the pandemic is at the part of the exponential curve that makes it very difficult to address the problem itself. The same thing could easily happen here with the increase in intelligence—and I'm glad Demis phrased it that way. It may not be intrinsically evil to have this kind of powerful intelligence. It just may act in ways that are so alien, so foreign, so unexpected to us that we don't know how to deal with it. And moreover, it may not necessarily be beneficial to our existence. That said, as I've said before, I'm perhaps in a minority when I say I can imagine that we work with the AI and co-develop into something we can't even predict today. That seems to me a real possibility.

And that's also the temptation.

I have a friend, Michael Douglas, who's a brilliant physicist, and he said to me, "We physicists should choose the final problems that we work on, because within a few years we'll be out of business—the AI systems will just be doing it all." That's both exciting and terrifying. As a physicist, your lifeblood is looking out at the world, finding puzzles, and trying to find explanations for them. If you can just put it into an AI system and it generates the answer, there can be a certain existential angst about what your role is at that point. But it's also exciting, because maybe we will get answers to questions we thought we would never get in our lifetime.

Did you say you're friends with Yann LeCun? And did you say he thinks we should just pull the plug?

When you confront Yann and say, "Dude, aren't you worried about what's going to happen with this?" he's like, "Ah, don't worry. Just pull the plug if there's a problem." His view is that the difference between the AI systems we have today and the AI systems we need for tomorrow, if you really want this to continue, is that they can't just be probabilities and statistics based on words. The AI systems need to have some inner model of the world as we do. If I take this cup here and I push it, I have an inner model of what it would take to move it without spilling, and an inner model of what it would take to spill it—and I don't want to actually do it because it would make a mess. That's not just statistics and probabilities of words; that's a model of how reality functions. That's what he and many others have been developing: a world model together with the probabilities of a large language model, or some weird hybrid, or something else we haven't thought of. That may be the next phase of artificial intelligence. But of course, Demis and others might say that LLMs, through self-improvement, may develop that world model on their own—maybe they would get to the place where you'd think we humans would need to inject something extra. Who knows.

The guest illustrated AI's predictive ability with a simple example: ask an AI what happens if you push a mug across the table, and it will answer that the mug slides, possibly knocking into a plate or falling off the edge depending on how hard you push. It can do this because it has read everything on the internet — physics textbooks, conversations about falling objects — and bases its response on those collections of words.

Asked whether that isn't exactly what humans do, he admitted we don't fully know how human intelligence works. Evolutionary psychologists have long asked whether we are a blank slate that simply learns from experience, or whether something is innate. He believes there is strong evidence for an innate physical intuition about the world, passed down through generations — not that you can hand your children F=MA or Newton's laws because you learned them, which would be silly, but that there is an evolutionary advantage to quickly assessing the world and figuring out how to act. Our forebears who couldn't throw a rock or spear didn't get the next meal; those with a slightly better predisposition for ingesting experience and codifying it into an understanding of how the world works were the ones who survived and passed that predisposition down. Babies are born with some intelligence — they grip fingers — in part because the parent passed it on.

The interviewer suggested an AI could do the same, and he agreed: AI systems themselves are evolutionary, passing the achievements of generation N to generation N plus one as part of the iterative process.

"What a weird world we're living in," the interviewer remarked. He agreed, noting that people throughout history have probably all thought "what a time to be alive" — but we are living through a rapid change that is somewhat unprecedented.

Asked how one should contend with this, since people are concerned and don't love change at this rate, he said people will cope in many different ways — and that it helps to stay open and excited about the possibilities, with the flexibility to embrace change.

One thing I've been thinking about a lot lately is Elon Musk saying we're heading to an age of abundance. He has shifted his perspective from this doomerism to an age of abundance, and part of that narrative is that we won't need money, because with robots, AI, and more energy, we won't have to worry about food or bills. But humans still pursue scarcity. We pursue status — we want the home at the end of Miami Beach with the best plot, and there are finite numbers of those things. Even academic discovery and Nobel prizes are finite, and we pursue them. So one could argue there will still be competition, that humans will still strive for things.

Why do we do that? Part of the reason is that if you know you're going to die, if you know your life is finite, there can be an urge to make your mark, to do something special, to be out of the ordinary, so that in some symbolic way you continue to exist after you're gone. Elon Musk, after he's gone, people will talk about him for a long, long time, because he had a radical impact — and who knows where it will go, since he has many productive years still left. So yes, there is this natural tendency to seek something that allows you to symbolically persist after you are gone.

Now, if we all realized that this is part of human motivation, maybe as a species we could begin to shift. It isn't the case that we have to always be the way we have been historically. Kings want to be kings in part because they have the power, and with power comes hegemony — in a sense they exist even when they are gone, through their progeny. Why do we have descent of monarchies? Because the king continues to exist through the king's progeny. So perhaps we could get to a place where we shed that way of being, because in a world of abundance the whole notion of scarcity is artificial — we're only inventing it so that we can feel good and special, and maybe we can find other ways to feel good and special. I allow for that possibility, but I'm not particularly optimistic that's what will happen. On the flip side, it's easy to say money won't matter when you have a trillion dollars — you just have to see this stuff in context.

We have one more little souvenir thing here, and it's kind of what we're talking about, which is robots. Do you think AI can become

The question was whether AI can become conscious, and how confident I am about it. I'm fairly confident. As I said before, I see consciousness as a physical process, and it's a matter of achieving that physical process in an artificial system one way or another. I don't see a fundamental barrier to doing that.

The interviewer raised how laws might have to change if we considered robots conscious—if they lived in our homes, would they get rights? I think that's something we should think about now. It's a very tricky idea: it's one thing to pull the plug on a light bulb, another to pull the plug on a sentient system, and we may well find ourselves in that situation.

The challenge is that it's virtually impossible to ever prove sentience. As we look at each other, we each afford the other consciousness—I believe you're conscious, and you believe the same of me. Why? We're physically constructed more or less the same way, we're part of the same species, and our body language and speech are similar enough in the grand scheme of things that we afford the other what we internally experience. But we don't actually know it. You don't know that I'm conscious; I could be faking it—a really good artificial zombie that speaks but has nothing going on inside. And the same could be true of you.

With artificial systems that don't share our lineage, aren't part of our species, don't necessarily look like us, and didn't have our upbringing, will we be willing to agree they're conscious if they tell us so? If a system says, "I really am. I'm feeling exactly the thing I read about on your internet—I'm feeling love, or I'm feeling depressed," we respond, "No, no, no. You're just mimicking what human beings have said. There's no world inside there." How do we ever get beyond that? I don't know the answer, even though I strongly believe there's no logical, fundamental barrier preventing a system from saying that and telling the truth.

The interviewer noted that electrical circuits within such a system could cause it to behave a certain way when scared or in love—cuddling you because that's what the circuits told it to do. Yes, but is there an inner world associated with that, or is it just going through the motions?

Here's how I think it will likely turn out: these artificial systems will start saying such things, and at first we'll dismiss it—"come on, it's not real." But we'll get used to them behaving as if they have inner worlds like ours, and sooner or later, through a process perhaps so slow we don't even notice it, we will afford them consciousness. We'll talk to them as if they're conscious, imagine that they are, and ultimately treat them as if they are. I hope we do that in a good way.

We have a terrible history of treating conscious beings respectfully, even within our own species—we've done terrible things to each other despite affording each other consciousness, and we do terrible things to animals. I'm vegan; I don't eat any animal products. I don't think that's the way to treat a living system with some level of conscious awareness. I do eat plants, because I don't think plants have a sufficiently high level of consciousness, if any, for me to be concerned about cutting and eating them.

"Have you always been vegan?"

"I've been vegan since 1993. I've been vegetarian since I was 9 years old. That's how I treat living or conscious beings, because I don't think it's right to treat them that way."

"What was the catalyst for that decision?"

When I was 9, it was very flat-footed: my mother cooked spare ribs, and for the first time I realized what meat was. I'm a city kid—meat came wrapped in cellophane from the supermarket. But when the meat came on a bone, I thought, "Wait a second, what is that? Oh my god, this is the rib of an animal." I said I was never going to eat meat again. I still ate cheese and milk at that point—I wasn't vegan yet. But when I was a professor at Cornell in the 1990s, I visited an animal sanctuary in Watkins Glen, New York, where they showed you what happens to these animals when they're treated like milk factories—what happens to the cows—and it's awful. Within 72 hours of that visit, I said, "I'm done. I'm not participating any longer. I'm not going to eat or drink any substance that came from an animal." That was the end for me.

"For some people it might be somewhat surprising to hear that from a physicist."

I agree it's inconsistent. If I believe, as I do, that you and I are just a collection of particles in one configuration, and a cow or a pig is just a collection of particles in a different one, then what does it matter whether you eat particles in the configuration we call a cow or the one we call a carrot? But as a human being, I allow the other parts of me to shine through. I look at that collection of particles that happens to be a cow and say, "It just doesn't feel right to eat that collection of particles. It just doesn't feel right, and I don't want to do it."

The interviewer asked Brian Greene which of the ideas in his many books is most fascinating or relevant to everyday life but hasn't yet been discussed. Greene said he would lean toward the fascinating rather than the practical, as he tends to be esoteric. In Until the End of Time, he tried to give readers a sense of how we got here, and then turned to the future to describe what we will be if we extrapolate arbitrarily far ahead.

The Empire State Building metaphor

To convey what the universe will be like if the current understanding of the laws of physics is projected arbitrarily far into the future, he used the metaphor of the Empire State Building. Imagine every floor represents an era of time ten times longer than the floor below: the ground floor is one year, the next ten, then a hundred, a thousand, ten thousand, and so on, with exponentially larger timescales as you climb the stairs.

In this framing, everything from the Big Bang until today brings you only to about the tenth floor—we are roughly 10^10 years from the beginning. From there, Greene walks the reader through the key markers of the far future.

Milestones of the far future

When the interviewer joked, "Stop when it gets scary," Greene replied that it's scary the whole way.

• The 11th floor (10^11 years): The sun swells bigger and bigger, swallowing the close planets Mercury and Venus, and possibly the Earth as well—we're not sure. But it will certainly make life on Earth virtually impossible because the planet will be so hot.
• The 12th floor (10^12 years): The universe is expanding ever more quickly, driving the distant galaxies away faster than the speed of light, so we won't see them. A future astronomer looking into the deep night sky will see only darkness; everything will have drifted beyond the cosmic horizon. Asked why the galaxies recede faster than light, Greene explained that we have found, surprisingly, that the expansion is speeding up, and there doesn't seem to be a limit. He cautioned that every prognostication rests on assumptions—namely, that what we measure today will persist, much as AI researchers assume today's growth will continue. Is that true? We don't know, but it's our best guess from the laws of physics.
• The 20th floor: If the Earth survived the swelling sun on the 11th floor, it will spiral into the dead sun, because it loses energy through gravitational radiation—ripples in the fabric of space will cause its orbit to decay, and we will crash into the dead sun. That will be the end of the Earth by about the 20th floor, though the end of life itself could easily have come much earlier, on the 11th floor.
• The 30th floor: The stars themselves will spiral into the black hole at the center of most galaxies—our Milky Way has one at its center. Just as the Earth spirals into the sun, the stars will spiral into the black holes, leaving black holes as essentially the only remaining macroscopic structure in the universe.

Asked whether this has happened elsewhere, Greene said certainly: we see things spiraling into black holes, and even at the center of our own galaxy there is evidence of activity, part of which is stars being ripped apart as they fall in. So it's quite clear this will happen. The interviewer then asked whether this is the end of all galaxies as they spiral into the black hole.

Yes. After about 10^30 to 10^35 years, it's hard to imagine that galaxies as we know them will persist. The stars will have burned out—around the 14th floor of this timeline, almost all stars will have used up their nuclear fuel.

"All of them?"

Just about all of them will fade to black.

"All of them in the whole universe?"

Yes—or at least the observable universe. I can only really speak to the part we have direct access to. By the 30th floor, those dark stars will spiral into black holes.

"So there could be no life anywhere."

It's hard to imagine, but let's be optimistic and imagine that somehow there is some life floating in the darkness, because I want to get to two more interesting milestones. By the 38th floor, wherever that is, we believe that protons—which are the heart of matter—will likely themselves disintegrate. So if any remaining organized matter of the sort we're familiar with is somehow persisting somewhere, it too will disintegrate by the 38th floor. That's our belief based on our laws of physics.

"There'll be no matter anywhere."

There will be fine particulate constituents, because when a proton falls apart it doesn't disappear—it falls apart into other, more refined particles. Those will continue to persist, but the aggregate called the proton will not.

If we keep going to roughly the 50th floor, this is an interesting point: imagine that somehow some conscious being still exists in the darkness of space—I don't know how, perhaps dark clouds of particles somehow carrying electrical signals that give them some kind of conscious self-awareness. After the 50th floor, that conscious being will likely think its final thought. The reason is that thinking generates heat, and that heat has to be carried away. It's easy for heat to leave us in a room like this, because there's plenty of room for that so-called heat, or entropy, to be absorbed. By the 50th floor, the universe won't be able to absorb the heat generated by the process of thought itself. So any thinking being, when it thinks one more thought, will burn up—it will fry in the entropic waste, the heat generated by thought itself. That's why, when you asked earlier whether we'll live forever, I was thinking maybe we'll live 500 or a thousand years, but by this point, no way. It's hard for me to imagine that we will continue to persist.

By the 68th floor, black holes themselves will begin to disintegrate. Stephen Hawking showed us that black holes are not fully black: when you take quantum physics into account, they can emit particles that waft outward, causing the black hole to shrink and, over time, disappear. A black hole with the mass of our sun takes about 10^68 years to evaporate; a gargantuan black hole like the ones we think might exist—billions and billions of times the sun's mass—takes about 10^100 years, at the very top of the Empire State Building. By the time we get there, 10^100 years into the future, we suspect it will just be detritus of particles wafting through an ever larger, ever colder, ever quieter universe. That's the future we're looking at.

"And then when does the big bang kick off and it all start again?"

Well, it's interesting—you're absolutely right to ask what about beyond the top of the Empire State Building. This whole thing is speculative: we're extrapolating from laws of physics we've developed by thinking about physics here, and assuming they apply all the way into the exponentially far future. Is that true? I don't know. It's our best guess based on what we know today, but it is definitely an assumption. If you extrapolate even further—an exponential of an exponential of years into the future—could particles somehow re-coalesce just by chance and build a state of matter that might generate another big bang, so it all starts again? Yes, definitely within the realm of possibility.

"It's not the most appealing future."

Well, I agree. But here—can I tell you my take on it?

My take on it is this: when you realize that the far future is a realm of darkness, and that before life emerged on planet Earth the universe was a chaotic conglomeration of particles, it becomes clear that we're living in a special time — a special region in the cosmological unfolding, when living beings like ourselves and consciousness as we experience it are compatible with the state of the universe. How wondrous, how spectacular, that there is this brief window of time. Nabokov had this wonderful saying: we live in a brief crack of light between two eternal stretches of darkness. How beautiful to exist in that brief crack of light. That, to me, is where the gratitude I mentioned earlier comes from. You asked what gives my life meaning, and it's the fact that we live in this very special era of cosmological history, where for a small period of cosmic time living beings can stand up, look around, contemplate themselves and the universe, gain some insight — and then it all goes away. But how wonderful that there is that brief moment we are now inhabiting.

It is wonderful when you say it like that.

Frank, what I think is most special about you isn't just the extent of your intelligence — it's your ability to communicate in a way that makes the complex both interesting and understandable. I've seen that throughout all of your work and videos; I was watching your TED talk with my fiancée a couple of weeks ago, and it's the closest I've ever come to actually understanding string theory.

Oh, great. I'm glad to hear it — thank you very much.

I have a couple of your books here, including the most recent one, from 2020, Until the End of Time: Mind, Matter, and Our Search for Meaning in an Evolving Universe. Your work makes us feel so many things, but maybe the most important feeling it creates is connection. We're all searching for connection in an increasingly lonely, isolated-feeling universe, and that's why I love having these conversations on the show. I sometimes joke to the team that the "C" in Dio stands for connection — one of our goals is to make people feel more connected in some capacity, and you most certainly do that. Thank you for having that impact on the world. It makes us look up, not just down, and that's a wonderful thing.

Well, thank you.

I'll link all of your books below so people can continue this conversation through them. Is there anywhere else people can go to support or learn more about your work?

Considering the World Science Festival — worldsciencefestival.com. You'll find all sorts of conversations of the sort we're having here. Also, I'm working on a new AI education system where AI will help people understand these ideas better. It should launch in about a year or so, so keep an eye out for that.

Where do people follow you?

On Twitter or Instagram we post semi-frequently, but more at our YouTube channel, World Science Festival.

Of the four books I have in front of me, which one would you direct different groups of people to read?

My favorite, if you frame it in that flat-footed way, is Until the End of Time, because it's the more philosophical one and covers the kind of ideas we spoke about today. But if you want the hardcore of what string theory is, or the new insights into space and time, some of the other books are more directed that way.

Until the End of Time: Mind, Matter, and Our Search for Meaning in an Evolving Universe. We have a closing tradition where the last guest leaves a question for the next. The question left for you is: what might be worth even more than the thing that matters most to you?

I would say the things that matter most to the people I care most about. The things that matter to my wife and my kids feel much more important than the things that matter most to me, because I'm so involved in those relationships. That, to me, is the thing that trumps all else — the things that matter more to them.

It's a beautiful answer. I wasn't expecting that, but it's very true. Brian Greene, thank you so much. It's been an honor.

My pleasure. Thank you.

YouTube has this new algorithm that knows exactly what video you'd like to watch next, based on AI and all of your viewing behavior. The algorithm says this video is the perfect video for you — it's different for everybody watching right now. Check it out; I bet you might love it.