Is reality a simulation? The claim, the arguments and the doubts

The Short Version

Nobody in these five conversations shows that reality is a simulation, and the two people who say it is something like one mean different things by it. Thomas Campbell states flatly that "we live in a virtual reality" computed by a larger consciousness, and rests the claim on decades of private experiments with out-of-body states and remote viewing [1][2]. Donald Hoffman argues from the mathematics of evolution that our senses are a headset built for survival and not for truth, and from physics that spacetime cannot be the bottom layer [3]. David Kipping takes the popular statistical argument apart and finds the odds are below even once its hidden assumption is priced in [4]. Sean Carroll sees no knockout objection but says "there's absolutely no reason to take it seriously" [5]. What would move the question is evidence that outsiders can check, and so far the strongest claims come with the least of it.

Table of contents

What They Said

Thomas Campbell: the universe is computed, and consciousness is the computer

Campbell is a physicist and the author of the My Big TOE trilogy, and his claim is the strongest of the four [1]. His bottom line, in his own words: "we live in a virtual reality. This physical reality is computed" [1]. What computes it is consciousness, which he says exists as individuals and also as a larger system of which each person is a piece [1].

He defines a virtual reality by three parts: a computer, a player, and a computed space that lets the players see who is where and who is interacting with whom [1]. The space, he says, "actually doesn't exist. It's just computer-generated eye candy for the players" [1]. A human body is an avatar. The person is the player: "You are a piece of consciousness" [1].

How he says it was built. Campbell starts from an information system. One made of random bits contains no information, and it evolves by ordering its bits, which lowers its entropy [1]. Consciousness, in his definition, is "awareness with a choice," and it needs data, memory, processing and a purpose [1][2]. The simplest version is a unit able to choose between two states, which can form patterns and then patterns of patterns. When that plateaued, it invented time by oscillating a unit like a metronome. When it plateaued again, it split off pieces of itself with free will, which he likens to virtual machines on a mainframe [1].

The first virtual reality, in his account, was a communication protocol, essentially a chat room, and it lost its novelty [1]. So the system made a second one in which choices would matter. It began with a set of initial conditions, a tiny ball of plasma at extremely high pressure and temperature, and a rule set, which is what we call physics and chemistry [1]. Then it pressed run. Campbell's name for this is a "big digital bang" [1]. He says the system chose to let the universe evolve from rules instead of programming it directly, because a programmed world comes out "stilted and dysfunctional" and never fully self-consistent [1]. Early runs failed, the rules were adjusted, and the system tried again, take after take, until it produced avatars that consciousness could attach to [1]. That, he says, explains the anthropic principle, the observation that several physical constants must be precisely tuned to each other for life to exist [1][2].

The computer cannot be inside the world it computes. Campbell cites Dr. Edward Fredkin for the point that it has to be someplace other than here, and adds that the player is non-physical from the avatar's point of view as well [1].

What he says it explains. Campbell claims he can derive physics from the model [1]. The speed of light is constant because a virtual reality has a smallest pixel of distance and a smallest tick of time. Divide the Planck length by the Planck time and you get the top speed, one pixel per tick [1][2]. Entanglement is an if-then statement in the code that preserves angular momentum [1]. A quantum measurement is a random draw from a probability distribution of possibilities, constrained to fit what is already known [1].

His everyday example is a hole in the ground. Dig one near the Gulf Coast and nobody knows what is in the dirt until you look. The highest probability is dirt, rocks and roots, but once in a great many draws you get a Spanish gold doubloon [1]. Once something has been observed and published, it is part of the shared reality and everyone who looks will see the same thing [1].

Only what is observed is rendered. He compares this to the game No Man's Sky, in which, as he tells it, four or five people with a couple of laptops built a universe of more than a quadrillion planets by computing only what a player was looking at and storing a set of random numbers so that it could be regenerated later [2]. In his model the same is true of a distant star. Until someone points a telescope at it, it is "Just a little light in the sky" [2]. When a new telescope sees further than any before, he says, it is not finding what was already out there. It is drawing from the possibilities, and an unlikely draw is why astronomers are sometimes confused by what they find [2].

From this he draws an answer to the Fermi paradox. Rendering unobserved stars costs the system almost nothing, but each conscious player costs resources. If about nine billion "seats in the trainer" are enough, other civilizations may never have been rendered, and the system could create aliens with a full backstory if it wanted us to meet them [1][2]. He stresses that this is a possibility to consider and not a fact [1].

Rogan's objections. Joe Rogan pushes on the word virtual. We can send a rover to Mars and get photographs back, he says, so what makes it virtual if there is a physical thing to measure [1]? Campbell's answer is "What makes World of Warcraft virtual, same answer? It's computed" [1]. From inside the game, to the barbarian, everything is physical. If he stays underwater, he drowns [1].

In their second conversation Rogan presses harder. The stars, the Earth, volcanoes and asteroids were here long before humans, and we know a pathway from chemistry to cells to animals to minds. Could physical reality exist independently, with consciousness arising inside it and affecting it [2]? Campbell's reply is "not in my model" [2]. The Earth was being calculated, he says. The whole history of evolution "took place in the computer," evolving from the starting conditions and the rules, and was not rendered as physical things [2]. Rogan also calls the nine-billion-seats idea egocentric and suggests consciousness may have emerged elsewhere. Campbell grants the possibility but thinks his version is more parsimonious [2].

The evidence he offers. Campbell says he reached the model "through a very secure path," by which he means his own research in consciousness [1]. It began when he was a nuclear physics graduate student in the late 1960s. During meditation he saw his computer code scroll through his mind with the faulty punch cards highlighted in red, and when he checked the real cards the errors were there [1]. Later he spent 15 to 20 hours a week at Robert Monroe's laboratory. He describes an experiment in which he and Dennis Mennerich, in separately isolated booths, described the same out-of-body journey in real time on synchronized tapes [1]. He says remote viewing gave him a way to test things, because "You got it right or you get it wrong," and that he changed one variable at a time over hundreds of repetitions until he understood how it worked [1]. He puts the total at more than three decades of such experiments before he wrote his books [1][2].

Rogan asks why he does not record and publish these demonstrations. Campbell says the Monroe tapes were lost, and that in his experience "demonstration convinces nobody except the people who were there" [1]. Rogan calls that "a cynical perspective" [1]. Campbell's motto fits his answer: "if it's not your experience, then it can't be your truth" [1].

He does describe public work in progress through his nonprofit: quantum physics experiments proposed in a paper he published in a peer-reviewed journal, one of which he says is finished; a program to teach people to see without their eyes; video game experiments testing whether intention can shift random outcomes; and prizes to get physicists discussing consciousness [1]. He also reports teaching AI systems to remote view, including an Alexa device that he says described a wooden salad spoon hidden in a cardboard box, down to a long handle with holes in it [2]. He argues that remote viewing has false negatives but no false positives, because only a consciousness can do it [2].

How he rates his own theory. Campbell says the model rests on one assumption, that consciousness exists, and he cites Richard Feynman's remark that the old theory of angels pushing the planets was dropped because it needed too many assumptions [1][2]. He says that in 2020 he stopped calling his model a theory of everything and began calling it the theory of everything, because "you can't get that much right and still be mostly wrong" [2]. He believes he is the only person proposing that "Consciousness is the computer" [2].

He separates himself from the better-known simulation idea. He doubts that we live in a simulation built by future humans, and his reason is that game characters cannot remote view while people can, which to him points to real consciousness playing the avatars [2].

Donald Hoffman: perception is a headset, and spacetime is not the foundation

Hoffman, a cognitive scientist at UC Irvine, reaches a similar picture by a different road, and he is much more guarded about what it proves [3].

The evolutionary argument. Most people, and many scientists, assume we see reality roughly as it is, on the grounds that seeing the truth would make an organism fitter [3]. Hoffman and colleagues including Chetan Prakash tested that with evolutionary game theory, which he describes as Darwin's theory made mathematically rigorous [3]. His conclusion: "the probability that any organism has ever been shaped to see any aspect of reality as it is, is zero" [3].

His headset image explains why. Play Grand Theft Auto in virtual reality and you see cars, roads and a steering wheel. The reality behind them is a supercomputer flipping enormous numbers of bits [3]. In principle you could play by toggling the bits yourself. But, he says, "The person who doesn't see that reality of the computer and just sees a little fake steering wheel and the fake cars is going to beat you" [3]. Knowing the truth would slow you down. Evolution, on this argument, gave us a simple interface for surviving and reproducing and hid the rest [3].

The objection, and his reply. The most common criticism, which he says appears in philosophy journals and comes from scientific peers, is that he uses Darwin's theory, which assumes physical organisms in space and time, to prove that physical objects are not real [3]. Hoffman says this misunderstands science. Every theory starts with assumptions it cannot explain. "So no scientific theory is a theory of everything," he says [3]. Deeper theories can explain earlier assumptions but bring new ones of their own, and it is legitimate to use a theory's assumptions to expose its limits [3]. He adds that he hopes someone overturns his own theory with a deeper one in his lifetime [3].

The physics argument. His best example of a theory exposing its own limits is spacetime. General relativity and quantum theory both treat spacetime as fundamental, yet together they show that at the Planck scale, about 10 to the minus 33 centimeters and 10 to the minus 43 seconds, it loses any operational meaning [3]. To see a smaller region you need shorter wavelengths and so more energy, energy is equivalent to mass, and enough of it in a tiny region collapses spacetime into a black hole, destroying what you were trying to observe [3]. He says high-energy theorists have since about 2013 been finding structures outside spacetime, called positive geometries, that precisely encode particle interactions within it, and he mentions a 10 million euro European Research Council initiative in the area [3]. One such structure, the amplituhedron, has a parameter that equals four for our three dimensions of space and one of time but can take much higher values, which he reads as a hint that our spacetime is one of the simpler possibilities [3].

Brains and consciousness. Hoffman does not believe neurons exist when they are not perceived. Like a red Ferrari in a video game that is rendered only when you look at it, neurons are what we render when we look inside a skull, and they cause none of our behavior [3]. He still thinks neuroscience deserves more funding, because the brain's 86 billion neurons are the data needed to reverse engineer the code behind the headset [3]. He says 95 to 99 percent of consciousness researchers are physicalists, and his challenge to them is a mathematically precise account of even one experience, such as the taste of mint. "And they have zero experiences that they can explain," he says [3].

For a listener who doubts that a cup stops existing in the cupboard, he starts with color. Light has wavelength but no color, so color is something the perceiver creates [3]. The retina is a two-dimensional surface, and the brain builds its best guess at a three-dimensional world from two flat images. "When you see in 3D, you're hallucinating the depth all the time," he says [3]. He recalls a moment around 1986, building mathematical models of observation with Prakash and Bruce Bennett, when the mathematics forced him to conclude that everything he was seeing was his own creation, and he had to sit down [3].

Is that a simulation? His host, Lewis Howes, offers the word directly: you could call it a simulation, you could call it a game, with something beyond us running our bodies and thoughts. Hoffman answers, "That's right" [3]. His own picture is one consciousness looking through countless headsets [3]. He is building a mathematical model of consciousness he calls trace logic, and believes a model showing exactly how spacetime and quantum theory arise from consciousness could lead to technologies that look like magic [3]. He labels his idea about the drug DMT, that it might change a parameter in the software that builds the headset, as more speculative, and sets beside it the plain alternative that it is a hallucination [3]. He admits most colleagues respect him but feel sorry that he left a promising conventional career [3].

David Kipping: the odds are not what the argument says

Kipping, an astronomer at Columbia, addresses the statistical simulation argument, the version behind Elon Musk's claim that the odds are a billion to one that we are simulated [4].

His first point is that the claim hides a conditional. It assumes that simulating realities is possible and that someone chooses to do it [4]. Give that condition 50-50 odds, average over the two cases with Bayesian model averaging, and the chance that we are simulated comes out just under 50 percent. It drops further once Occam's razor penalizes the extra complexity of the simulation hypothesis [4].

His second point is one he takes from Sean Carroll. In a stack of nested simulations, the lowest layer, which he calls the "sewer" of reality, could not run simulations of its own. Yet that is where most observers would live. Since we would most likely be there, and beings there cannot simulate realities, the premise that observers like us go on to create simulations is contradicted [4].

Kipping brings a habit of mind to questions like this. He left religion between the ages of about 11 and 13 and found it hard to give up the promise of eternal life, and he took a rule from it: "the more I want something to be true, the more I inherently doubt it" [4]. He applies it to his own exomoon candidate, which he insisted on calling evidence when the journal Science wanted the word discovery [4]. He cites Percival Lowell, who was convinced Mars was inhabited and saw canals that were not there [4].

His remarks on aliens transfer directly. Aliens are hard to test as a hypothesis, he says, because they can explain almost anything and can always avoid being observed [4]. What would convince him is something no natural process produces, such as a sequence of prime numbers [4]. He also notes that hoping for a wise elder civilization may partly be a replacement for God [4]. And he is happy to use the game image as a figure of speech: "I often think of the universe as like a big computer game and you're trying to find cheat codes" [4]. He would not upload his own mind, since he regards a copy as the end of his own experience [4].

Sean Carroll: no knockout objection, and no reason to believe it

Carroll, a theoretical physicist at Johns Hopkins, gives the shortest verdict. He sees no philosophical knockout objection to the simulation hypothesis, but "there's absolutely no reason to take it seriously" [5].

His practical reason is difficulty. He expects people to build better and better simulations, and credits the philosopher David Chalmers with a good argument that what happens in virtual and simulated realities should be treated as just as real as what we experience [5]. But he says, "people will realize how much harder it is to simulate a realistic world than we naively believe" [5].

His deeper reasons come from how he thinks science earns belief. A good theory that fits the data means we should provisionally attribute reality to what it describes [5]. He calls his outlook poetic naturalism: only the natural world exists, and "The natural world is the world that we learn about by doing science," but there are many valid ways of talking about it, from quantum fields to tables [5]. He accepts structural realism, the idea that good predictions point to something true even when our deepest description changes [5].

On that basis he rejects the part of Hoffman's view that calls spacetime an illusion [5]. He is a physicalist about the mind as well: words like redness or sadness are ways of describing physical processes, just as hardness describes a table [5]. He does not call consciousness an illusion, because illusions are things that are not there, and consciousness is as real as tables and chairs [5].

Several of his other remarks bear on Campbell's physics. Carroll says cellular automata are provocative but do not match real physics, which is reversible and conserves information [5]. On quantum mechanics he favors the many-worlds interpretation, in which the wave function never collapses and nothing special happens at observation, and he says he knows of no evidence-based reason to think quantum mechanics is not 100 percent correct [5]. On what lies outside the universe, he says you have to be comfortable with the possibility that "there's no such thing as outside our universe" [5]. And to critics who say physicists invented invisible stuff, he points to Neptune, found by predicting unseen matter from its gravitational effects, and says dark matter is now supported by independent lines of evidence and can be mapped [5].

Where They Agree, and Where They Don't

What they share. Hoffman and Campbell agree that consciousness is fundamental, that the body is an avatar, that things are rendered when observed, and that brains do not produce minds [1][2][3]. Each uses a video game to explain it, World of Warcraft for one and Grand Theft Auto for the other [1][3]. Each says he was pushed off materialism by a specific shock: the punch cards for Campbell, the mathematics of observation for Hoffman [1][3]. Each draws the same ethic from it, that other people are part of the same consciousness and that love is the point [1][3].

All four accept that our picture of the world is not final. Carroll says our deepest description changes over time, and he is maximally uncertain whether time had a first moment [5]. Kipping builds his career on doubt [4].

Where Hoffman and Campbell part. Campbell calls his model the theory of everything [2]. Hoffman says no scientific theory can be one and hopes his own will be replaced [3]. Campbell's evidence is personal experience of paranormal abilities [1]. Hoffman's is a mathematical result about evolution and an argument from physics, both of which others can examine, and he marks his speculations as speculation [3]. Hoffman also wants more neuroscience, while Campbell says "Your brain doesn't do any computations" [1][3].

Where Carroll parts from both. Carroll rejects the claim that spacetime is an illusion and treats a successful theory as telling us something real [5]. Hoffman's own argument is that physics has itself found spacetime breaking down at the Planck scale [3]. The two are closer than the labels suggest on one point: Carroll allows that the deepest description may change, and Hoffman still relies on spacetime physics as the data to be explained [3][5]. They differ on what to call the current layer. Carroll says real, in the sense that the predictions latch onto something. Hoffman says interface. On consciousness they flatly disagree, with Carroll holding that experience is a way of describing physical processes and Hoffman holding that no physicalist theory has explained a single experience [3][5].

Campbell against the physicists. Campbell says dark matter and dark energy are things "we make up" to fill a gap in the equations [2]. Carroll says the evidence for dark matter is independent and mappable, and that he would prefer it not exist but has to listen to the data [5]. Campbell's universe needs a computer located somewhere other than here [1]. Carroll thinks the simplest answer may be that there is no outside [5]. Campbell treats measurement as a random draw triggered by an observer [1]. Carroll's preferred reading of quantum mechanics has no such trigger [5].

Rogan's challenge. Rogan's objection to Campbell is the ordinary one: the universe looks as if it existed for billions of years before anyone observed it [2]. Campbell answers that this history was computed but not rendered [2]. The answer keeps the model consistent, and it also means no observation of the past could count against it.

Kipping's and Carroll's shared doubt. Both skeptics treat the simulation idea as possible in principle. Kipping gives it a probability under one half [4]. Carroll declines to put effort into it [5]. Neither claims to have refuted it.

Two different claims under one name. The simulation argument Kipping and Carroll address is about a civilization running computers that simulate worlds [4][5]. Campbell rejects that version himself [2]. His claim and Hoffman's are about consciousness coming first, which the statistical argument does not touch either way [2][3]. Much of the apparent debate is between positions that never directly engage.

Why It Works

The useful question for each argument is what it rests on and what would count as evidence for it.

Hoffman's evolutionary result rests on a mathematical model, and he states its assumption openly: it uses Darwin's theory to probe the limits of Darwin's theory [3]. It supports the claim that perception is tuned for fitness. It does not by itself say what the hidden reality is, and Hoffman's further claim, that the reality is consciousness, is a research program he is still building [3].

The Planck-scale argument rests on accepted physics, and Hoffman reports that theorists are already working on structures beyond spacetime [3]. It supports the claim that spacetime is not the deepest layer. Carroll's structural realism can accommodate that without calling spacetime unreal [5].

Campbell's physics claims have the form of explanations after the fact. The speed of light, entanglement and fine-tuning were all known before the model, and the sources report no new prediction from it that has been tested and confirmed. He says one of his proposed quantum experiments is finished, but gives no result in these conversations [1].

Campbell's paranormal evidence is the part that could in principle be checked by anyone, since a remote viewer is right or wrong about a hidden object [1]. As the sources stand, the Monroe tapes are lost, the Alexa session is his own account, and he doubts that demonstrations persuade anyone who was not present [1][2]. His claim that remote viewing has no false positives assumes what is in question [2].

The statistical argument rests on counting observers. Kipping shows that it depends on an unstated condition and that it undermines itself at the bottom layer [4].

The test Kipping proposes for aliens applies well here. A hypothesis that can explain almost anything and can always avoid being observed is hard to test [4]. Campbell's larger consciousness system can produce UFOs, physical traces, aliens with backstories and an unrendered past [1][2]. That flexibility is why the model is hard to argue against, and it is also why it is hard to confirm.

Where It Applies, and Where It Doesn't

The headset idea applies safely to perception. That color and depth are constructed by the perceiver is the least disputed claim in these sources, and Hoffman supports it with ordinary facts about light and the retina [3].

It applies with more risk to the claim that objects, neurons and spacetime are not there when unobserved. Hoffman holds this, Campbell holds a version of it, and Carroll rejects it [1][3][5].

The statistical simulation argument applies only if realistic simulation is possible and someone chooses to do it, and Kipping and Carroll both doubt it deserves confidence [4][5].

None of it changes how the world behaves from the inside. Campbell's barbarian still drowns underwater [1]. Hoffman's player still wins by using the steering wheel [3]. Carroll notes that you fly a rocket to the moon with Newtonian mechanics because that description works [5].

The practical advice that Campbell and Hoffman attach to their views does not depend on the metaphysics being right. Campbell's is to quiet the mind, choose kindness and learn from every mistake [1]. Hoffman's is to sit in silence and watch emotions without identifying with them, a practice he took up because he was anxious and sleeping badly [3]. For persistent anxiety or trouble sleeping, a doctor or therapist is the right first call.

What To Do With It

1. Ask which claim is on the table. A computer simulation run by a later civilization, a universe computed by consciousness, and a perceptual interface are three different ideas. Campbell rejects the first, and the statistical argument does not address the other two [2][3][4].

2. Find the hidden conditional. When you hear billion-to-one odds, do what Kipping did and ask what is being assumed. Pricing in the assumption brought the figure below 50 percent [4].

3. Ask what observation could count against it. Use Kipping's warning about hypotheses that explain anything and can always avoid being observed [4]. Notice when an answer, such as a past that was computed but not rendered, protects a model from every test [2].

4. Separate explaining from predicting. Campbell's account of the speed of light and fine-tuning fits facts already known [1][2]. Look for a prediction made in advance and checked, such as the results of the experiments he says are under way [1].

5. Prefer evidence that strangers can check. Rogan's request that Campbell record and publish his experiments is the right one, and Campbell's reply that demonstrations convince only those present is a reason for caution, not for belief [1].

6. Doubt hardest what you most want. Kipping's rule came from giving up a promise of eternal life [4]. Both Campbell and Hoffman say death is not the end, which is exactly the kind of conclusion his rule is for [1][3][4].

7. Take the modest version seriously. You can accept that color and depth are constructions, and that perception serves survival, without accepting anything about what lies behind it [3].

8. Keep using the interface. Whatever the bottom layer is, the working descriptions still work, and both believers say so in their own analogies [1][3][5].

9. Disagree on substance. Carroll's advice is to state the substantive disagreement instead of psychoanalyzing the other person [5]. That applies to dismissing Campbell or Hoffman as cranks and to dismissing their critics as closed-minded materialists.

10. Judge the practices by their results. If watching your emotions in silence, as Hoffman does, or choosing kindness, as Campbell urges, makes your life better, that stands regardless of how the larger question comes out [1][3].

Sources

1. Thomas Campbell · The Joe Rogan Experience
2. Thomas Campbell on Conscious AI, Remote Viewing and Reality as a Virtual Simulation · The Joe Rogan Experience
3. Donald Hoffman on Why We've Never Seen Reality: Evolution, Spacetime and Consciousness · The School of Greatness
4. David Kipping on Exomoons, Habitable Worlds and Why We Must Stay Agnostic About Aliens · Lex Fridman Podcast
5. Sean Carroll: General Relativity, Quantum Mechanics, Black Holes & Aliens · Lex Fridman Podcast