2016 Isaac Asimov Memorial Debate: Is the Universe a Simulation?
2016 Isaac Asimov Memorial Debate: Is the Universe a Simulation?
[Neil deGrasse Tyson speaks at a podium.]
NEIL DEGRASSE TYSON: Welcome back. This is the 17th Annual Isaac Asimov Panel Debate. And we’ve been going strong ever since the year 2000, when an idea surfaced in the hearts and minds of the family of Isaac Asimov, exploring a way for his memory to be preserved in the programs of this institution. And Isaac Asimov was a friend of the American Museum of Natural History.
Much of the research for so many of the books that he wrote took place in and around the halls and in our libraries. And so perhaps there’s no more fitting tribute to him and to his memory, than to keep this celebration going. So, thank you for attending.
We are also streaming live on the Internet. And I’m your host for this evening, Neil deGrasse Tyson. I’m the Frederick P. Rose director of the Hayden Planetarium.
[APPLAUSE]
TYSON: Just a couple of newsy notes. This year we sold out in three minutes. And it’s not a particularly sustainable model. So, we’re going to have top people looking at how to improve that next year. We don’t know how yet, but the least we can do is offer it live streamed on the Internet on amnh.org. So, I welcome everyone from the Internet universe, as well as the universe gathered here. Tonight’s topic is: Is the Universe a Computer Simulation? Yeah.
[LAUGHTER]
TYSON: Do you want it to be a computer simulation? I mean, this topic is—we’re going to—you’ll see. We’ve got some highly thoughtful, talented, respected people to weigh in on this. I will introduce them individually, and then we will start the panel. By the way, unlike most debates you might have heard about or read about, where there’s point/counterpoint and an argument is presented and attacked, that’s not what’s going to happen here. We’re using the word debate loosely. Think of yourself as eavesdropping on scientists at a break-out room in a conference on this topic. So, we’ll all be sort of arguing with one another, and you’re listening in. That’s really what’s going on here. And you get to see how scientists think. You get to see how arguments are contested. You get to see how resolution arrives, if it arrives at all.
So, afterwards we will have a brief time for question and answer before we adjourn before 9:00 Eastern time zone— Eastern daylight time.
So, join me in welcoming my first panelist this evening. He is a professor of philosophy at New York University, where he’s also director of the Center for Mind, Brain and Consciousness, David Chalmers. David, come on out.
[APPLAUSE]
[David Chalmers walks on stage and shakes hands with Tyson.]
DAVID CHALMERS: Hey. Looking forward to this.
TYSON: Thank you. Next we have a nuclear physicist, who’s a post-doctoral research associate at MIT up in Cambridge, Massachusetts. And let’s give a warm welcome to Zohreh Davoudi. Zohreh.
[APPLAUSE]
[Zohreh Davoudi walks on stage and shakes hands with Tyson.]
TYSON: Next, we have someone who is actually no stranger to this panel. This may be his third visit to it. In part, the topic of this year was selected because he brought it up a couple of years ago. And I said, man, we could do a whole subject on that alone. Let’s give a warm welcome back to James Sylvester Gates.
[APPLAUSE]
[James Gates walks on stage and shakes hands with Tyson.]
TYSON: Another non-first timer is professor of physics up at Harvard, a specialist in nuclear particle physics. Give a warm New York welcome to one of our own, a graduate of Stuyvesant High School, Lisa Randall.
[APPLAUSE]
[Lisa Randall walks on stage and shakes hands with Tyson.]
TYSON: Did I do this out of order? No, we didn’t. Good. And last among the five—yeah, I did do it out of order. My bad. Yeah, sorry. You guys know where you need to sit. Talk among yourselves while I do this.
There’s a friend and colleague, an astrophysicist, also from MIT, who’s done some deep thinking about this very subject and has even written a book on the topic. Let’s give a warm New York welcome to Max Tegmark.
[APPLAUSE]
[Max Tegmark walks on stage and shakes hands with Tyson.]
[All panelists sit on high stools. Tyson stands and walks around the stage.]
TYSON: By the way, we are sort of lit for live streaming. And the intensity of the lights on the stage is such that two of our panelists—I think they just want to look cool, but they said they need to wear sunglasses for this event. And that’s cool. Later on I might join you. I brought my pair with me as well. If I’m feeling cool I might do just that.
So, Zohreh, I’d like to start with—no. who should I start with here? Yes, let me start with you, Zohreh.
Could you tell me why this topic interests you? Just give a couple of minutes just as an introduction here.
ZOHREH DAVOUDI: Sure. So, as Neil said, I’m a theoretical physicist. My interest is in nuclear physics. In fact, I got my PhD in 2014 from Institute of Nuclear Theory in University of Washington.
And the research I was focused on there, and at the moment, is trying to use the knowledge of the laws of nature and, in particular, strong interactions to start from a bottom-up approach and try to see what comes out in a physical system.
And that’s actually relevant to why I got interested in the simulation idea. And, in fact, by just watching the progress that researchers in this field of simulating a strong interactions have made in several past few years, we started to wonder how could we not think about the universe itself based on the laws that we’ve discovered not simulated.
So, that the way that we actually simulate the universe, it might actually give us hints that the universe itself could be a numerical simulation. And then you would start thinking, well, let’s make assumption that if that scenario is the case, and if that simulation is actually—has similarities with what we do in our research and just drawing parallels between our algorithms and techniques that we use to simulate laws of nature, and making assumption that they are similar, then what can we actually conclude about the universe as a simulation. Can we actually make predictions for the signatures that we should go after and test?
So, that’s that approach we took. And it was a fun idea and fun paper became of it with my collaborators Martin Savage and Silas Beane at the University of Washington. And that’s basically why I’m here. I’m trying to—
TYSON: So, the prospect of this being true didn’t freak you out at all?
DAVOUDI: No, I think it’s a fun idea.
TYSON: Okay. Just it’s fun for you?
DAVOUDI: Yes.
TYSON: Okay. Fine. So, Max, you’ve got a book on this, too, right? So, what’s going on with you?
MAX TEGMARK: Yeah. Well, already as a kid I was always very fascinated by these very big questions about what’s really going on with this reality. I remember actually lying in this hammock I had put up between two apple trees back in Stockholm, Sweden when I was 13, reading Isaac Asimov actually. I’m very honored to get to be here.
It really makes you think about these big, big questions. And the more I learned about later on as a physicist, the more struck I was that when you get deep down under the hood about how nature works, down to looking at all of you as just a bunch of quarks and electrons, the rules—
TYSON: And you, too. It’s not just us. Yeah. Looking at you as a quark, no, you would come under this category as well.
TEGMARK: Yes. I am a quark blob, too, I confess. But if you look at how these quarks move around, the rules are entirely mathematical as far as we can tell. And that makes me wonder, if I were a character in a computer game, who starting asking the same kind of big questions about my game world, I would also discover eventually that the rules seemed completely rigid and mathematical. I would just be discovering the computer program in which it was written. So, that kind of begs the question: How can I be sure that this mathematical reality isn’t actually some kind of game or simulation?
TYSON: So, you’ve analogized yourself to Super Mario in a—that’s who you are?
TEGMARK: I don’t know if that’s a good thing or a bad thing.
TYSON: So, Jim, I just remembered you started all of this a few years ago, in my mind at least, just triggering the idea that in your research you found things that forced you to consider the likelihood that somebody programmed us. Could you—
JAMES GATES: Well, first of all, I would disagree with you. I’m not sure somebody programmed us, but that’s—you and I had a conversation where I pointed out that in my research I had found this very strange thing. Physicists, I like to say we all belong to a company called Equations-R-Us because that’s how we make our living, is by solving equations. And so I was just going through solving equations, and I was then driven to things that Max knows about, these things called error-correcting codes. Error-correcting codes are what make browsers work. So, why were they in the equations that I was studying about quarks and leptons and supersymmetry?
And that’s what brought me to this very stark realization that I could no longer say that people like Max were crazy.
TEGMARK: Okay.
[LAUGHTER]
GATES: Or stated another way, if you study physics long enough, you, too, can become crazy.
TYSON: That’s a corollary to that idea. Yeah.
GATES: And I’m also a science fiction fan like Max, who talked about his encounter with Asimov. I was reading at age eight, as opposed to 13, sir.
TEGMARK: I hang my head in shame.
TYSON: Snap.
TEGMARK: Got off to a slow start.
GATES: I was reading at age eight a science fiction book by an author named Paul French. And some people in the audience might know that’s a pseudonym for Isaac Asimov.
TYSON: Oh.
GATES: So, science fiction drove me into science in some sense. And then now in my 65th year of life, I find out I have to make friends with Max and people like that.
TYSON: So, Lisa, I kind of brought you on the panel because I knew you—I mean, you’re a rationalist in all this. And so I was expecting—I don’t know what to expect. I just needed to anchor this in somebody who I knew was not going there. So, where—
LISA RANDALL: Yeah. So, actually—well, I can’t say I decided to be on the panel because I think I said what date is it, and they were like, “Thank you for agreeing to be on the panel.” But I have to say I’m curious not so much about the question of whether we’re a simulation because I think it’s only interesting insofar as there are ways to test it. And we can come back to that, I think, very much in terms of how the laws of physics operate and whether we can actually distinguish that. But I actually am very interested in why is so many people think it’s an interesting question. Like why is the audience here? Why is this panel here? Because really to first approximation we can’t really distinguish it.
So, I think the interesting question is: Why do we feel compelled to want this to be true, or even think this could be true? And how do the laws of physics operate? And are there really ways that we could eventually test whether there is something that distinguishes just a true universe? But I have to just say if the inference is simulation, I don’t understand why it gave me a cold today.
TYSON: Okay.
RANDALL: So, my voice might go. But I also think sometimes some of the ridiculous things in the universe and think, really, why would that be part of the simulation? And I realized that if I was doing a simulation, I would definitely put those things in. So, there you go.
TYSON: Okay. Well, thank you for that. Now, we couldn’t have a panel without a philosopher. David, we needed some philosophical—
DAVID CHALMERS: I know how you love philosophers, Neil.
[LAUGHTER]
TYSON: I’m on record for some comments about philosophers that got him a little ticked off. But, anyhow. So, David, what do you—philosophers have been at this for a while, yourself included. So, how do you see all of this happening or fitting in to the worldview?
CHALMERS: Well, philosophers like to ask the big questions about the world; the foundational questions. And this is one of them. Actually, I blame Isaac Asimov for all this, at least in my case. I got into thinking about these big questions when I was a kid. I read just about everything that Asimov was writing. Not just the science fiction, but the science fact, the history, the detective novels. I read multiple volumes of his autobiography. But throughout Asimov’s work, this was a guy that was just interested in the big questions about the nature of reality at all levels. And that, ultimately, drove me to think about questions about consciousness and the mind, which I could approach as a philosopher because philosophy allows you to step back and say what is the science here telling us.
But this question about the simulation corresponds to another of the great questions of philosophy, which is basically how do we know anything about the external world at all [unintelligible] said how do you know you’re not being fooled by an evil genius into having an impression of this world around us? Even though none of it really exists.
Well, the contemporary version of that question is: How do you know you’re not in a simulation like The Matrix? In which case, allegedly, none of this really exists. And, to me, that question is just extremely interesting because it seems nothing we could know could rule out the hypothesis that we’re in a simulation.
But you also want to think about what follows. Some people think if we’re in a simulation, then none of this is real. I think if you adopt the kind of perspective which, say, Max was suggesting a second ago, where the universe is all mathematical or informational, this allows us to reorient our attitude to this question and say, okay, maybe we’re in a simulation. But if we are, all this is perfectly real because all the information is there in the simulation.
All the math is there. All the structure is there in the simulation. So, I’d say, well, maybe we’re in a simulation. Maybe we’re not. But if we are, hey, it’s not so bad.
TYSON: If I do this, you feel that.
CHALMERS: Yeah.
TYSON: Okay. So, that’s real. That was a real punch. Yeah. So, Zohreh, let me ask you, I see you coming to this almost from the most pragmatic side. You’ve done experiments with your colleagues. Or you’ve had hypotheses with your colleagues. Could you just detail for me where you landed in one of those papers that you guys published?
DAVOUDI: Sure. So, what we did is not actually doing the experiment. We proposed that experiments could go and look for the signs of possible underlying simulation for the universe. And the reason we thought about this, as I said, is because we’ve been simulating strong interactions, which means that instead of just looking at the larger structures, we’d start from the underlying degrees of freedom of our theory, the quark, gluons, and that we understand. And there are very simple laws governing the interactions among these particles.
However, when you think about all these complex systems of atomic nuclei and larger systems in the universe, the ordinary matter in our universe, it all emerges from those simple, fundamental building blocks and these interactions.
So, we’ve been trying to just input those simple mathematical structure with a few degrees of freedom, these quarks and gluons, and then see how these, for example, atomic nuclei emerge from these simulations.
TYSON: So, you’re building the universe from the ground up?
DAVOUDI: Exactly. But what are the limitations? We don’t have infinite computational resources. We have very large super computers in the national labs, for example, that we can compute these interactions basically and build up these systems.
However, we are still limited. And the reason is that if you’re interested in simulating the universe, and you don’t know what the size is—it could be finite or infinite. However, we are limited to a finite size.
On the other hand, if you think about even a finite side, there are infinite numbers of points on these in this finite size that you have to simulate to get the physics right. However, we are not capable of inputting infinite number of information in our computers.
Also, we want the simulations to be quantum, which means that there is not just one single path of evolution from one point to the other. There are infinite number of paths. Some are more important than others. And, therefore, there’s another type of infinities that we have to implement in our simulations to get the answer right.
TYSON: Yeah, but just because you can’t—we can’t do it because we’re limited, why should that mean the whole universe is limited?
DAVOUDI: So, wait. So, this is the point.
TYSON: I’ll wait. I got time.
DAVOUDI: All right. So, we can do it, and then you—based on assumption that if there is an underlying simulation for the universe that has this problem, that has the problem of finite computational resources—just as we do—then what happens?
Then the laws of nature, the quantum mechanics and whatever interactions have been going on, has to be put on a finite set of space-time points in a finite volume, and then just a finite number of quantum mechanical paths to a process can be evaluated.
So, these are the assumptions. So, if the simulator of the universe, in whatever form it is, is just finite computational resource and not infinite, then it’s limited to simulate the universe in this kind of limited scenario, just as we do. And then by making that assumption, and then going back and look at our simulation and see what kind of signatures we see in the observables we calculate, that could tell us that we started from a non-continuum space-time. Then apply it to an underlying simulation of the universe and make the same assumption, then what would you see?
And that’s basically what we look for, and list a few observables in our universe that might lead to actually constrain this scenario under this assumption. And one of which is looking at the spectrum of cosmic rays. Because what happens if these very high energy cosmic rays that approach the earth, they are actually traveling in a discrete space-time, as opposed to a continuum. Then their equations that basically special relativity that would describe the relation between the energy and momentum of this particle is modified.
And then you would ask what would that modification mean in terms of the observation we make in our observatories, for example, spectrum and distribution of these cosmic rays. And if we see something that would be hint, that would be consistent with the scenario of a limited computational resources of the universe. And then you might think about other signatures and maybe taking this scenario more seriously and think about [unintelligible]—
TYSON: So, cosmic rays, it would be your pathway to the limits of what has ever been measured.
DAVOUDI: Exactly.
TYSON: And then seeing at that limit you’re probing the limits of the programmer of the universe.
DAVOUDI: Right. Because these cosmic rays are the most energetic particles that we’ve ever been able to observe. We can’t even produce them in laboratories. These are very high energy cosmic rays.
TYSON: They’re higher than anything we produced in our particle accelerators.
DAVOUDI: Exactly.
TYSON: Yeah.
DAVOUDI: Yes. By orders of magnitude. And, therefore, because these are very energetic, they can actually probe the fabric of space-time. This is our way of probing if the universe—if the underlying space-time is discretized or just a continuum.
TYSON: So, Max, like I said, you’ve written a book on this. Yet, you told me offline that you have an argument that would argue that—
TEGMARK: That maybe we’re not simulated after all?
TYSON: Yeah. Maybe we’re not a simulation after all. So, where does that land?
TEGMARK: Yeah. So, before giving a counter argument, let me give the pro argument. Of course—
TYSON: So, you can give arguments in both directions here?
TEGMARK: It’s fun to argue with yourself.
TYSON: Okay.
TEGMARK: Of course, we all—as David mentioned—have seen the argument, the idea, of us being simulated in The Matrix and in science fiction going even far beyond that. But the guy who really started foreseeing scientists to take this a bit more seriously, and gave this idea a bit more scientific street cred, I think, is Nick Bostrom, my fellow Swede—Nick Bostrum—who published this very dry academic article that’s pointing out that—
TYSON: He’s a philosopher?
TEGMARK: Indeed, indeed. And he pointed out that it seems like the laws of physics allow us to build amazingly powerful computers way beyond what we have now; solar system-sized things, which could simulate minds that would feel just like us. And then he went on to say it seems overwhelmingly likely, if you don’t wipe out here on earth, that in the future the vast majority of all computations and all minds will be inside of such a computer. And, therefore, he said if almost all minds are simulated, we’re probably simulated. So, that’s the pro argument.
Now, it sounds good, but—
TYSON: So, just to clarify, so what you’re saying is if simulating universes becomes a pastime among those who have access to high powerful—to highly powerful computers, and we are in a universe, we’re probably in a simulated universe, even if one of those universes is actually real.
TEGMARK: Right. That’s basically—
TYSON: Is that a fair—
TEGMARK: That’s a fair summary, yeah. And if you’re not sure at the end of the night whether you’re actually simulated or not, my advice to you is go out there and live really interesting lives and do unexpected things so the simulators don’t get bored and shut you down.
[LAUGHTER]
TYSON: Is that the cause of death? Okay.
TEGMARK: But now in terms of the counterargument, if you just take Nick seriously—
TYSON: That’s the cause of death.
TEGMARK: There’s something fishy here. Because suppose you buy into this and you’re like, okay, I’m sold on Nick’s argument. We are simulated. Let’s talk then about our simulated universe. We’re measuring the laws of physics here in the simulated world. And we find that in the simulated world we can build all these supercomputers in the future, and there’ll be all these simulated minds and so on. And we can make the same argument all over again and convince ourselves that actually we’re doubly simulated. And then we’re a simulation in the simulation, and then you can repeat the argument again and say, well, okay, we’re in a simulation in a simulation.
But in the future, there’re going to be all these simulated, simulated computers and they’re going to have all these minds. So, we’re actually triply simulated. No, we’re quadruple simulated, and it goes on and on all night.
TYSON: So, the turtles all the way down.
TEGMARK: Turtles all the way down. And at this point, I get this sinking feeling that there’s something rotten at the core of this argument.
TYSON: Okay.
CHALMERS: The answer is we’re at level 42.
TYSON: Good answer.
GATES: No, no, 137.
TYSON: One-thirty-seven. That’s the fine structure constant.
GATES: Of course.
TEGMARK: And I think where the problem lies is that when you make this argument about what kind of minds are really the most common, the most simulated and non-simulated, it assumes to answer that you have to know what the actual laws of physics are.
But if you start making these other arguments, we have no clue as to what the laws of physics are. It doesn’t matter what the laws here in our simulation—if it is one—are. We need to know what the real laws of physics are in the basement universe that’s the foundation. And, if so, we don’t really have access to that.
So, that’s the philosophical nitpick, which seems to be swept under the rug here.
TYSON: Jim, where—
GATES: Where am I?
TYSON: Yeah.
GATES: Well, first of all, I have a finger. And I look at it, and it seems to be real. And so my point of view is very conservative. It was Carl Sagan who once said that, “Extraordinary statements,” and I’m paraphrasing—
TYSON: Claims, yeah.
GATES: Right. “Extraordinary claims require extraordinary evidence.”
Now, Zohreh has told us about a kind of evidence. And that’s the kind of evidence that would convince me as a physicist. But what I do is sort of a mathematical model of physics. And in our previous encounter here on this stage, I had a chance to tell you about these error-correcting codes, which are very specific kind of digital data. It’s not just general digital data. It’s a very specific kind that seem extraordinarily unlikely.
And I have to tell you that one of the reasons I enjoy talking to audiences like this is they get us experts out of our comfort zone. And so one of the first non-physicists that I talked to, or that I read reflected on my comment, said effectively—this is not exact words, but effectively he said if the simulation hypothesis is valid, then we open the door to eternal life and resurrection and things that formerly have been discussed in the realm of religion. And the reason is really quite simply. Because if you think about a computer—if we are a simulation, then we’re like programs in a computer, as long as I’m a computer that’s not damaged, I can always rerun the program.
So, if you really believe that we are in a simulation, and there’s some structure that runs that simulation, unless something damages that structure, then we can be repurposed. And so it starts to break down a very funny barrier between what people often think as the conflict between science and the conflict between faith.
TYSON: So, what you’re saying is that if we are simulated, that means there’s a code that’s doing it, and that code was started at some point. And in principle, it could just be rebooted, and then all of this would happen exactly the way it happened before because it’s running the same computer program. In principle.
GATES: If one accepts the simulation hypothesis as an accurate description of nature—
DAVOUDI: I would say that’s a useless exercise. What would be more interesting is to actually—
TYSON: The word was useless, Jim, in case you missed that.
[LAUGHTER]
TYSON: Okay, you heard that. Okay. Emphasis on useless exercise. Go, Zohreh. Go.
DAVOUDI: Trying to repeat what you’ve already done with huge computation resources is useless. What is more interesting is to go and change the parameters of the simulation—the input parameters. Just put the same laws of nature, and then just change a little bit the value of the parameters—the very fundamental parameters of our universe. And then let it run and see what happens. It’s actually very interesting idea—
TYSON: It’s a fun thing to do, as a scientist.
GATES: But in changing those parameters you might cancel out my existence, in which case I don’t think that’s very useful.
[LAUGHTER]
TYSON: The universe without Jim. So, Lisa, isn’t this some of the foundation—couldn’t we account for a multi-verse in this very way? That multiple-verse is multiple universes as I understand them will have slightly different laws of physics. Maybe they are themselves the experimenter’s search.
RANDALL: Okay. So, let’s slow down a bit here. So, first of all, I actually want to address some of the things that have come up already. One of the questions is probability; Bostrum’s argument or whatever, that we’re likely to be in a simulation. I mean, part of the problem is that probabilities have to have a well-defined meaning, or are only useful when they have a well-defined meaning.
So, among all possible scenarios we can actually say which one is more or less likely. When we run into infinities, when we run into—it stops making sense. I mean, I could say really by probability I’m very likely to be Chinese because there’s a lot more Chinese than Americans. But I’m clearly not Chinese. So, probabilities are tricky, and you have to be careful what you mean when you’re saying them.
Another thing is I actually find the egotism of thinking that if there was simulators around that they’d come up with us kind of audacious and ridiculous. I mean, I think it’s a very self-centeredness to this whole thing that kind of I find hilarious.
[LAUGHTER]
RANDALL: But in terms of feedback—in terms of error-correcting code, I think it’s very likely that there were going to be feedback mechanisms in whatever universe survives because if there aren’t, I mean, there’s always going to be mistakes. And if mistakes can propagate and just cut things off, those universes don’t survive. So, there have to be—I mean, for any universe, simulated or non-simulated, there has to be error correction. So, that has to be part of it.
TYSON: Right. That assumes that the programmer makes the same kind of programming—is susceptible to programming errors and programming bugs that we are.
RANDALL: It’s not even intentional. It could be just that the computer itself is subject to error. I mean, it’s only firing things somewhat random—I mean, ultimately, there’s uncertainty in everything. Nothing is created perfectly.
TYSON: Quantum uncertainty.
RANDALL: So, [unintelligible].
GATES: Can I jump in here?
TYSON: What?
GATES: Because she’s raised—in fact, I think an incredible point about this.
RANDALL: As long as you come back to me afterwards.
GATES: Maybe I take a few times? I’ll [unintelligible] minutes back later.
TYSON: Yes, okay.
GATES: This point about error correction is something that when people have—general public has looked at my work, they say, “Oh, you must believe in simulations.” And I’ve said, no, actually I don’t. And the reason is because precisely the point the Lisa points out.
If you look in all of nature and ask are there any other places in nature—not in engineering, not in computers, not in the things that we build, but in nature herself, is there a discussion in science about error-correcting codes? It turns out there’s one place and one place only that I have been able to identify. That’s in evolution and genetics. And there’s been a discussion—
RANDALL: Or any biological system.
GATES: Right. Or any biological—right. And it’s not that we think life is some kind of programmed simulation. It’s because the universe itself, as Lisa had said, has to have feedback mechanisms that basically sustain a structure that propagates faithfully forward in time. And I think that’s in fact the most critical point. And you have your time now.
RANDALL: Thank you. And anyone who wants to take my time to agree with me—
[LAUGHTER AND TALKOVER]
RANDALL: But as far as the multi-verse theory goes, so we have to be careful by what we mean by that. I mean, at some underlying level we still think it’s physics in action. Now, what might change in different universes, we might actually have different forces. We might actually have different strengths of interactions; the kind of thing that gets simulated. I mean, we simulate strong interactions the way that were described.
TYSON: Just to be clear, strong interactions are the forces that bind atomic nuclei.
RANDALL: So, protons.
TYSON: Yeah, protons that are the same charge that are sitting right next to one another in a nucleus. And how’s that even possible when we were taught that like charges repel?
So, there’s got to be a really strong force down there holding it together. And there is a really strong force. It’s called the strong force. Okay, so go on.
RANDALL: Which is strong.
TYSON: Yeah. Okay. Just to be clear.
RANDALL: So, and there can be different possibilities for what these parameters can be. It’s still underlying you still believe that there’s the laws of physics that are operating.
So, the question—I mean, so it’s not a simulation. It’s just—I mean, it’s in principle possible that there are universes we don’t communicate with that are so far away we’ll never send a signal, they’ll never send a signal. So, for all intents and purposes, there just are different universes. That doesn’t mean they’re simulated. It just means they’re different from ours and they can have different properties.
To really distinguish a simulation, you really do have to see just our whole notion of the laws of physics breaking down, or some of the fundamental underlying properties. So, it would be extremely interesting to look for the kind of violations of [unintelligible] that were discussed earlier, or things like quantum entanglement no longer hold it. Not because of interaction of the environment, but just the computer just couldn’t keep track of stuff. I mean, that’s stuff that gets so—
I mean, a lot of the simulation idea—I mean, to simulate the universe, you need the computational power of the universe. So, all of the simulations are based on the idea that there are some approximations that we don’t see, but you have to be able to hide them. So, what we’re really looking for is the breakdown of the assumption that those approximation s are valid.
TYSON: But, David, what do your philosophical circles say about proposing an experiment that might falsify these ideas?
CHALMERS: Look, I don’t think you’re going to get conclusive experimental proof that we’re—we’re certainly not going to get conclusive experimental proof that you’re not in a simulation. I suppose we could get some kind of various—
TYSON: Well, why not? You just declared something. Why can’t a clever person come along and—
CHALMERS: Because any evidence that we could ever get could be simulated. That’s basically the reason. Sorry. Maybe—
TYSON: So, if I find evidence that we’re not simulated, the great simulator—
CHALMERS: They could have just planted that for you.
TYSON: —put that in.
CHALMERS: Yeah. They’re one step ahead. However—
TYSON: We’re done. We’re done here.
CHALMERS: Maybe we—we probably could get pretty strong evidence that we are simulated. If someone wrote up in the sky, “Sorry, guys”—the stars suddenly rearrange themselves into, “Sorry, guys, it’s all a giant simulation.” And then they took over the Internet and—
TYSON: Except it would be in Chinese to get the most number of people to read it.
CHALMERS: Then we’d probably have a pretty good reason to think we’re in a simulation. Either that or the weirdest non-simulated universe that anyone ever imagined. So, for a philosopher anyway, it’s not fundamentally a matter of experimental proof. It’s cool. I really like Zohreh’s experimental evidence that we’re in a simulation. But I think around here it’s really important to make a distinction that there’s a hypothesis that we’re in a simulation. There’s a hypothesis that the universe is computational.
Those are closely related. If we’re in a simulation, the universe is fundamentally computational. But it’s not true that this universe is fundamentally computational we’re necessarily in a simulation. Because the simulation hypothesis is a combination of two things.
TYSON: That’s an official thing, the simulation hypothesis.
CHALMERS: Yeah. The simulation hypothesis says we’re in a computer simulation. A computer simulation’s a computation that was created by someone for a purpose. So, basically the simulation hypothesis is that computation hypothesis, plus something else about someone who created it. And around here is where you might be able to get a little theological and say, okay, well, it’s a naturalistic version of the god hypothesis. But, anyway, my worry about Zohreh’s stuff, which is really cool, it’s really evidence for the much weaker hypothesis that the universe is some form of discrete computation and is completely neutral on the question of whether this is actually a simulation in the sense of something that was created—
TYSON: With intent.
CHALMERS: —by a simulator.
TYSON: So, Max, do you mind if I call you Mario from now on? Because if you’re Mario in the computer game—
TEGMARK: Starts with M-A, so [unintelligible 38:46] for the two letters, yeah.
TYSON: I imagine Mario—someone coming into a Mario game and calculating how high he jumps and how fast he runs and coming up with the laws of physics of the game, and possibly then questioning why is it that and not something else perhaps. And so, fine, but is there—why would that allow someone in the game to have any understanding of what’s outside the game?
TEGMARK: Yeah, that’s a really deep and good question. Mario might—if Mario can ever—even if he figures out exactly the rules of his world—
TYSON: Then he just figures out the rules.
TEGMARK: —he won’t even know if he’s running on a Mac or a Windows box or a Linux box because all he has access to is this higher level of this sort of emergent reality. And we might, at some level, be stuck in that situation in physics also. It’s quite fascinating to think that so much of what we’ve figured out, for example, about how a glass of water works with waves and vortexes and things, we figured out already without having a clue about the substrate. We didn’t even know there were atoms. But the same kind of questions that you’re asking, which I think are awesome, the kind of questions where you ask suppose this is actually somehow simulated, suppose the simulators cutting corners, how would that show up?
Actually, it has been incredibly useful in the past. If you imagine going back 200 years and trying to simulate this water as an infinitely—a continuous liquid where there’s a pressure and a density that has to be defined with infinitely many decimal places and infinite points, that sounds horrible to simulate. So, maybe whoever did this cut corners. Maybe there’s a smallest kind of chunk of object—let’s call it atom or something—you can figure out then what are the departures from this simplified continuous physics that I’m guilty of teaching my undergrads at MIT about this morning?
And you would figure out a way there’s this one little thing, which is different.
TYSON: He trained down a few hours ago from Cambridge.
TEGMARK: Yeah.
TYSON: Thank you for coming and for—
TEGMARK: Brownian motion that things should jiggle around in a weird way. And Einstein found that, got the Nobel Prize for it importantly. And I think that the sort of thing you’re doing is awesome.
Look for corner-cutting evidence. I suspect that whether we’re simulated or not there are a lot of things that are wrong about what we assume today. I am very skeptical that we really have a continuous space that can be stretched infinitely many times. It seems like some sort of simplification that we came up because it was easier to do the math.
DAVOUDI: But do you ever ask why should that be the case? Why do we need a discretized universe? I mean, if you put away the simulation hypothesis or a computational hypothesis, why should we even think about a discretized universe? Why not continuum? It’s [unintelligible].
TYSON: So, this is an important—
TEGMARK: Yeah.
TYSON: I don’t want to call it a problem in physics, but a reality of physics that our macroscopic world looks continuous to us. And that has a certain simplicity of modeling. And then as you get smaller and smaller and smaller, it’s no longer continuous and it’s discrete, which may be easier to calculate than being able to be divisible all the way down to an infinitesimally small bit. Because now you need that much bigger computer to do it. By the way, we have—
RANDALL: So, you know something that none of us actually know. This is actually a real question, whether space is discrete at really small scale.
TYSON: Well, we run into this problem when we do flyovers in the Hayden Planetarium. We have a data set for a planetary surface—let’s say Mars—and you had a given distance. And from that distance you can see Olympus Mons, the biggest mountain around, and Valles Marineris, and you say, fine, now I want to get closer.
Well, to get closer, and have more information come to you, you have to swap in a higher resolution map. And we try to do that continuously, so you don’t realize that. So, you keep doing this, and then you reach a point where we don’t have more resolution to give you. So, we actually hold you back, so you don’t go closer. But if you did, all of a sudden you see these discretized pixels of the Martian surface.
And that’s basically because we don’t have the data. We’re not there. It doesn’t exist for us.
CHALMERS: So, anyone’s who’s used one of these virtual reality devices, like the Oculus Rift, knows there’s something called the screen door effect. It’s like you can—if you look closely enough you can see the pixels, so it’s not a perfect simulation. So, I guess really what Zohreh is doing is saying, well, we can get empirical evidence for a screen door effect in real physics.
DAVOUDI: Yeah, I think it’s actually a deeper question than that. It’s not about not having enough data to resolve those distances, but to some extent that’s true. But the problems is something that even bothered Feynman a lot that why do you need infinite numbers of degrees of freedom, or infinite amount of information, to describe a very tiny chunk of the space-time? That just doesn’t make sense.
You can pretty well describe the physics without actually needing that infinite amount of information.
TYSON: What I meant to add is that when we’re zoomed down to Mars, it’s not only that we don’t have the data, even if we did have the data, you would need that much bigger disk space to have it ready and loaded to be able to go from the bird’s eye view down to any kind of small—I mean, we rapidly run out of capacity to calculate.
TEGMARK: And that’s a great controversy that even mathematicians have been really arguing passionately about for over 100 years. Gauss, one of the greatest mathematicians ever, said—or Kronecker actually said God had created the integers and everything else was just the work of man. All this continuous real numbers with decimal places and stuff.
I mean, frankly, as a physicist it feels kind of hubristic to say that you need an infinite amount of information to figure out the height of my wine glass or anything. Nature seems perfectly about to figure out what’s—
TYSON: There’s water in that glass, by the way.
TEGMARK: Yeah, what to do. And we have this toy model that you need an infinite amount of information to do things. I think you’re on to something very deep [unintelligible 44:56] and that nature actually—infinity is just something we made up for convenience.
And as we dig deeper, we’re going to find that maybe even space and time itself is at some level digital.
RANDALL: So, can I just say something by way of clarification? Which is just in physics we don’t actually prove any theory. We can rule out theories.
So, we can rule out a lot of alternative theories, but in any case you can always have the possibility that you can dig deeper and find that whatever theory you thought was the most fundamental has some underlying structure.
And so that’s why all the physics we’ve done works. That’s why we really don’t need to have an infinite amount of information at any time because we don’t have access to an infinite amount of information. And we can’t even ask the question or tell whether or not there’s this underlying infinite amount of information.
So, it’s not just we can’t just ask the question whether the universe is a simulation. We can’t ask if any physical theory is absolutely correct. We’ll never know the answer to that. All we can know is that we’ve tested it up to a certain level, at a certain level of precision, over a certain range.
And so these questions all [unintelligible], and that’s why I can describe this glass of water without knowing about atoms, because I didn’t have—wasn’t doing an experiment where the effects of the atoms became manifest. And the same might be true of the universe as a whole.
So, we can have in the back of our mind there may or may not be an infinite number of degrees of freedom. But that’s not what we’re actually testing.
TEGMARK: Let’s disagree on one thing, though. I think there’s one fantastic example where we can tell it makes a huge difference. I think the biggest embarrassment we have arguably in fundamental physics and cosmology right now is this fact that inflation, if it goes on forever, makes this multi-verse, and then we can’t calculate probabilities, like you so eloquently said in the beginning. That comes exactly from the infinity assumption; the idea that you can take a piece of space and just keep stretching it into twice the size forever. So, I think you should question that.
RANDALL: Well, it doesn’t have to be infinite. It could just be a large number. It could be 10 to the 500. I mean, it doesn’t really matter if we say it’s infinite. Why don’t we just say it’s a lot?
TEGMARK: But you can calculate probabilities as long as it never gets infinite. It’s exactly infinity that [unintelligible]—
TYSON: So, he’s cool with 10 to the 500, is what he’s saying, which seems like a really big number.
RANDALL: I know.
TYSON: That like equals infinity to me, I think.
RANDALL: But that’s exactly the point. That’s exactly the point.
TYSON: Jim, is there any functional difference at all between admitting that we live in a computer simulation and saying that’s basically a secular god? What’s the difference?
GATES: Well, first of all, I’ve decided my name should be Morpheus, not Jim.
TYSON: Okay. Well, let me—
TEGMARK: I’m Mario. Nice to meet you, Morpheus.
TYSON: Morpheus.
GATES: Exactly.
TYSON: Yes. You have to see the movie The Matrix and play video games to follow this conversation at this moment. Morpheus.
GATES: But as I said, for non-scientists—because I’m going to make this partition. I think for non-scientists, an acceptance of the simulation hypothesis as an accurate view of our universe is equivalent, I believe, to the notion of a deity. I don’t understand how, for a non-scientist, you can make that distinction. For a scientist, however, we are rather secular.
The definition of science is actually a secular definition. And, in fact, it’s the definition that comes to us from Galileo. Einstein quotes Galileo as being the father of all science because Galileo—and these are Einstein’s words—drums into us that contemplation alone, without observation of nature, is totally useless in trying to come up with an accurate view of nature. So, it’s that ability of us—our human ability to observe the universe that actually defines science. So, if you can’t give me something that I can observe, I don’t know how to do science.
TYSON: Okay. So, what you’re saying is that if in fact there is a programmer who would be philosophically equivalent to a Creator, and you can’t observe them, they’re just outside the realm of science.
GATES: I think that’s the definition.
TYSON: David, do you have to be defined by that?
CHALMERS: Well, I think there’s a theological reading, if you like, to the simulation hypothesis. It says all this was created, but what’s interesting is at the same time it can be seen as a kind of a naturalistic theology. A naturalistic hypothesis—from the point of view—
TYSON: Is that the first time the phrase has ever been uttered? A naturalistic theology.
CHALMERS: I think it’s out there already.
TYSON: Oh, it’s out there. Okay. All right.
CHALMERS: Simulation theology [unintelligible]. Simulation theology is the coolest kind of naturalistic theology, from the point of view of the—
TYSON: Actually, there’s a book in 1750—or who was it?
CHALMERS: Yeah, David Hume was into naturalism.
TYSON: No, there was—who was the fellow who wrote the book Natural Theology? There was a book with that very title.
CHALMERS: Yeah.
TYSON: But not natural simulation or simulated theology.
CHALMERS: If you think about is from the point of view of the simulated—I mean, we in this universe can create simulated worlds, and there’s nothing remotely spooky about that. People are already doing it with virtual reality and the Sims and Second Life. And whatever this is is just a far more sophisticated version of that.
So, we just need to move that picture to the next universe up and say, hey, maybe that’s what’s happening to us. So, we got a creator, but our creator isn’t especially spooky.
It’s just some teenage hacker in the next universe up whose mom’s calling him in to dinner.
TYSON: Working in the basement, yeah.
CHALMERS: So, I think you could be led to at least entertain this idea by perfectly naturalistic ideas as, say, Nick Bostrum was and say, okay, maybe this is the kind of theology which even someone who’s got no sympathy for spooks and gods and ghosts, needs to object to.
TYSON: So, that’s an interesting point because we don’t think of ourselves as deities when we program Mario, even though we have all power over how high Mario jumps. Because that’s a line in the code. So, you’re right. You just take it up a few notches. There’s no reason to presume they’re all powerful other than just they fully control everything we do, say and think.
CHALMERS: Could be they’re all powerful. I got into this from watching my five-year-old nephew playing with one version of the Sims or Sim Life or something. He’d make a whole town. He’d build up the buildings, and you got the trees and the jungles and the creatures. And then he’d say now comes the good part, and he’s send down fires and floods and such. I was like, finally, I understand the God of the Old Testament.
TYSON: Because it is true in our world we have fires and floods. I played one of those Sims—Sim City because I’m a city kid. And—the early, early low-res simulation. And there’s a feature, you build up the—you need money. You’re mayor of a city, and you construct buildings and you need the schools and the fire departments.
And then every now and then Godzilla stomps through your city and you say that’s not real. I’m trying to be real. But then it’s kind of real in the sense that some major disaster can—you will confront like Hurricane Sandy or 9/11. Now, you’ve got to redistribute resources.
So, I look at our real world, and these things actually do happen. So, are they just trying to mess with us? Is that—
CHALMERS: The way I think about—I mean, who knows if there’s actually a simulator who’s actually doing any of this. But if you do take the simulation hypothesis seriously, it’s got a couple of elements of a traditional god. This person could be all knowing about our universe, could be all powerful. The one thing which is probably missing is wisdom and benevolence. If there’s a simulator, I refuse to worship you. You may be out there, but you have established yourself as being worthy of worship. I refuse to [unintelligible]—
TYSON: Right. Because they’re all powerful and all knowing, but not all good.
CHALMERS: There’s no reason to think they’re all good.
TEGMARK: Cut him some slack. He’s only five years old.
[LAUGHTER AND TALKOVER]
CHALMERS: You’re going to be maturing one of these days.
TYSON: Zohreh?
DAVOUDI: Yeah. So, I think there is a big danger in trying to compare our idea of simulation with what comes with computer games, whether you’re talking—at least in my point of view and I think a physicist’s point of view.
What’s called the simulation is you just input the laws of physics, and nature and universe emerges. You don’t actually try to make it look like it’s something going on. You don’t try to—the same as with computer games. You don’t interfere with what you’ve created. You just input something that is very fundamental and just let it go, just as our universe.
TEGMARK: Like deism?
DAVOUDI: Yeah.
TYSON: In other words, you set the laws into motion and let the universe unfold.
DAVOUDI: Exactly.
TYSON: However those laws prescribe.
DAVOUDI: Because a priority—you don’t know what happens because the universe is complex. The laws of physics are simple, but you don’t know what kind of complexities you should expect. And then you just get it wrong and things emerge, and we just watch.
TYSON: But, Lisa, in the search for the Theory of Everything, isn’t that got a little bit of this in it? Once you find the Theory of Everything—and you’ve been on two of our Theory of Everything panels here—you’re going to find out the one equation that the five-year-old working in the garage wrote down that made our entire universe.
RANDALL: Well, you might recall, since I’ve done this a couple times, that the Theory of Everything, I think, is very badly named for a lot of these reasons. Because even with the equations, as was pointed out earlier, you could start your system in very different ways. You can have different conditions. And there’s a lot that we don’t understand.
I mean, even if I understood quantum gravity at a fundamental level and could derive all the equations, that’s still not going to help me predict waves at a practical level. I mean, the computer simulation will never be that detailed, in my opinion. It’s much better to go to different levels and figure out what’s going on at what I would call an effective theory approach. So, even with the fundamental equations—
Now, I mean, clearly if you had infinite computing power, then you would just be literally mimicking the universe. And possibly you could do that. But short of that, you’re going to have to find these approximations, these descriptions that are sort of somewhat in between. They’re still science.
They’re not something I’m just making. There’s still equations that work, and they ultimately are attributable to whatever is that fundamental equation. But that doesn’t mean it’s fundamentally how we’re computing it. It doesn’t mean it’s fundamentally how it’s working.
TYSON: But, Zohreh, you started this whole discussion by describing—trying to obtain an understanding of the basic forces of nature and the particles and build up from there. But isn’t there surely a gap between what you know drives the behavior of individual particles and what might be emergent features in a macroscopic system. Isn’t that true with the gas laws?
We learn gas laws in the first week of chemistry, but I don’t know that you can get the macroscopic gas laws by knowing every single particle at every single instant. I don’t know that they’re fully reducible to that. So, can you admit the possibility that there are gaps and that there’s emergent phenomena that—so, starting at the very basic level won’t get you there? Is that possible?
DAVOUDI: I do admit to that, and it is in fact—
TYSON: Okay, good. Thank you. You admit to it. No, go ahead.
DAVOUDI: No, this is indeed a field of research now, for example, in nuclear physics we know that these microscopic features about particles and building blocks of that would contribute in strong interactions, but we don’t know exactly how to get these complex system of nuclei.
And we have very good microscopy and [unintelligible] models that describe all these larger-scale phenomena, but we still don’t know how to get them from this phenomena. So, that’s what, as physicists have to—
RANDALL: In principle, if you could do it—I mean, if you had infinite computing power.
DAVOUDI: Yeah.
RANDALL: In principle, you could actually see a system that exhibited the gas laws. The question is whether we as scientists would call them—deriving the gas laws. It wouldn’t be a very useful description.
It would mean that we’d have to have these enormous computations every time to do it, rather than solve an equation that, as you know [unintelligible]—
TYSON: Oh, so I never heard that before. You’re assuming that if in fact we could compute the behavior of every single particle in a gas, that out of that would emerge the macroscopic gas laws.
RANDALL: Well, it would behave according to the gas laws. That doesn’t mean that you [unintelligible] know what those gas laws are.
TYSON: Okay. That’s confident. So, what you’re saying is it’s not emergent. Because I’m intrigued which of you mentioned the water—
RANDALL: Well, emergent means that it emerges from the fundamental laws.
TYSON: But because we understood, to a very high degree, fluid dynamics long before we knew that fluids were made of atoms.
RANDALL: Right.
TYSON: And I don’t know how much the public knows that atoms are—though, the idea is old, evidence that atoms are real is relatively recent. And even as recently as the year 1900, it was still kind of not sure.
And it wasn’t really until Einstein and Brownian motion in 1905 where there’s really good evidence that atoms were real things. Yet, we had full understanding of fluid dynamics in any way that mattered for us.
RANDALL: Right. But we also now can derive fluid dynamics from the atomic description, in certain cases. Not all of fluid dynamics, but some of the properties of condensed matter physics we can derive by that.
TYSON: Okay, I’m glad to hear that. So, we’re still talking about reducible things.
TEGMARK: They’re two separate things, though. We mustn’t conflate. On one hand, I think in principle one can derive all these higher level things, I think, even ultimately even consciousness like David Chalmers is working on, from starting out with a quark as the [unintelligible]—
TYSON: You’re going to bring consciousness into this?
TEGMARK: But in practice, on the other hand, whether we humans are smart enough to figure it out. That’s a whole different story. And I think that’s—I’m guessing that’s what you were getting at there. You weren’t saying that there’s some mysterious epistemological gap that we can’t—
DAVOUDI: Oh, no, no. That’s not what I meant.
TEGMARK: But that we might be able to understand.
DAVOUDI: We haven’t yet have the resources and probably enough tools and understanding to fill that gap. But the fundamental equations are there. It’s just a matter of when we actually get there.
TYSON: So, I’m curious—this brings me to a point that we did not discuss earlier in the notes that we shared. You can know everything you can about cell biology, about how life works.
And it’s not obvious to me that by just studying a single life form that you can derive evolution by natural selection. That that’s an emergent phenomenon given the system. So, if it’s emergent, then no one actually programmed it in to do that. That’s just something that resulted.
RANDALL: Right. So, the way I would describe it is I would say that the fundamental—whatever’s fundamentally there—that substrate—is essential to whatever happened, but is not necessarily essential to your description of what happened. And so the laws are following from this, but it’s not giving an explanation.
So, I can note that there’s atoms, but it doesn’t help me predict what will happen when I throw a ball. I mean, in principle I could probably figure it out based on that; put it all together, but it won’t help me. It’s so inefficient.
So, it’s much better to have a description of a solid ball, even though it’s made of atoms, which are actually mostly empty space. So, that solid ball description leaves all that out, and it works just fine. It tells me exactly where the ball will land [unintelligible] measure it.
TYSON: David, you and your consciousness cronies, is it generally recognized that consciousness is an emergent phenomenon of a complex brain?
CHALMERS: Yeah. Well, this word emergence is kind of word that people used to cover a huge variety of sins. I mean, sometimes I think it’s kind of a magic word we use to make ourselves feel comfortable with things we don’t really understand. So, ah, that’s emergent.
There’s different kinds of emergence. There’s the kind you get with, say, complex systems like the Game of Life; Conway’s Game of Life where the cells blip on and off, and you get complex phenomena like gliders that move along.
You know it’s surprising, and you wouldn’t have expected it, but you can put together the equation that it’s totally predictable. You run the game of life over and again with simple computational rules, it’ll be predictable again and again.
Evolution is interesting at the immediate case. Maybe given the laws of physics in certain initial conditions. You can run them again and again. I don’t know. Maybe you’ll get—maybe it’ll turn out evolution arises 60 percent of the time. If so, that’s incredibly cool, and then that ought to be explainable in principle.
Now, for consciousness, people sometimes say consciousness is emergent, but there’s a gap there of a kind that we haven’t even begun to close in the gap of consciousness. People can tell stories about life. People can tell stories about evolution. No one’s even begun to tell a story that enables you to predict the existence of consciousness from any number—any amount of underlying physical dynamics.
It explains the behavior. It explains how we walk, how we talk, but why that should actually feel like something from the first-person point of view, that is emergent in a much stronger sense. I’d say that’s strongly emergent in the sense of it might require new principles to explain.
TYSON: Max, is there any role of chaos theory in this? Because we know that in principle and in practice there’s some systems that are so complex, that you cannot accurately predict its future behavior. Now, is that true even if you had an infinitely powerful computer?
TEGMARK: No matter how powerful a computer we build on earth, we can certainly not predict—we could not have predicted that the Red Sox were going to win the World Series right after I moved to Boston.
TYSON: Okay.
TEGMARK: Because precisely of chaos theory, where tiny changes in the position of some particle made a huge difference later on. But—
TYSON: Just the Butterfly Effect.
TEGMARK: Yeah. If things—
TYSON: I’ve got to tell you real quick, there’s the Journal of Irreproducible Results, which is if you’re a scientist and you come up with something that you know isn’t right, but it’s a really cool calculation, you publish it there. And it’s like in there you’ll find the calculation of what happens if you strap a jellied toast to the back of a cat. Since toast always lands jelly side down, and cats always land on their feet, what would happen if this dropped? Okay.
And so in the paper, they hypothesize that the cat falls, and then hovers over the—so, it’s stupid fun calculations. One of them was—sorry for this interlude, but one of them was there was some major storm system that happened that hit the East Coast of the United States, and someone said, “We found the butterfly that caused this.” And they killed it and it was on display. So, go on. So, this Butterfly Effect—
TEGMARK: Yeah, yeah. I was just- apropos of complicated emergent phenomenon related to chaos and such. I just wanted to come back to what David was saying about consciousness here, and kind of connect it with what you opened with here. How can we test with scientific methods these ideas of whether we’re simulated or not? Or at least update our odds in one way or the other. I think one thing that’s great to do is what you’re doing. Again, looking for this evidence of a simulator cutting corners to make the simulation easier to run.
I think another thing we should do is if you want to test this computation—hypothesis that everything is a computation, or that everything’s mathematical, we should look precisely at the things where we’re the most clueless right now about how we would actually describe it mathematically. And I can’t think of anything we’re more clueless about right now than consciousness.
And try our very best to see if we can bring also that in under the type of things that we can describe with math. If we fail spectacularly on that, and can realize why, we’ll see, wow, our universe is not mathematical. Boom, done. Death to the simulation hypothesis. Whereas, if you and your cronies, as we’re told that they’re called, succeed, that would I think be a big boost for the simulation hypothesis.
CHALMERS: Yeah. And there are people who are pursuing the idea. As you know, the consciousness is fundamentally about information processing in the right way when information, for example, is integrated in just the right way. Maybe you get a kind of consciousness.
That’s still a very controversial idea, and a lot that it doesn’t explain. But if something like that is right, it goes very naturally, at least, with the simulation hypothesis because it’s very natural to suppose that in a simulation there could be all that information being integrated and giving you consciousness.
Certain other views—just say, for example, consciousness requires a certain very specific intrinsic property like a certain specific biology. Then there could be a simulation of the whole universe. But if it didn’t have that biology, then no consciousness. It would just be a world of unfeeling—
TYSON: Of world.
CHALMERS: —zombies.
RANDALL: I have a question, though.
CHALMERS: You know, unfeeling physical dynamic. So, it really makes a difference.
RANDALL: How do you ever show that something can’t be described mathematically? You’d have to believe you understood fundamentally what the degrees of freedom are. So, you might just have the wrong description. I mean, even in physics, I mean, we know classic examples where people thought certain things were impossible until just a new law of physics was discovered.
I mean, Darwin got the age of the world—our world closer than the greatest physicists of the time because Darwin just looked around, and Kelvin thought he knew the laws of physics, and he didn’t get them right. So, I don’t see how you’re ever going to be able to show that something has no mathematical description.
TYSON: But, Max, you’re big on the mathematical concept here. What you’re saying is everything is mathematics. And if everything is mathematics, then everything is programmable.
TEGMARK: That’s right. That’s right. And so I think as an answer to Lisa’s question, David put it very well in the beginning. In physics, we aren’t ever able to really prove that something is true. The only people who prove stuff are mathematicians.
But if David and Giulio Tononi and Christof Koch and others succeed in this endeavor to try to actually explain consciousness mathematically, it wouldn’t prove that things are purely mathematical, but it would certainly be yet the great boost.
RANDALL: I asked the other question of how you [unintelligible].
TEGMARK: Because if you just go back—let’s go back to Galileo again. We were eulogizing him earlier, right, for his great insights. When he wrote that our universe is a grand book written in the language of mathematics, that was 400 years ago because he was so impressed that things moved in parabolas and things like that.
He had no clue why oranges were orange and hazelnuts were hard and some things were soft. That seemed like it was beyond what he could do with math. Then we got Maxwell’s equations, the Schrodinger’s equation, the standard model of particle physics. More and more has been explained by math. I think Galileo would be really impressed if he were on stage. So, it’s really cool to look at what are the things left.
TYSON: I’ll invite him next time.
TEGMARK: [Unintelligible]. Well, you can reincarnate him and bring him on.
GATES: Just simulate him.
TYSON: We’ll just have to simulate him. That’s what we’ll do.
TEGMARK: So, it’s really cool to look, well, what’s left. Like consciousness, for example, and see if we can also make some progress there. There’s no better way to fail on anything, including consciousness understanding
than to tell ourselves, oh, we know it’s impossible because of some principles, and let’s not try.
TYSON: Yeah, those aren’t good scientists who behave that way.
CHALMERS: I think we have to distinguish, though, between the two claims that you can give a mathematical description of everything, and you can give a complete mathematical description of everything. Even consciousness, obviously, give many mathematical descriptions of color space has certain geometrical properties, the light, the feeling of the light is more or less intense. You can give a very rich mathematical description of it.
And that’s what, say, someone like Tononi is doing. But can you give an exhaustive mathematical description of it? Once you’ve given a full mathematical specification of consciousness, have you understood everything about it, or is there some further nature like the redness of the red, or the blueness of the blue?
TYSON: What Max is saying is that previous frontiers in that question were ultimately breached when enough smart people came along to figure it out. So, whatever’s our state of mind today, it would be unwise to suggest that it somehow transcends any access that the future of math might—
RANDALL: I mean, on some level we don’t have an exhaustive description of anything because we understand that there can always be something more fundamental, something we haven’t seen yet.
GATES: I agree.
TYSON: In fact, the very word atom in Greek means indivisible.
CHALMERS: Yes.
TYSON: So, yeah, with that—how long did that last?
RANDALL: And unchanging.
TYSON: Yeah.
GATES: While we have been all bowing at the altar of mathematics—
[LAUGHTER]
GATES: a number of us are aware of this result by Gödel called the incompleteness theorem. And it even says in some sense mathematics is incomplete. There are things in mathematics that you cannot prove. That’s what the theorems say. And so we, as humans, I think—
TYSON: In fact, Gödel proved it.
GATES: Yes. Yeah, right. He proved it. That’s exactly right.
TYSON: Gödel proved that math cannot be proven.
GATES: That’s right.
TYSON: Yeah.
CHALMERS: If it’s consistent.
GATES: Right. If it’s consistent.
[TALKOVER]
TEGMARK: In defense of our universe here—
TYSON: Somebody’s got to defend our universe, so go ahead.
TEGMARK: Standing up for our universe. There’s actually no evidence that our universe is inconsistent, or that mathematics is inconsistent. Gödel said that we humans, we cannot prove ever that mathematics is consistent.
TYSON: Right.
TEGMARK: We cannot prove that—that’s impossible to prove that one equals two. But I think that’s probably more of a reflection of our own limit—of the limitations that thinking beings have, rather than our universe has some kind of identity crisis. Our universe seems to know exactly what it’s doing. It doesn’t seem very inconsistent except when I watch the Presidential Debates.
TYSON: Okay.
GATES: Oh, boy, I’m not going there. But, Max, that was precisely my point, that maybe what we’re talking about is in fact part of our limitations. Not limitations on the universe, but—in science it’s very funny because the way we do science—well, when I give public talks, I like to say if you look at family in their house, you might be an anthropologist and record what they do, and they turn the appliances off and on. And you might come up with some big record book of this.
But then when everybody’s out of the house, you might just go to the house and watch how it behaves. The thermostats go up and down, and maybe you have a timer that does other things. And so the house has a set of rules for operating when you’re not there. And in some sense, in science, that’s what we’re doing.
And when we do this split between science, non-science, in some sense we’re talking about how the universe behaves as if we could take our consciousness outside of the universe. And that’s a very sudden point to appreciate. And so maybe what this whole discussion has been about is actually just our limitations.
TYSON: So, we’re all stupid, is what you’re saying.
[LAUGHTER]
GATES: Actually, the universe made us very clever, at least most of us.
[LAUGHTER]
TYSON: So, Jim, I got to ask you something. Your discoveries of the checks—error-correcting code within the laws of physics themselves, at the depths that you’re researching them, what I wonder is we live in the age of IT, of information technology. So, we all have a certain fluency. So, it’s in our brains to think that way at some level.
Could it be that how the saying goes, if you’re a hammer then all your problems look like nails, and you solve them by hitting them. If now we are in an IT revolution, and you’re finding IT solutions to your problems, maybe it’s just the fad of the moment. And you’re forcing a solution that is either not real, or there’s a better one awaiting in a revolution that has yet to occur.
GATES: Sure. So, the last time I was here I actually misspoke. I used the name of Shannon when I meant Hamming code instead. So, first, let me correct that for this wonderful audience and mention it—
TEGMARK: Error correction in action.
GATES: That’s right.
TYSON: Error correction—
GATES: Error correction in action, absolutely. But in—look, in our work, first of all, we don’t know it’s the physics of our universe. There is a large experiment underway that Lisa knows a lot about in Geneva because she has written papers about possible outcomes in these observations.
TYSON: Lisa, are you flying to Geneva tomorrow?
RANDALL: I am, but not for that.
TYSON: Not for that, okay.
GATES: So, the Large Hadron Collider is going to explore more of the structure of the universe. So, first, the mathematics that I have done will only become physics, or relevant to nature, when the LHC or some other observational device says the idea of supersymmetry is correct. Then it will kick in. So, that’s a big if. There are lots of physicists who don’t believe the universe will be supersymmetric. In which case, all I’ve done is an interesting mathematical fairytale.
TYSON: So, supersymmetric proposes a whole other regime of particles that are counterparts to the particles that we’ve come to know and love?
GATES: Correct.
TYSON: Okay. And they’re yet to be discovered, but they could be describing a whole other parallel reality, awaiting our discovery?
GATES: Well—
RANDALL: But even that—I mean, I just want to clarify. We may or may not find evidence at the Large Hadron Collider, which is what’s being discussed. But that doesn’t even mean that supersymmetry doesn’t exist. It means that we can’t find the evidence at the scales that we can probe.
GATES: Exactly.
RANDALL: So, it could be that there is some fundamental symmetry, and it’s broken at such a high scale that we cannot access any of the evidence of it. And that’s the world we live in. I mean, that’s what we do as scientists. We try to simulate what we can.
We try to derive what we can. We try to measure what we can. And then we have to allow for the possibility that we just haven’t had the accuracy. We haven’t had the cleverness, or we haven’t had the resources—
GATES: Technology.
RANDALL: —to be able to test certain ideas. And so I think that’s right, that it’s a combination of what’s out there and what we can actually do.
TYSON: So, I don’t who among you to ask this direction, so I’ll just put it out there in front of you like a piece of raw meat, and you can chase after it, if you—
RANDALL: You think we’re dogs? [Unintelligible].
TYSON: Or you vegetarian—some raw carrots. Okay. So, you can chase after it. I don’t know if any of you are vegetarian. So—
RANDALL: Can you cook the meat at least?
TYSON: I’ll cook the—okay. We’ll cook the meat. My question is I remember physics 101 and 102 and 201 and 202, and as you learn the laws of physics, every now and then something pops up that’s just kind of weird. All right? You learn Maxwell’s equations, which describes the behavior of electromagnetic radiation, the behavior of light, and they’re really beautiful except there’s an asymmetry in there. There’s like you can have particles that have electric fields like electrons, but you don’t have isolated particles that are their own magnetic fields. There’s always a plus and a minus stuck together.
So, they’re not symmetric that way in the equations. And it’s like you cringe when you see that because part of us wants some beauty and symmetry to the universe if it is—I don’t know. We’re holding it in very high—holding very high expectations for what we want to find.
And then you go back to the early universe, and you find out that one out of 100 million—one out of 100 million photons did not become a photon because symmetry was broken, and it made only one matter particle. Whereas, all the other interactions had matter and antimatter they annihilated and became photons.
And we are made of this one in 100 million stuff that’s left over. Something broke in the early universe. And I ask you why aren’t these bugs in the program that we’re dealing with?
RANDALL: So, I’m going to actually answer that.
TYSON: You’ve got an answer for that? Very cool. Very cool.
RANDALL: So, it’s definitely not a bug in the program because in both these cases, the underlying laws actually do exhibit symmetry. As Jim knows really well, that it has to do— in our description of electromagnetism, you have electrically charged particles.
There’s an alternative description where the fundamental particles would be magnetic. That’s not the universe we find ourselves in. So, a lot of the symmetry is broken by the actual state of the universe we live in. So, it could be that the laws of nature have some underlying symmetry that gets broken at some point.
TYSON: So, who’s breaking it?
RANDALL: Who’s breaking—the universe. The universe is [unintelligible]—
TYSON: No, that’s not the answer.
[LAUGHTER]
TYSON: I’m looking for a little more insight into who’s breaking the laws of the universe than just the universe.
RANDALL: Well, here’s a simple example, okay.
TYSON: Well, just to be clear, we come up with what we saw are laws, and then if we see an exception we say that there’s a case where the law is broken.
GATES: No, no.
RANDALL: Okay. Let me give you a simple example.
TYSON: And we’re okay with that.
GATES: It’s the symmetries that are broken.
RANDALL: Suppose I have a pencil—
TYSON: Oh, sorry. Symmetries that are broken.
RANDALL: So, say I have a pencil standing on end. I have rotational symmetry, right? We’d like to believe everything’s rotationally symmetric. Why should one direction be different? So, I have a pencil standing on end. It’s going to fall down. It’s going to fall down in some direction.
Now, who made it fall down in that direction? No one made it fall down in that direction, but it was going to fall down in some direction. So, the symmetry is broken. We didn’t ask the symmetry to be broken. The fundamental laws were perfectly symmetric, but the symmetry is broken.
And there’s many things in the universe that are like that. The fundamental laws are symmetric, but the actual universe we live in has broken.
TYSON: So, we can’t look for weirdness because if it is a program that’s running, which came up earlier, and we’ve all had programs that crashed, what happens if our program crashes? Do we all disappear like instantly?
What are the consequences to this being a program if someone unplugs it, if there’s a bug that crashes the entire system? Is there any piece of the universe where that part of the program failed?
CHALMERS: I have it on good authority [unintelligible 79:24]—
TEGMARK: A big spinning wheel here on the stage going round and round and round.
CHALMERS: I have it on good authority it’s crashed five times during this panel discussion, but, fortunately, it rebooted perfectly and we have no memories of it. That’s just good error correction.
GATES: No, no, but, Neil, the point you raised, in fact, is for me one of the most uncomfortable ideas about the simulation hypothesis. That it’s running on some device, and that the errors would then—how would it manifest itself? Well, in the way that I think most of us think about it, it’s kind of the end of the universe.
And, for me, the universe that I have studied for 50 or 60 years is a kind of a—it’s a place of mystery, but it’s not a place of the fundamental kind of insane, unleashed chaos, that kind of end.
Now, we know that—we talk about, for example, false vacua. That’s something, again, that Lisa knows a lot about because it was pioneered largely by Sidney Coleman, a professor at Harvard before Lisa got there.
We know that these possibilities are out there, but the breaking of the symmetries are so—one thing that’s really odd about this is if you don’t break the symmetries you don’t get us. You don’t get a universe with creatures like us in it unless you break these symmetries. And so maybe the question we should—
RANDALL: The simulation’s like, “Why did I break those symmetries?”
GATES: That’s right. The simulation’s like why am I breaking those symmetries?
So, the fact of our existence says something very deep about the mystery of this place we call the universe because the laws—the symmetric laws, they’re beautiful. We write them with simple equations on one or two lines. But if those laws held exactly, we’re not here.
TYSON: In fact, it’s just a universe of photons.
TEGMARK: I think that’s a very good point you bring up there. At first, it looks like if someone’s simulated this, they have been drinking too much or whatever, or really wasteful because you might ask why—if they just wanted to simulate us, did they bother simulating all this dark matter? Six times as much matter, obviously, increase their CPU cost, what they had to pay for their über cloud services, whatever. Who needed that?
TYSON: Plus we came really late in the universe.
TEGMARK: But every single thing we’ve discovered, like dark matter, for example, that seems superfluous, we’ve since discovered that if it weren’t there we would be dead. Or, in fact, we wouldn’t even have evolved in the first place.
If there were no dark matter, for example, then its gravity would have not been there to help pull our galaxy together, and the Milky Way wouldn’t even have existed.
So, it’s an interesting question, I think, to ask is this the simplest kind of simulation you could run that would actually get some interesting life? Or is there something in our universe, which is really just bells and whistles that you could optimize out?
CHALMERS: Someone was just doing—this kid was just doing a science experiment. He ran a million simulations overnight, and exactly one of those universes produced, you know, broke the symmetries in the right way to produce conscious beings and, hey, here we are.
RANDALL: Why did they make [unintelligible] so difficult to simulate on the [unintelligible]?
GATES: That’s a scientific question, guys.
TYSON: So, Zohreh? Yeah.
DAVOUDI: So, maybe just adding something to this. If someone was just looking at the weirdness that we observe in the universe, maybe more fundamental question to ask—again, we can ask why the parameters of our universe, mass of the electron or the cosmological constant and things like that, why should they have the value they have?
In terms of the simulation scenario, you can sort of start to think this is just an input as many other input. Or the other way to interpret it is that we don’t know, at a fundamental level, what’s going on. Maybe there is embedding theory that would arise to—that is simpler. It has fewer input, or maybe just one, and then gives you the values of the standard model and all these theories that, no, to be exactly the same that we observe in nature.
TYSON: Well, just to be clear, when we—if any of us program a computer, a simulation of anything, there’s a set of parameters that are established up front. And then you watch what happens thereafter, and then you sometimes tweak the parameters if necessary. Some other parameters are non-tweakable. Almost all of our codes, there’s a line that gives the value of pi that’s not tweakable.
DAVOUDI: We don’t have any mathematical—sorry.
TEGMARK: My sons tell me, for example, that in Minecraft when you create a Minecraft world, I’m taught by Philip and Alexander, you have to input a world seed.
TYSON: Okay.
TEGMARK: Yeah. And if you put in a different one, different universe.
DAVOUDI: Basically, it means that we don’t have the mathematical equations, for example, to say that the mass of the electron should be what we measure and things like that. So, we don’t have yet a description as why these have these values.
TYSON: But why should we be—this might have to go back to David. Why should we be the measure of what an intellect is, and then judge what is hard or what is easy? So, in other words, just because we think something is hard because of all of these physical constants that come together, so that we exist many billions of years after the universe forms, maybe that’s just trivial for anybody who’s programing the universe.
CHALMERS: Yeah, I mean, it probably is trivial. They’re probably got Sim universe technology. Everyone’s running it on their desktop, I mean, someone at Google in the next universe up—
TYSON: The next universe up. That’s now a phrase.
CHALMERS: —created a Sim universe by, okay, set a few parameters around the universe. No big deal. Some of those universes produce nobody. Some of those universes produce somebody. And those somebodies have to reverse engineer their universe, and it turns out reverse engineering is really hard, whether you’re in a simulation or not. But that’s just—if you look at it this way, it’s just a matter of perspective.
TYSON: Because when I think of the game Tic-Tac-Toe, to a child this is a challenging game. They don’t know what move to make next, and then they might win or lose, and then they cheer, or they’re sad. And then you realize this is a pointless game you can play so that you’ll never lose, or win. But then it’s no longer fun.
But to a child, it is a complex—it’s a game that challenges them. And then we have the game of chess, which challenges us, but you go up an intelligence level, and then it’s just a trivial exercise. I don’t care how many possible moves there are. It’s trivial if the brain is a greater brain than ours.
And I’m reminded, which one of you may remember—just to correct me if I don’t get it right—was it Feynman who first analogized the laws of nature to our attempt to understand the laws of nature would be like coming upon a game of chess and you know nothing of the game, and you’re just watching people move, and you don’t have the rule book. And you have to deduce what the rules are.
And so pretty easily you can see, well, this piece moves this way and this only goes diagonal. You get that, but occasionally one of the pieces jumps two squares instead of just one. Well, why did it do that? So, you make a note of that, right? And then later on that little piece that jumped too, it reaches the other end of the board, and then it becomes a whole other piece. That’s kind of freaky. It’s rare, but it happens, and it’s an important rule of the game that most of the time you don’t see.
And so I’m left wondering how much of a chess game, without the instruction manual, is the very universe in which we live. To get your answers one each from that. Yes, Zohreh?
DAVOUDI: Maybe starting from the philosopher’s side.
[LAUGHTER]
TYSON: All right, David, you led off with that.
CHALMERS: I would say that’s basically the situation we’re in. We call that—this is the game of reverse engineering the universe, and we call it science. A little bit for philosophy, too.
There are clues we could get about the—we certainly get the equations and so on, but, hey, there are clues we could get about the grander structure. Hey, maybe at a certain point we’re going to find one of those constants that has an arbitrary value, and find there’s a coded message in there in the simplest possible language saying, yeah, you guessed right. It’s a simulation.
If so, then that’s part of the reverse engineering, too, but it’s actually a miracle we can understand anything about the world we’re in from this perspective. The world could have so much complexity, which is completely beyond us. And it could be that we are simply scratching the surface. But I think to do science, you’ve just got to take the optimistic.
TYSON: But you left me very disappointed earlier in saying that anything—I was trying to find evidence to show that we’re not, and that evidence that we’re not could be put in by someone who is.
CHALMERS: Yeah, I’m afraid—
TYSON: So, that was disturbing. So, I’m getting back to Zohreh’s point. What’s the point of thinking that way?
CHALMERS: Well, you might take on board Occam’s razor, which is if we don’t need the hypothesis that we’re in, a simulation, then we should just do without that hypothesis. Maybe science is going to tell us a bunch of math. It’s a bunch of equations. Yeah, that could be combined with the further hypothesis we’re in a simulation. It’s a lot simpler if we’re not. So, Occam’s razor at least says why bother.
Then, on the other hand, you’ve got Bostrum’s statistical argument. We might actually produce a whole lot of simulations, and have very a good reason to believe there are a lot of simulations in the universe, at which case you can just raise the question. We know some people are in simulations. We cannot assign at least probability of zero to us. At that point it becomes a statistical question.
So, I think the standard scientific reasoning of Occam’s razor might give some reason to reject it, but the statistics starts to maybe balance that in another way around. At a certain point, you’ve got to start doing some math about the probabilities.
TYSON: So, Occam’s razor actually goes very far back, I think, to the 14th century, where it was the Earl of Occam who suggested—
CHALMERS: William Occam.
TYSON: —that of all explanations, perhaps the simplest has the best chance of being correct. And that was well before the methods of the scientific method and others. So, we’ve been using it an invoking it ever since. So, Lisa, is the universe a chess board and we’re—
RANDALL: So, I was very much with you until the very end.
[LAUGHTER]
RANDALL: So, I think it is indeed true that we don’t know the answer, and we’re just going to keep doing science until it fails. And it hasn’t failed yet. Seems to—we make progress, and so we’re going to keep doing that. In terms of are we trying to figure out, I actually love the idea that we’re a simulation where they actually kind of saved in efficiency by making us not quite smart enough to figure all this stuff out.
[LAUGHTER]
RANDALL: So, [unintelligible] a lot of it.
TYSON: That’s a built in—
RANDALL: But, really, that’s too much computational power, so let’s make them a little bit dumber.
But I do have—I understand it’s a possibility there’s a simulation, but there is a problem with the statistical argument. I mean, I think if you asked any statistician, there’s just not based on well-defined probabilities here.
And actually one of the key—so, Bostrum’s argument would say that also that you have lots of things simulating, lots of things that want to simulate us. And I actually really have a problem with that. Why simulate us? I mean, there’s so many things to be simulating. None of us actually get together and say—I mean, we simulate processes or whatever, but we mostly are interested in ourselves. I don’t know why this higher species would want to bother with us.
TEGMARK: Maybe they don’t care about us. They just simulate a bunch of physics, or a bunch of laws. And, hey, we came along as a by-product.
RANDALL: Yeah, it’s in the realm of possibility.
TYSON: So, we grew out of their Petri dish.
RANDALL: But I do think that, again, ultimately what—as physicists, as scientists, we’re interested in the things that we can actually test. So, to the extent that it gives us an incentive to ask interesting questions like do we see cosmic rays at different energies, or from different directions, going at slightly different speeds, or anything of that nature.
Or do we find the laws—I mean, that’s certainly worth doing to see what is the extent of the laws of physics as we understand them. But that is what we’re doing anyway, but maybe we’ll frame—maybe we’ll be presented with a bunch of other questions. But that is—I mean, so it’s a little bit of a systematic way of figuring out the chess game because in the case of the chess game, you have many games. You can watch many games.
Here, we have this one universe, and we can try to make little tests within that universe to try to test laws, but those apply in those little realms. I mean, one of the brilliant things about Galileo was he realized there’s many ways to do science. There’s thought experiments, there’s observations. But he actually came up with the idea of experiments themselves. [Unintelligible]—
TYSON: We were bringing him to the next panel.
RANDALL: Yeah, I know. I know. I’m totally excited about that one, too.
[LAUGHTER]
RANDALL: So, that’s just the nature of science. So, I think we are trying to figure it out to the extent that we can.
TYSON: Jim?
GATES: I think you nailed it with the chess game analogy. One thing that I think that a lot—often times, when I talk to people—and Lisa alluded to this very well, is that a lot of people think it’s all about them. They really do. They think it’s all about them. They have to understand things that’s somehow related to them.
It’s all about them. And I think that one of the things that science actually teaches us is that it’s not all about us. We may be struggling with the description that we’re trying to construct, but the universe doesn’t care whether I understand or don’t understand.
The universe doesn’t care whether I exist or not exist. The universe, at least as I have studied it, is- I’m going to retreat into Einstein, it’s a- at the end of the day, it is an extraordinary mystery. That’s the sense that I get from having studied science for now 50 years almost. That we live in this place of mystery, and we need to accept a humbleness about our efforts to go out and explain that chess game that you described.
TYSON: Ooh. Max?
TEGMARK: I fully agree with you that the world would be a better place if we humans could be a bit more humble. At the same time, I also feel that the very soul of physics is this audacity to look for hidden simplicity in things. So, I think the metaphor of chess is a beautiful one.
We have as a goal in physics to look at this very complicated, messy-seeming universe and look for hidden simplicity, look for rules of chess, which are actually simple. Not the list of one googolplex different possible things. We’re not just saying it’s all random.
And, of course, we don’t know yet whether there are rules that are simple enough that our human minds can understand them or not. But I’m an optimist, and I feel it’s actually much healthier as scientists if we have this innate optimism instead of saying, well, it might be they were too dumb to ever figure this out, so let’s just not try.
I think it’s much healthier to say there is a real possibility that there is this hidden simplicity, and, in fact, Galileo and Einstein and so many before us have found simplicity far beyond what their ancestors every dreamt to. So, let’s keep looking for even more hidden simplicity.
Maybe this is actually all computational mathematical, which that’s anyway the ultimate audacity to hope for that because that would mean that in principle, at least, it really is possible to figure out the rules of the game. That’s that attitude I’d like to take. Consider the possibility that it is possible, and then try our very, very best to actually figure it out.
TYSON: Thank you, Mario, for that and, Morpheus. So, Zohreh?
DAVOUDI: So, yeah. I would like to deviate a little more from your question very quickly because it didn’t came up in the discussions, but I wanted to distinguish between the idea that we can simulate a universe, as opposed to the universe being a simulation, because there are fundamental limits to our capability to actually compute things.
And these are based on the physical laws that govern our universe. Basically, we can’t have infinite power, the energy, the laws that govern, basically, the uncertainty principle can limit the rate that we can process logical operations, and also the entropy and thermodynamic laws can limit the amount of memory that we can hold in a given amount of the space-time.
And thanks to ideas like this that were discussed by [unintelligible] and other people, and therefore we might not be able to actually—and there are other actually limits when you think about larger-scale expansion of the universe, whether or not it can ever causally connect to parts of the universe. It’s expanding. And, therefore, store and process those information, be able to actually re-simulate the universe that we have.
So, it’s a different idea. I don’t think that based on the physical laws of our nature this could be possible, but that doesn’t mean that our universe could not be a simulation inside another universe that has another laws of physics that doesn’t actually limit the amount of computation that is required to simulate a universe. So, these are two different ideas.
But just to come back to your question, I think as a physicist and thinking about the simulation idea, I think it doesn’t change the way I think about the science, and I do my every day job as a scientist.
I think just the notion of whether or not we’re real or just simulated, it’s kind of irrelevant because what we are observing is no different from being real or imaginary. We just go and discover things that we already don’t know about the universe, that laws that we haven’t discovered. But at some point, maybe we find some sort of more strong evidence that could connect us to a higher level that says something about whether or not the universe is computational based and there is some simulator besides us.
These are the ideas that require more thinking and more thinking out of the box, I would say, at the moment, but maybe at some point in future when we have more understanding of the laws of our universe, we can have more rigorous way to go and look for those evidences and say something meaningful.
At this point, there is not such evidence. We’ve just started to make assumptions by just comparing our simulations of the universe and see what would be the consequences of those kind of assumptions. But at the moment we don’t have such evidence, and it would be wrong to put a lot of focus on this idea. But it’s definitely a very fun and curious idea to think about as a scientist and, therefore, I think that’s why I do science.
TEGMARK: I just have to alert you she knows the answer to your chess question, and she’s just not telling us because you’re Persian and you Persians invented chess.
[LAUGHTER]
TYSON: Okay.
DAVOUDI: Yeah. I actually played that game when I was very, very young.
TYSON: So, she does have the answer.
[LAUGHTER]
TYSON: So, let me just end before we transition to Q&A. I want to get the likelihood that you think we are in a simulation. Ten percent chance? Twenty percent? Just give me a number. Just a number. Go.
DAVOUDI: I can’t give you that number. I don’t have any answers.
TYSON: No.
[LAUGHTER]
TYSON: She’s not authorized to divulge that information. Okay, so you’re giving no answer. Max?
TEGMARK: Seventeen percent.
TYSON: Seventeen percent.
[LAUGHTER]
TYSON: Jim? Morpheus?
GATES: One percent.
TYSON: One percent chance.
RANDALL: I’m going with effectively zero.
TYSON: Effectively zero. David?
CHALMERS: Forty-two percent.
[LAUGHTER]
TYSON: I think the likelihood may be very high. And my evidence for it is just it’s a thought experiment, and it’s simple. We’ll just end with this reflection—and I’m elsewhere like on YouTube saying this, so you can check it out later, if you choose.
I just think when I look at what we measure to be our own intelligence, and we tend to think highly of it, getting back to Jim’s point, there’s a certain hubris just even in how we think about our relationship to the world. And that’s understandable perhaps, even in the search for intelligent life in the universe.
It comes with the assumption that we’ll find life that also thinks we are intelligent. Well, if we look at other life forms on earth with whom we have DNA in common, there is none that we would rank ever in the history of the fossil record, or life thriving today, that we would rank with us and our level of intelligence.
So, given our definitions, we’re the only intelligent species there ever was because we have poetry and philosophy and music and art. And then I thought to myself, well, if the chimpanzee has 98-whatever percent identical DNA to us—pick any animal. It doesn’t matter. Dogs, it doesn’t matter. Mammals have very close DNA to us. They cannot do trigonometry.
Some people can’t do trigonometry. Certainly not these animals. So, if they cannot do trigonometry, and they have such close genetic identity to us, let’s take that same gap and put it beyond us and find some life form that is that much beyond us that we are beyond the dog or the chimp. What would we look like to them? We would be drooling, blithering idiots in their presence.
The smartest chimp can do maybe some sign language and stack boxes and reach a banana, put up an umbrella, like our toddlers can do. Our toddlers do that.
So, maybe the smartest human—bring Stephen Hawking forward in front of this other species, and they’re chuckling because they’ll say, oh, this happens to be the smartest human because he’s slightly smarter than the rest because he can do astrophysics calculations in his head, like little Timmy over here.
[LAUGHTER]
TYSON: Oh, you’re back from preschool? Oh, you’ve just composed a symphony. That’s so—let’s put it on the refrigerator door. We just derived all the principles of—oh, that’s cute.
And so that is not a stretch to think about. And if that’s the case, it is easy for me to imagine that everything in our lives is just the creation of some other entity for their entertainment. It is easy for me to think that. So, whatever the likelihood is: zero percent, 1 percent, 17, 42, no answer, I’m saying the day we learn that it is true I will be the only one in the room saying I’m not surprised.
Thank you all for coming tonight, and thank the panel.
[APPLAUSE]
TYSON: We will bring up the lights, and in this transition between the formal part of the panel and the Q&A that we’ll be taking from you, I just want to tell you what it takes to run this thing.
As I told you earlier, it was formed by an endowment created by Isaac Asimov’s widow, Janet Asimov, and friends of Isaac Asimov. And we’ve been going strong ever since. And so I just want to say we have people who run this thing.
We have Suzanne Morris, who’s director of Hayden programs. We have Suzanne. Suzanne is out there.
And we have Emily Haidet. She’s also part of this team that make this work. We have my executive assistant, Elizabeth Stachow. We have Laura Jean Checki, who’s our stage manager.
[APPLAUSE]
TYSON: We have some fans of Laura in the audience. Good. We have Miriam Poser, who has been with us like forever and is a die-hard supporter as a volunteer of our programs. Lydia Marie Petrosino—did I pronounce that right, Lydia? I only ever call you Lydia Marie. Lydia Marie is good. And, of course, we have Betty Walrond.
These are people who make this happen every single year. I just want to collectively give them applause.
[APPLAUSE]
TYSON: We have about 15 minutes for Q&A, so let’s go straight to it. We’ll go back and forth, left and right. No one will hear you unless you speak into the microphone, so let’s start here. Let’s go.
AUDIENCE MEMBER: Okay. So, my question is does it really matter—would you view the universe differently if we knew it was simulated? Could we do the math differently?
TYSON: Ooh, I like that. To make this efficient, we’ll pick one person to answer, so how about Max because your book is tilted—what’s the title of your book?
TEGMARK: Our Mathematical Universe.
TYSON: Yeah. So, he’s the guy to answer this question. Okay. If you knew, would your math be different?
TEGMARK: Well, I’ll take your question. I mean, if I knew, would that make me super depressed or super excited? Would it change the way I feel about everyday life? My answer is absolutely not.
I feel that when we look at these rather sterile equations or the computer code, or whatever it is that’s running this, there’s no meaning or purpose built into that. We shouldn’t look through our universe creating the meaning for us. It’s we who give meaning to our universe. So, the way we feel about things, and the meaning we create, is the same regardless of whether we’re simulated or not. And I think this is very much the point that David was making earlier also.
We shouldn’t dis things just because they’re simulated.
CHALMERS: The math is a little bit different if we’re simulated, on the other hand, because there’s also the math of the simulating universe.
TEGMARK: Right.
CHALMERS: There’s a math of our universe embedded in a bigger one, which raises some exciting prospects like, hey, maybe we could get out and explore that one. So, that would be cool.
TYSON: Wait. So, you’re swaying—can you embed a complete system of mathematics within a higher system of mathematics?
CHALMERS: Yeah. Maybe the embedding universe is a vastly higher level of complexity than ours, in order to have the computational power, for example, to simulate ours. Maybe they’ll let us out one day. We’re just computers in their world. They’ll give us input devices.
TYSON: So, you feel like you’re in prison?
CHALMERS: No. it’s just like earth is cool, but the galaxy’s even cooler.
TYSON: Okay.
[LAUGHTER]
TYSON: Yeah, a quick one here.
RANDALL: I think it would feel different. It’s always interested me that if you miss a basketball game and you know it already happened, it’s less interesting to you to watch, even though you know it happened already because it’s not happening in real time. And you might not even know the result, but I think there is a sense in which psychologically the idea that it is preprogrammed would be disturbing, at least to me.
TYSON: My analogy to that is if you go to the Smithsonian Air and Space Museum in Washington, and you see the Apollo 11 Command Module that went to the moon and came back, and there it is, if you made an exact replica of that and put it on display anywhere else and say you cannot really tell the difference except microscopically, but it’s a fake, it means—it’s different to you even though you can’t tell the difference. The knowledge that it’s really seems to matter to us than if it’s a simulation or a model, in that case.
So, I have to agree. I reluctantly agree. I don’t want to agree, but I have to agree.
GATES: I disagree for one reason. I don’t do science to make me feel something. I do science because I think it’s an investment in the long-term survival of our species. Our science underlies our technology. If our environment changes, we will use that technology to survive, whether that’s true if we’re simulations or not. That’s why I do science.
TYSON: Are you also running for president?
[LAUGHTER]
RANDALL: No, because they can’t talk about science.
TYSON: Who, by the way, Jim Gates is on the President’s Committee of Advisors of Science and Technology, PCAST. And you’ve been there for almost the entire administration, so keep up the good advice that you’re giving.
[APPLAUSE]
TYSON: Let’s go right here. Hey, how are you doing?
AUDIENCE MEMBER: Good.
TYSON: What grade are you in?
AUDIENCE MEMBER: Eighth grade.
TYSON: Eighth grade, cool. So, what do you have?
AUDIENCE MEMBER: So, you were saying about bugs in the code of the universe, if it is a simulation. How come if it probably statistically would not be perfect, how come we have not so far seen any corruption or glitches maybe in the far-looking like the cosmic background radiation? How have we not seen anything that just seems like it couldn’t be there?
TYSON: Great question. Jim, what do you got?
GATES: That’s an easy one to answer. Up until September of last year we have never seen gravity waves either. The point is that our technology was not sufficient, and might not be sufficient now.
TYSON: Okay.
[LAUGHTER]
TYSON: So, what he’s saying is that it may still be there. We just haven’t found it yet. That’s a cop-out answer, I think, between you and me. But don’t tell him that I said that. Yes, Zohreh? Yes?
DAVOUDI: Yeah. So, exactly the line of investigation that we have in our favor is whether or not the simulation is imperfect, as you say. So, given that we haven’t already seen something doesn’t mean that we might not see something and, therefore, as I said, we look for evidences that tells us that there are some imperfection in the universe because the simulator or the amount of computation that could be done to generate our universe has been finite, it hasn’t been infinite and, therefore, there might be some evidence. But it doesn’t mean that the fact that we haven’t seen it doesn’t mean we shouldn’t go and look for it. It’s been difficult, but—
GATES: But I thought the question was why haven’t we seen it. That’s what I answered.
TYSON: Well, yeah. And don’t you both agree? You’re saying if you haven’t seen it, it doesn’t mean it’s not there. Keep looking.
GATES: Absolutely.
DAVOUDI: Right.
TYSON: Good. I’ve got a question from Twitter that came in. Evan Quinter from the Twitterverse asked- I think I’ll direct this to Max. “If the universe is a simulation, does that mean there’s a limit to how far our universe can reach?”
Because we speak of an infinite universe beyond our horizon all the time. So, you’re ready to say if it’s a simulation, it can’t be infinite, and there is the limit to the code.
TEGMARK: It’s a great question. If we are being simulated in one universe up on finite computational resources, yeah, then either the size of our universe is actually finite, or there’s some other trick like it just keeps repeating itself over and over again.
TYSON: Like the background in the Flintstones when they’re riding in the car?
TEGMARK: Precisely.
TYSON: The background just repeats. I was so angry. I said you can’t draw me a—what’s with your budget?
TEGMARK: That’s right.
TYSON: Have you ever seen the backdrop of cartoons when people are running? It just repeats. When I was a kid, that disturbed me.
CHALMERS: Maybe it’s a just-in-time simulation. It’s kind of like the Truman Show or something. They started off simulating me in Australia where I was born [unintelligible].
I came to New York, so suddenly they had to, like, simulate all of you guys and—
TYSON: For you?
CHALMERS: Yeah, exactly. Or for whoever. Or maybe they started off with the earth, and then we go to—the Voyager just reached—the thing just reached Pluto, so now they had to simulate Pluto.
TYSON: I see.
CHALMERS: Just like that. So—
RANDALL: Let’s not start with Pluto.
CHALMERS: Start small and go bigger.
TYSON: Yeah, don’t get me started on Pluto, first of all.
[LAUGHTER]
TYSON: But what you’re saying is they might just be laying down the bricks in the road as we drive along.
CHALMERS: Yeah. Just-in-time simulation, they call it.
TYSON: Just in time.
CHALMERS: Yeah. Simulate only as much as you need.
TYSON: Okay. Right here. Yes?
AUDIENCE MEMBER: Hi. So, we briefly discussed infinity and how it relates to this topic. But I was just wondering on the other end of the spectrum why is nothing not a thing? Is there not nothing when you come down to the fundamental question?
TYSON: We had an entire Asimov panel on that very subject. Where were you? The title of that was The Existence of Nothing, and we had all the experts on nothing on the stage at the time. I’m just saying.
[LAUGHTER]
TYSON: So, okay, maybe he didn’t know that, so—
RANDALL: I actually have a probabilistic argument.
TYSON: Okay. So, we’ll entertain it, but go online. The whole thing is there. It’s called The Existence of Nothing. Okay, yes?
RANDALL: So, in my book Dark Matter and the Dinosaurs I have a section on cosmology, and I actually talk about the probability of nothing. And I think nothing is just very unlikely.
I mean, first of all, we wouldn’t talk about nothing because we wouldn’t be here. But nothing is just one—if you think of a number line, zero is just one point on it, and nothing is just—I would say it’s very unlikely. And if you have an explanation of why there’s nothing, then there’s something there that allowed you to have the rules to explain it.
[LAUGHTER]
TYSON: So, you’re saying the act of posing the question of why there’s something is proof that there could not have been nothing?
RANDALL: Right. So, there’s two answers.
TYSON: Okay. Crystal clear now.
RANDALL: That’s one. And the other answer is probabilistic.
TYSON: Okay. Yes, sir?
AUDIENCE MEMBER: Hi, Neil. How are you?
TYSON: Hi. Good.
AUDIENCE MEMBER: So, I think I’m going to say this every day now after this. I’m going to say computer end program when I wake up in the morning. But the question is say we assume that we are in a simulation—we don’t try to prove it anymore—would it be possible to come up with equations, knowing what we know from the past, to predict what inputs might be in the future, assuming that this is an original idea and it’s not an input from the programmer and we’re not on a holodeck within a holodeck within a holodeck within a holodeck.
You think that would be possible and maybe escape the simulation?
TYSON: David, what do you have?
CHALMERS: Well, I think we’re probably stuck for now with the laws of the actual simulation of the simulated universe. If it’s a perfect simulation, we’re not going to be able to do better than that. We’re not going to get information about the character of the simulated universe.
Now, if it’s a buggy simulation, or if it’s interactive simulation—if they’re sending messages down here in the way that God was supposed to and so on—then all bets are off. All I can say is so far I’ve not seen evidence that we can use to make predictions, hey, tomorrow the simulators are going to call the whole thing off.
So, all we can do there is speculate as far as I can tell. If Zohreh’s work pans out, maybe we’ll suddenly have a lot more evidence.
DAVOUDI: And I would add that at the end of the day we are living in this universe. So, we are constrained by the laws of this universe. So, the concept of escaping from this universe doesn’t seem logical to me because we are bound to be evolving according to the laws of this universe, and not something beyond that.
TYSON: Plus, some other universes have slightly different laws of physics. I don’t want to be the first to visit them.
[LAUGHTER]
TYSON: Send something else then, and we’ll figure it out. We only have time for a couple more questions, one of which I’m going to take from our Twitter list.
So, let’s go right here. Sir, yes?
AUDIENCE MEMBER: Sorry, Neil, I tend to disagree with your conclusion about the universe simulation percentage, and more agree with Lisa’s zero percentage. Here’s why. Using your chessboard analogy, yes, there are 64 pieces on a chessboard. You can assign numbers to each piece.
TYSON: Sixty-four squares.
AUDIENCE MEMBER: Sixty-four squares, right. And you can assign values—numerical values to the chess pieces, and then use computers to see into the number applies: 3, 4, 5, and then run the simulation, and you get the end result of who’s going to win and so forth.
However, just look at the humans on this earth. We have seven billion people now on Earth, and even if we can mathematically model one person, when we have interactions of each other, we have interactions of more than each other because every action there’s a reaction, and then we have seven billion people.
So, the result is going to be totally unpredictable. In fact, it doesn’t look too good because the earth’s with its limited resources, and the earth population growing at a logarithmic rate.
TYSON: Exponential rate.
AUDIENCE MEMBER: The conclusion—exponential rate.
TYSON: Yeah.
AUDIENCE MEMBER: It’s going to be one conclusion, and that is—the result is not going to be good.
TYSON: But it may be that these multiple interactions that transcend our native ability to compute are no different from the Tick-Tack-Toe game being played by the five year old.
We’re just too stupid to know.
AUDIENCE MEMBER: Well, that’s true.
TYSON: Okay. That’s right.
AUDIENCE MEMBER: But, again, when you look at the complexity and try to mathematically model seven billion people’s characteristics—
TYSON: Yeah, we can’t do it because we just have human brains.
AUDIENCE MEMBER: Absolutely. Thank you very much.
TYSON: Yeah, sure. Forgive the rest of the people online. The last question is going to have to be from our Twitter stream here. And this one I’m just going to send to Lisa. Lisa, you will take us out with your answer to this question.
This is from Ashley Cannino, “Is dark matter”— There are multiple ways you could probably get to this, but let me say as it’s written: “Is dark matter transparent or simply rules of the game, a computational structure? Like an operating system.”
And think of dark matter and dark energy, these permeating elements of the universe that we don’t understand at all. We don’t know what’s causing them, but we can measure their existence. Could that be the blood of the operating system throughout the universe?
RANDALL: So, it’s an interesting question, and people have asked that question. And you have to take a little bit of an Occam’s razor approach here. So, first of all, dark matter is indeed transparent matter. It’s matter that just light just goes through.
There’s evidence for it not because we see it, because it doesn’t emit or absorb light, but because it has gravitational influence. And we can observe the gravitational influence. Now, you can ask is that because we got the laws of physics wrong and there really wasn’t matter, and the we got the laws of physics wrong.
First of all, it’s a lot simpler to believe this matter that we have no reason to believe shouldn’t be there is there, than to think we got the laws of physics wrong. Because the laws of physics work incredibly well over many distant scales. So, there’s no evidence that those laws of physics are wrong. But, furthermore, there’s actually—there’s more and more evidence that makes it look just like it’s matter.
One of the things is known as the bullet cluster, or other clusters, which are really mergers of clusters of galaxies. Clusters of galaxies are bound states of many galaxies held together.
And when those things go through each other, a cluster of galaxies has gas in it and it stores in as dark matter. When it goes through it, you see the gas get stuck in the middle. You can see that through X-rays. Through gravitational lensing, you can see the dark matter just pass through. It acts just like you would expect matter that’s not interacting to act. It goes right through, the gas stays in the middle. Now, you can try to mock up equations, or some simulation or something that does that, but it looks just like matter would look. It’s exactly what you would predict.
TYSON: Except we established earlier that we don’t form our own galaxies without the existence of dark matter creating the womb in which we collect. So, isn’t that kind of like an operating system, enabling matter to do its thing?
RANDALL: Well, gravity is, in some sense, the operating system. But gravity is responding to the existence of the dark matter. And dark matter does play a big role. There’s more of it. It collapses. It doesn’t attract with light, so it can form structures more easily than normal matter.
TYSON: I like that. Gravity’s the operating system. Well, how about dark energy?
RANDALL: Dark energy is another just thing that’s in there that’s responding to gravity. So, dark energy is—for those who don’t know—smoothly distributed. It’s not like matter that clumps together.
TYSON: It’s the operating system.
RANDALL: It’s not the operating system either.
TYSON: Operating system is everywhere you touch on a computer. So is dark energy.
[LAUGHTER]
RANDALL: I’m just getting confused now. It’s part of what I put in, at least in my initial state, and then I let the gravity equations work on it.
So, you have this distribution of energy. You have this distribution of matter. And then you can ask what is the effect of this energy that we don’t observe directly. In some sense, we observe the fact that it is responsible for the acceleration of the expanse of the universe.
But gravity is the only law there. The other stuff is just stuff. Dark energy is stuff. Dark matter is stuff. The gravitational equations are acting on that, and it’s actually creating the gravitational force.
TYSON: Well, so if gravity is the operating system of the universe, I can’t wait for Universe 2.0.
[LAUGHTER]
TYSON: Thank you all for coming this evening. Thank the panel.
[APPLAUSE]
TYSON: Zohreh, Max, Jim, Lisa, David.
[APPLAUSE]
TEGMARK: It was so much fun.
DAVOUDI: Thank you.
TEGMARK: That was really—
TYSON: Thanks for coming out. That concludes the 17th Annual Isaac Asimov Panel Debate. Good evening to everyone watching live stream. Goodnight to you all here in New York. We’ll see you next year.
[Credits roll.]
Neil deGrasse Tyson hosts and moderates a panel of experts in a lively discussion about the merits and shortcomings of a provocative and revolutionary idea—that perhaps the universe as we know it is a computer simulation. While the debate may have started as a science fiction speculation, it has become a serious line of investigation among physicists, astrophysicists, and philosophers.
This debate took place at the Museum on April 5, 2016.