Into the Impossible With Brian Keating - David Chalmers: The Matrix is REAL! (#370)

Episode Date: November 23, 2023

Is the Matrix really real? And if so, which pill would David Chalmers take?  Join us for a mind-bending journey through virtual worlds, human consciousness, technology, philosophy, and religion, and find out!  David Chalmers is an Australian philosopher and cognitive scientist specializing in the areas of philosophy of mind and philosophy of language. He is a Professor of Philosophy and Neural Science at New York University and co-director of NYU's Center for Mind, Brain, and Consciousness (along with Ned Block).  Key Takeaways:  Intro (00:00) Judging the book by its cover (01:49) Defining the hard problem of consciousness (04:50) Why is there tension between physics and philosophy (07:20) The Chalmers equation for the simulation hypothesis? (11:57) What have we learned about consciousness from computer analogs? (21:24) AI physicists (32:08) What is it from bit? Or is that bit from it? (40:10) What does it take for the maker of the simulation to be god or godlike? (47:50)  Why create the world (or the simulation)? (53:49) Can there be a substrate-free creator (simulator)? (57:43)  Why do you claim we can't know if we are in a simulation? (1:06:32) Would you take the red pill? (1:08:41)  Outro (1:16:43) — Additional resources:  📢 Ownership of your health starts with AG1. Try AG1 and get a FREE 1-year supply of Vitamin D3K2 and 5 FREE AG1 Travel Packs with your first purchase 👉 https://drinkag1.com/impossible 📚 Reality+: Virtual Worlds and the Problems of Philosophy by David Chalmers: https://a.co/d/3pzidPf ➡️ Follow me on your fav platforms: ✖️ Twitter: https://twitter.com/DrBrianKeating  🔔 YouTube: https://www.youtube.com/DrBrianKeating?sub_confirmation=1  📝 Join my mailing list: https://briankeating.com/mailing_list  ✍️ Check out my blog: https://briankeating.com/blog.php  🎙️ Follow my podcast: https://briankeating.com/podcast  — Into the Impossible with Brian Keating is a podcast dedicated to all those who want to explore the universe within and beyond the known. Make sure to follow so you never miss an episode! Learn more about your ad choices. Visit megaphone.fm/adchoices

I think the human brain is a big complex machine. It's an amazing machine. It's a creative machine, but it's still made up of these neurons, which appear to be computational units, hooked up in amazing ways. The imagination, as far as we can tell, is itself a kind of simulation. When we imagine things, when Einstein was imagining things in free fall, he was running a kind of simulation himself, running on this incredible computer,

which is far more sophisticated than any computer that we would. computer that we ourselves have developed to date. But you know, AI is moving fast. I can't believe where it is now compared to where it was 10 years ago. Any sufficiently advanced technology is indistinguishable from magic. Open the pod bay doors. I am your fearful host, Professor Brian Keating of the Arthur C. Clark Center for Human Imagination talking with a renowned intellect. And this is a real treat when I found out that he has new book was coming out, I just had to get him. And it is none other than Professor David Chalmers, who first formulated the problem of the hard problem of consciousness, which we're going to get

to in a paper facing up to the problem of consciousness. And way back in 1995, expanded upon it in his wonderful book, The Conscious Mind, 96. His works are provocative, influential, and some of the greatest luminaries of all time, including past guest Stephen Pinker, have called it so praiseworthy and his acclaim and renown, no, no bounds. And so first, I want to welcome you today, Dave. How are you doing today all the way in New York? Thanks, Brian. Yeah, great to be here.

It's kind of rainy out here, but I hope it's sunnier where you are. Yeah, La Jolla, that's a fair game. You know, I usually say, you know, the hardest job in the world is being a San Diego sportscaster because we've never won a championship in any sport. But the easiest job is being San Diego's meteorologist because 72 and sunny, 70, one and a half and so, you know, it's, it makes up for its lack of variance by its consistency of beauty. So yes, it is quite lovely. But I am a native New Yorker. I'm a born and bred in New York, so sometimes the accent will come out. And Dave, as I told you, right before we started,

my audience loves it when I play the game, which we call judging books by their covers and titles and subtitles. So this book has a butterfly, very provocative and beautiful butterfly, which is kind of the only character that makes its way through a dream sequences throughout the book. The title is Reality Plus, the subtitles, virtual worlds, and the problems of philosophy. Dave, explain to us what is the title, the meaning of the title, and that mischievous butterfly doing on it. The title is Reality Plus. Actually, I started, I had a working title for a long time that was Reality 2.0. because this is about, you know, the book is in large part about virtual and artificial realities,

these could be the second class of realities. First reality, physical reality, but then we start creating our own artificial realities. That's reality 2.0. Another idea is it could turn out that we ourselves might be in simulated artificial reality, so that our reality is reality 2.0. Now, the only trouble is there's a fatal flaw with this title, which has probably occurred to all of your listeners already, which is reality 2.0 suggests the people, AOL, 1990s, you've got mail.

Dial-up, yeah. It's so retro that it sounds like it's a step backwards rather than a step forward. Even our colleague, Max Tegmark, went to life 3.0 to avoid this problem. I thought, reality 4.0, maybe not. But someone had the great idea of, yeah, well, reality plus. Because although putting a plus on the end of a title may be becoming a ubiquitous cliche too, at least in streaming services, you know, Disney Plus, Apple TV Plus, Paramount Plus, and so on. At least it's a 2020s cliche and not a 1990s cliche. And it kind of

suggest this idea of there's more to reality than you think. There's physical reality, there's virtual reality, there's simulated realities. And yeah, if in the future if somebody comes up with a reality streaming service where you're able to choose between the virtual realities than you want to live in, then Reality Plus might be an appropriate name for it. So in the end, I thought, okay, got to go with Reality Plus, got to go with the plus sign. And that gives a hint at the enhancement that we're going to be experiencing in the maybe augmentation and supplementation, but maybe even full-scale replacement of our notions of reality. And I've had on, as I said, Don Hoffman, believes reality does not exist. And it's sort of an avatarish scenario, a desktop

scenario, which is provocative. And you refer to him towards the end of the book. We'll maybe get a chance to get into that. And neglected to mention that you are the co-director of NYU's Center for Mind and Brian consciousness. Oh, no, brain consciousness. My mom named me, you know, Brian so that people would make that mistake. And it happens about two times a day. And you were elected a fellow of both the American Academy of Arts and Sciences, but in your native land down under of Australia. You were elected the Australian Academy of Humanities as well. And I want to start by, because I can't resist. It's like if we were to hear go to a concert by your countryman, ACDC, and they don't play. you shook me all night long.

You're going to leave unsatisfied. I have to take it from the master. I want you to define the hard problem of consciousness. Yeah, so the hard problem of consciousness is the problem of explaining how physical processes in the brain give rise to subjective experience, to the subjective experience of the mind and the world. So for me, the core of consciousness, that word means many things to many people. but the core of consciousness is subjective experience of perceiving, of feeling, of thinking, of acting.

It's anything that feels like something from the inside. And the easy problems of consciousness are things like explaining behaviors which are associated. How can we walk? How can we talk? I get a stimulus. I can point to it. I can act on it. I can report it.

I can say, yeah, there's a red object over there. we got a beat on how to explain those things in terms of physical processes in the brain. Specify a mechanism, show how it does the job, but those were the easy problems. The hard problem is why is all that accompanied by subjective experience? Why doesn't it all go on in the dark without any subjective experience at all? I mean, it doesn't. At least in me, I know that I'm conscious. I experience all this.

And I assume it does for you as well. But why does it? Right now, we don't know. That's a hard problem. Yeah, and you refer to work done by Thomas Nagel and others. And I want to write a book, you know, my dream, Dave, maybe with your help, I want to write a book. What does it like to be Thomas Nagel? And the author would be A-BAT, because, you know, I think a lot of these things, you know, to physicists tend to be a little bit unsatisfying.

In other words, we kind of really tie into the physical, the Boltzman brain aspects. We'll get into that, the It from Bit. And I think the more interesting, you know, a bit from it. But, you know, how do we understand at a core grasp, you know, and reconcile things that are kind of this, yeah, subjective. And I want to maybe begin the core conversation by talking about this kind of maybe fitful tension between physicists and philosophers. So a famous quote by Galileo, when he invented not the telescope, but he perfected the Galilean. telescope, here's an example of one. He didn't invent it, but he used it for the first time astronomically. And what he said he did with it is resolve questions that had vexed philosophers

for many generations and caused them unending suffering. In other words, he was kind of like pejoratively putting down philosophers. And nowadays, we see it in people like Lawrence Kraus and others and talking about the utility or lack thereof. Why is there a tension between physics and philosophy? And do philosophers have antipathy towards those of my ilk physicists? I don't think so, and I don't think there has to be attention here at all. I mean, the great physicists of the past, many of them were great philosophers. You know, Isaac Newton considered himself a philosopher. He was professor of natural philosophy. But I mean, along the way, and when he was born, the problems of space and time were problems for a philosopher.

Now, Newton was a good enough philosopher that he managed to make progress on this problem with new methods, formal methods, experimental methods, that would basically turn this problem of philosophy into something we could make progress on and thereby kind of birthed the core of the science of physics. And this happens in philosophy all along. So many times, you know, disciplines like economics, psychology, linguistics. parts of logic kind of started as philosophy. Some philosopher made progress on it. And then we spin off a science. Okay, so now what's happened is that a number of these, a lot of philosophers have been successful enough

that we now have the successful spinoff going in physics, in psychology, in linguistics, and so on. And what's left in philosophy, almost by nature, is the too hard basket, the stuff that we haven't figured out how to turn into a science yet. But on the other hand, there's quite a lot of things which are just to the interface. And I would like to think the science of consciousness is precisely one of those, starting with a very core

philosophical question. How could there be consciousness in a physical universe? And we're managing to at least bite bits and pieces of it off. So we now have a thriving science of consciousness that involves neuroscientists, psychologists, and also philosophers right around that borderline. And there are still bits of the science of the study of consciousness, which are in the philosophy camp, but bits in the science camp. And actually my experience is I've had a lot of great experience interacting with neuroscientists, psychologists. Right now I've got a big project with a number of neuroscientists where we're trying to design experiments to test some of the leading philosophical theories of consciousness and come up with experiments that will do that.

So historically, look at all the physicists who were great philosophers. Heisenberg, Schrodinger, Bohr, Einstein, early, first half of the 20th century, they were all great philosophers. One thing that happens over time is that things get a bit more professionalized. Maybe physicists have to focus a little bit more these days and some philosophers focus a bit more. So there's less room for maybe there's a bit less of that kind of crossover. But still there's so many physicists who are brilliant philosophers. too, you know, John Wheeler was a brilliant philosopher. Look now, someone like Max Tegmark wrote the mathematical universe.

That's a great work of philosophy. People like David Deutsch who are at the interface between the two. Philosophers like David Albert and Tim Maudlin, who know physics forward and backwards. So I still think there's a real productive corner. Yeah, I would agree. And I mean, if they look at the root of the words, meaning, you know, love of wisdom, love of knowledge, I think that's something that physicists should aspire to. Of course, yeah, as you say, we have to focus, we have to niche down,

we have to choose a profession and so forth. So typically those questions, I call them the 3 a.m. Dorm room couch questions, you know, those kind of dissipate as we mature. And then people look back with somewhat disdain, you know, or just like, oh, you know, David, the first exposure I had in college to philosophy was, you know, Philosophy 101, my first day of freshman year at Case Western Reserve University. and then the professor was this guy who looked like, you know, the professor in Welcome Back Cotter. His name, I think, was even Cotter or something like that.

And he'd never taught before. And we would just have these, you know, infinitely long exams on a quarter, you know, on a cut by monthly basis. And they were all true false. It was like, you know, the categorical imperative is something that can subsume the ontological. It's true. I don't know. And I would get less than a 50% on each quick. I don't understand how that's possible.

It was only 50-50 choice. Hello, Students of the Impossible. It's Professor Brian Keating here with just a tiny little homework assignment to interrupt your podcast. And that's to make sure that you're subscribed to the podcast or following us on your podcast app of choice. Get some research and actually only about 50% of you are actually following or subscribing to The Impossible podcast.

And really mean a lot if you could subscribe. And keep up to date with me and with all the greatest content. I'm putting out tremendous amounts. Podcast has grown in popularity, but it can be better and bigger with your help. Do that. Please do it now. Don't wait. You'll forget.

If you're looking to really boost your position on the grade curve for some extra credit, make sure to leave a rating or review of the podcast. Really helps. Thanks a lot. Now back to the show. But to recapitulate one thing you said is that, I mean, some of these things are coming back and really providing a sound basis for physics research. In other words, they're encountering aspects of the multiverse, of alien worlds in this book,

certainly of the substrate dependence or lack thereof, all topics we're going to get into. And I want to start with this kind of really amusing section in the book where you go through kind of what I call the, I'm going to call, I'm going to hope this trends, but I'm going to call it Chalmers equation. And it's kind of the Drake equation, but for the probability of simulation to exist and virtual worlds to exist. and you go through all the different kind of steps in exactly the way that Frank Drake, 60 years ago, this year, came up with his eponymous Drake equation, which is to settle on the number of, you know, alien civilizations that could potentially have extraterrestrial intelligence. And of course, that's still an open question.

You know, Paul Davies, who was on the show recently, wrote a book called The Erie Silence, you know, the Fermi paradox. Why don't we see them? I wondered, was that kind of an inspiration for the, you know, you don't call it the Chalmersy equation, but it's sort of of a same thing. is it not? Yeah, it is kind of a similar statistical equation here, I guess, in thinking about the number of simulations. I should say that here I got a lot of inspiration from the philosopher, Nick Bostrom, who's thought about the simulation himself.

And he comes up with his own equation for what fraction of beings in the universe we should expect will be simulated. And he ties it to some complicated things like the number of beings who become post-human and so on. I don't fully agree with how Ostrom does it, which is why I come up with my own way of coming at it. But the equation takes a fairly similar form. First, we need to take the probability.

There's this background assumption, which is, you know, the question is, if you had to simulate how many universes, how many simulated beings are going to have consciousness just like mine in the history of the universe? Well, one assumption there is that simulated beings can be conscious at all. I don't know. How likely is that?

Not everyone is going to grant that. But I say, okay, let's grant that. I think it's at least 50% probability that simulated beings can be conscious at all. Put your own probability there. Then conditional on simulated conscious beings being possible, what I call simulated human-like conscious beings being possible, beings with experience like us. what is the probability that there will be many of them, that they will greatly outnumber humans. And well, again, that depends on some assumptions.

It depends on beings getting to the point where they can have the capacity to build these beings and that they will choose to build them. But I say, okay, let's give that at least whatever probability you get for that. I say, let's give it at least 50%. Multiply those out, you then get a 25% chance that most beings with conscious experiences like mine are themselves simulated being, and then running the probabilities a little bit further, you basically get just under 25% chance that I'm in myself in a simulation.

Okay, that's just a back of the envelope case, but it's a back of the envelope case for 25% chance, we're in a simulation. If you wanted to be more conservative, you could dial some of those back to 1 in 10, and you still get a 1 in 100 chance, And that's at least interesting. Yeah, as I always say, though, I gave a talk on, you know, maybe enemy territory.

I spoke at the SETI Institute six years ago now, maybe, and I laid out a case for, you know, my complaints against the Drake equation. And this is at the institute that Frank Drake and Jill Tar, you know, built. So I had to tread kind of carefully. But I said, you know, to scientists like myself, the number is interesting, but the error bars are much more important. Because the error bars subsume all your ignorance, all your uncertainties, all the most. model dependencies, all the calibration biases, et cetera, et cetera. And I went through an example where I calculated, let's do the Drake equation for how many

people are in the world famous San Diego Zoo down the road for me at this very moment. I go through the calculation. People can see it on YouTube at SETI Institute's website, our YouTube channel. And I come up with a number, and I come up with a number of people there. It's like 8,000. But the uncertainty was plus or minus 12,000. There could be negative people. There could be holes and virtual people, you know, and kind of, if,

suffusing the void there. So the point being that the error bars are, you know, in some case, more important than that number itself. So what would you ascribe the error bar on either the 25% or the 100% or the 1%? You know, some people say it's even much higher than either one of those two numbers, you know, the 25%. So talk about the errors in that calculation, which is what really matters to scientists in the end. Yeah. And in fact, in this case, the error bars in a way play in my favor. Because the conclusion I want to argue for here is that we don't know that we're not in a simulation, that we can't rule it out. So to the extent that there's a whole lot of uncertainty here, well, that just increases the likelihood that, yeah, we don't know.

There are possibilities here. We can't exclude. But I agree that there's all kinds of uncertainty here. One source of uncertainty is just on the consciousness side, right? We don't understand consciousness, so we don't know whether a simulated being could be conscious at all. I go 50-50, but someone might say, come on, well, I believe that. Maybe that's under 10%. Another source of uncertainty is whether simulated universes will be possible. Will the laws of physics ultimately be computational? Of course, Roger Penrose thinks that the laws of physics are not computable. So we'll never have a simulation of the universe, at least on a classical computer. At this point, I'd like to think that even if a classical computer can't do it, can't do it. Just say, you know, Penrose is a

right that the true theory of quantum gravity will have some non-computable element in it. I'd like to think that whatever that element is, we could then, you know, we could then capture that element and use that to build new computers, quantum gravity computers that would thereby exploit this non-computable element and we could get universe simulations that way. But okay, there's a big source of uncertainty there. And the third is maybe tied to the sociology, tied to the future. Is it the case that we will actually have beings, even though these simulations could exist, is it the case we'll get to the point where we have beings who can build them and do build them? Maybe all the beings of the future will be self-protective enough, maybe,

that they choose not to create advanced, artificial general intelligence, thinking that this is just too dangerous. Or maybe they'll be ethical enough to say, man, we can't create these simulated universes. that would be playing God. So, yeah, maybe this sociological part is probably the most uncertain of them all. I mean, I'm just thinking that naively these populations will be gung-ho. Someone's going to do it eventually. It's going to be useful for all these purposes, help predict the future, help understand the world.

Scientists love to run simulations now. So much science is simulation-based. Once we can simulate the whole universe, of course scientists are going to want to run some universe simulations to see what happens. But who knows? Maybe the institutional review boards will come in with the grants will become. It's peak pollination season and my business is scaling fast. To keep the nectar flowing, I need a phone plan with top priority data speeds. That's why I chose GoogleFi wireless. My connections stay strong even when the hive is buzzing. Plus, unlimited plans started $35 a month. Now that's a deal that doesn't stay. Explore GoogleFi wireless plans today. Plus taxes

government fees. GoogleFi Wireless is not subject to data traffic deprioritization during times of high network usage. You only get a very good, not an excellent, you know, from the Galactic Science Foundation. Right. But in the case of the Drake equation, what a lot of people point to is excessive things like the Kepler mission and calculating the number and observing actually real data, unlike, you know, some branches of really theoretical physics, like string theory or whatever, or very little data actually exists, and it's all returidictive and not predictive. But

in any case, I'm not going to go off on a polemic about string theory right now. But on the other hand, we've reduced some of these terms in the Drake equation to basically what I call the only thing that's remaining are the sociological terms in the lifetime. You go through this really

and it's just such a great book. I can't recommend it highly enough, especially for my fellow physicist and those nerds like me out there, because Dave goes through in great detail, but it's so entertaining this notion of, well, what is the brain viewed as a supercomputer? And, you know, we had this notion that the brain, the human brain is so far surpassing. It's got more neurons and connections between neurons than there are galaxies in the observed bowl of the universe,

et cetera, et cetera. And yet, you make a very convincing calculation that the human brain is sort of equivalent to a 10-pet-flop computer right now. And there are computers that we use to analyze data from the universe, the cosmic microwave battery back, background owned by the Department of Energy that are 10 petaflop computers, you know, or, you know, that scale. And that's, what, 60 years after general, you know, Turing machines were kind of first conjectured by Alan Turing and so forth by Neumann and others. So in such a short amount of time, we're already surpassing in raw computation power. And that's one of the terms, I think, that would have to go into a Chalmers equation for calculating a number

of similes. So what other things remain? As I said, the sociological term, the lifetime of of intelligence civilizations before they blow themselves apart, create some runaway AI, create some virus or whatever. That's the limiting factor in our understanding and the maximum contributed to our error bars. What is in the Chalmers equation, again, you don't call it that, but I can call it that. What is the limiting factor? And what have we learned most recently, you know, from analogs of, you know, consciousness, Kepler, or so forth, like what are the current experimental and theoretical computer

science or sociological terms that have shrunk the error bars in that equation. That's interesting. Yeah, I guess on the first factor, there's like as simulated consciousness possible, a simulated physics possible, and the limiting factors there are going to be tied to our knowledge of consciousness and of physics. On the sociological side, there's going to be, will we get to the point where we can make these and will we choose to make them? But maybe, yeah, the technology side is interesting.

I mean, the exact number of petaflops required to simulate a brain or a universe. I mean, in certain respects, you might say the simulation argument doesn't need overly strong arguments here. Just say it's a vast amount that's required. Because we don't need to hypothesize that the universe in which a simulation is run is the same as ours. It seems actually very likely it might be quite different. Might be quite different from ours. And, I mean, obviously, if the universe is finite, there's a pretty good reason to think no finite universe can run a perfect simulation of itself. But some people say, ah, therefore the simulation hypothesis is bunk.

But of course, the relevant hypothesis is the complex universes can run simulations of simpler universes or parts of themselves and so on. We also should hold open the possibility that the simulating universe is infinite. That's quite different laws of physics from ours, perhaps. and that is somehow trivial in these infinite universes with infinite resources to create universes with finite resources such as our own appears to be, at least in the known universe. So from that perspective, it's like even

if those resource bounds don't hold, then all we need to do is move to say, for example, the infinite simulation hypothesis, and we should still give some probability to that. But it is interesting, actually, simulation technology has been moving along very, very, very fast in recent years. People actually build cosmic simulations.

And yeah, every few years, the cosmic simulations get a whole lot more detailed and more fine-grained. And yeah, they're multi-scale simulations are becoming more and more and more of a thing. And then from the bottom up, it's like, well, virtual reality technology is also developing. For me, that's very relevant that actually we now have these simulations that we can enter and experience immersively, those are not yet, VR is not yet a copy of the physical world, but if VR just gets better and better for 50 years or 100 years, we'll probably

have the version of VR, which is indistinguishable from the physical world. Once we have that technology in front of us, we will actually be able to put people into simulations that feel like their simulations of physical reality. At that point, no one's going to be able to say, hey, this is just way out, this is science fiction, we're actually going to have the technology in front of us. It'll be happening to some people. And we can then raise the question. Maybe just that is happening to us. Now, some people say, oh, isn't the simulation hypothesis meaningless because it's unverifiable or undetectable?

Once we can actually put people into these simulations, no one can say it's meaningless anymore. It's like we've got it right here. I've got it right here. So I'm going to now strap on my Oculus. You have one there too, Dave? We can enter into VR. You did a VR interview with Cassandra Vittian here in UC San Diego for the Arthur C. Clark Center. And Arthur C. Clark makes an appearance in your book as well, one of his books, Childhood End, I believe, you quote from, which presage is an ocean of VR and computer simulation and so forth.

But now this isn't super realistic, you know, when I put these on. In fact, I've tried to use it for, you know, to do exercise. eyes and then you get all these cookies and I tried to use it for meditation and it's like blindingly bright and I'm taking you have a more recent version than I do. Mine is the Oculus. It's just the Oculus

Go I think from three years ago which is totally outdated. You have the quest? Yeah, I mean they're fun. My kids there's a fundamental problem with them and that you know that you're wearing this giant thing and as we know

now I don't know if you've checked your IRA or whatever, but Facebook, which is now known as meta, you know, for the Metaverse is cratering and going, you know, going down, which is making a lot of people really happy. But that's sort of a virtual where you could put this on. You can talk to your Facebook friends, maybe, but it's so cumbersome. And even if you use your voice, it's kind of cumbersome. Now, I mentioned Galileo before. It turns out Galileo may have been the first person to invent a VR or maybe an AR headset because he, there was the problem of computing time, as you know, in the 16 and 1700s, and there was this longitude

prize that was offered for the first kind of reliable clock. And Galileo said, well, hey, wait a second, I've got this telescope, and I can see the moons of Jupiter, and they're as periodic as you would like them to be. And in fact, you can see them, you know, all over the earth. And so he invented this helmet with two telescope or one telescope attached, called like a chronoscope or something like that. And he used it to, you know, trying to claim the longitude wise. And of course, it was rejected because people's, oh, you can only see the moons, you know, nine months out of the year. And what do you do those other months? He's like, well, you could, you know, calculate tables and interpolate. But, you know, back then, if you look at,

well, how much has really changed in augmented reality. Now I can, you know, go to IKEA's website. I can take a picture of my, you know, carpet and then it'll suggest it how a chair will look. It's kind of being used as Facebook's using it for, like, advertising. And I wonder when we scale up these digital computers, as I understand it. Moore's Law starts to deteriorate because as the power of the computer grows, and it does grow at this exponential rate, the number of use cases, the number of users and the demands that they're putting on it are always scratching, you know, the upper bound of the envelope. So in other words, you're almost like the utility of them becomes so much higher in addition

to their raw power. And it's growing at a rate that's, you know, sort of causing their bounds to saturate when it comes to the actual number of, say, output. And you could distill that into any term, not just pedophlops, but into how many papers get ridden, how many, you know, full simulations come to conclusion. And those are actually tapering off for these large supercomputers. And so we have to keep building more and more so that the plateau gets higher and higher. Can you see that as a term in the Chalmers equation that would eventually limit the, you know, kind of optimistic prospects of a, you know, world in which we are already living in such a simulation?

Especially if we're talking about, yeah, building ultimate full simulations of the human brain, full ground level simulations of physics. Yeah, in the book, I give some calculations that rely on something like Moore's Law continuing. So, yeah, every decade we get this multiplication factor. And given that, it's not too hard to say, okay, I'm a philosopher. I take the long view. I don't care what happens in the next 10 years. I care what happens eventually. if it's 100 years, 200 years, I'm fine, as long as it continues.

But yeah, if it asymptotes at some point, then obviously there are serious potential limitations there. I mean, I kind of hope then at a certain point we get to harness the power of the sun and do amazing things that we couldn't do before. But, yeah, if it turns out there are principled limits, then, I mean, we know their current physics has principled limits like the speed of light and the plank scale and so on. So either we eventually, if we eventually hit them, then I think we need to get more creative on the software side, so to speak. If we hit ultimate hardware limitations, then we do things on the software side. For example, we find speed ups and shortcuts in our simulation as people do already. Maybe people will find ways where, okay, we're only going to simulate local parts of the universe. Okay, simulating the whole universe out of reach.

Let's just simulate the solar system in detail, and everything else is just kind of a sketchy background coffee. It'll obey a few basic principles of cosmology and not too much. Hey, that could have been what messed up those experiments a few years ago, right? Exactly. In addition to the dust that you talk about at the end of the book, we'll get into that. So you mentioned, Sir Roger Penrose, his three or four-time guest, Into the Impossible. And, yeah, you're right. He does talk about the limits of computability and his apocle.

book, The Emperor's New Mine, which was the first science book I ever read. It was kind of a treat to have him endorse my first book way back when before he won a Nobel Prize. And it's quite interesting. And he and I have talked, and others have talked about this notion of AI physicists. So, you know, what is the limitation of, and our friend Max Tagmark, who's also a guest and friend? You know, he is very sanguine about AI Feynman, AI Galileo, AI. But I always remind folks. I say, you know, what, I don't know if you're familiar with this, Dave, but Einstein called the happiest thought of his life. And I've got a little Einstein, a virtual Einstein. Here he is. My audience always loves when I break out the virtual

areas. So, yeah, I only have one of these that has actually been on the show, and that's Noam Chomsky, but maybe someday there'll be a Chalmers Finger Puppet from the Unemployed Philosopher's Guild. This is called the Unemployed Philosopher's Guild. Yeah, who knew that Einstein was a So Einstein was speaking about the equivalence principle, which undergirds all of general relativity. And he said, he described what he called the happiest thought of his life. And that was that an observer, you know, freely falling in space would experience no gravitational field. Meaning that, you know, gravity and acceleration are, you know, one and the same. And that geodesics are manifestations of the shortest paths in a curvature-free space time in there.

And when you add in curvature, due to gravitation. mass, then it changes and alters the perception of both spatial intervals and time. But I posit to you that I don't believe it's possible for an AI Einstein to exist. Because, I mean, first of all, how do you visualize the notion of free fall as a silicon-based entity? And then, two, what does it mean to be happy? Like, what does an AI mean when he or she or it or Z or whatever feels happy? So isn't that, kind of a counter-proof that we could ever come up with a computer, a simulator, could come up with such laws that the human brain has done over and over again. I think the human brain is a big, complex machine.

It's an amazing machine. It's a creative machine, but it's still made up of these neurons, which appear to be computational units hooked up in amazing ways that can learn in amazing ways. The imagination, as far as we can tell, is itself a kind of simulation. When we imagine things, when Einstein was imagining things in free fall, he was running a kind of simulation himself running on this incredible computer, which is far more sophisticated than any computer that we ourselves have developed to date. But, you know, AI is moving fast.

I can't believe where it is now compared to where it was 10 years ago. I did my PhD in an AI lab in the 1990s with Doug Hofstad. And back then, people used to say, a year spent working in AI, a year spent working in AI is enough to make you believe in God. Because AI is so hard. It's so hard to get machines to do human level sophisticated things. But suddenly, in the last 10 years, yeah, machines doing image recognition, they're doing speech recognition, they're doing speech generation, they're doing game playing, they're doing navigation, they're writing their own code. there suddenly all these hurdles are, hurdles are falling. And yeah, I mean, we don't remotely have an AI Einstein yet,

but I'm not going to be the one who bets against it in 20 or 30 years time. And yeah, we do have emotion is certainly a limiting factor in AI right now. But we do have computational models of emotion, at least treating happiness and sadness as forms of valence. And I don't think anyone would claim that we yet have AI systems of experience genuine emotion, but I don't see the principle limit towards AI's ultimately. There are what they call affective computing as a field, and people are working on this. Machines can be made to behave in certain ways, which at least seem to reflect irritation,

satisfaction, frustration. Are they really experiencing it? Well, that's a question about consciousness. Right. So, my late father used to say, well, if you want to simulate, you know, pleasure and pain response, which is some psychologists and social philosophers claim that's, you know, all we're responding to is conscious entities. He used to, my father used to say, you know, well, just like when it does something wrong, you know, you blow a transistor or you, you know, you collapse a wave

function. If it's a quantum, you know, do something that is a painful, you know, cost function for that computer, it's still not clear how to, you know, simulate, you know, happiness and freefall. Is it really the sum totality of only pleasure with no pain? I mean, obviously we, can have super positions, you know, right now I'm really happy, but I could be, you know, unhappy very easily, and I could be much happier very easily, too. So it's not, is it clear that I have to be in the ultimate state of happiness to experience it? You mentioned God, you mentioned, you know, believing in God is kind of an off-off put, but I do feel like, you know, it's pretty remarkable that these, you know, Bronze Age, you know, itinerant, Semitic Wanderers, you know,

three thousand years ago, you know, came up with a notion of, you know, pleasure, pain, experience, reality, free will, commandments, moral and ethical imperatives, you know, 3,000 years ago, were they creating the earliest version of the Sim hypothesis? I don't know how far back some of these simulation ideas go, but you can find elements of it, ancient Chinese philosophy, the Duongja and his butterfly. How do I know I'm not a butterfly dreaming that he's Duangir? That is where, by the way, that's where the butterfly on the cover came from. That is meant to be a virtual reality version of Duang's butterfly.

I could be a butterfly dreaming he's Duangri. You find an Indian philosophy, ideas about illusion. So many Hindu tradition has God's mind supporting all of reality. And maybe this also connects to what you find in the Abrahamic traditions. Where, yeah, basically God is so all-powerful. God supports God supports reality. Of course, the question is,

how did God create reality? I guess I always remember let there be light and there was light. In the book, I speculated a bit about, maybe what actually went on was let there be bits. And God creates some bits.

No, actually, exactly. So in the Hebrew tradition and the Talmudic tradition that I'm more familiar with, what God does in the first six days of creation is separate. So the second line after let there be light or information, and everyone will say it, is, and the world was chaos and void, and he created order. So in other words, there's chaos, which could be a state, like a one, and there's order,

which would be lower entropy, it could be a zero. And he separates between man and animal, between vegetation, water and land. And so it's not at all a stretch to hypothesize that that is, yeah, there you go. Yeah, show that. That's... I got a great illustration of this in the book. On one side we have, there is a traditional god creating the universe by creating some bits. And then the tables and chairs and animals and everything else come from there. On the other side, we have a simulator, a simulator god who here is like a teenage girl on the next universe up. She's creating reality by programming her computer to create bits that create the trees and dolphins.

I mean, a shannon, you know, stipulated. Your summer starts now with Memorial Day deals at the Home Depot. It's time to fire up summer cookouts with the next grill, four-burner gas grill, on special buy for only $199. And entertain all season with the Hampton Bay West Grove seven-piece outdoor dining set for only $499. This Memorial Day get low prices guaranteed at the Home Depot. While supplies last, price invalid May 14th or May 27th, U.S. only exclusions apply.

See Home Depot.com slash price match for details. I believe, yeah. You only need zeros and ones. And you talk about, there's a Leibniz who also, yeah, came up with the primitive Boolean construction. So, yeah,

you only need these, you know, fundamental units, chaos and order, if you like. And so it's not altogether, you know, impossible,

even to be in, you know, some sort of resonance with a traditional religious, you know, conception. And so I saw a funny thing

in one of these, like, onion-type magazines. It was like, you know, scientists praise God for creating an ultra-realistic metaverse that he calls the universe. But I want to ask you, there's a very, very provocative

middle section in the book where you start talking, you basically make the claim that a creator plus it from bit equals a simulation. And yet, you still personally don't believe that such a God is worthy of worship, nor do you worship an ordinary Abrahamic or other faith creator. You also don't believe that is necessary or worthy of worship potentially. But before we go there, Dave, don't ask you, what is it from bit? Why is it so central? It's a wheelerism, like many things. But how do you instantiate, you know, a computer program from the absence of a computer program? I've had on, and I'll ask you to kind of steal man my opponent, maybe your opponent's ideas, but if you're if you're game but you know i've had on intelligent designers i've had on people that

believe in god believe in uh creation and and even people like mitch yokaku's your you know fellow new york professor down there somewhere in new york city but this notion that um that there's nothing we've ever discovered that can't be traced in some way to something with a design uh that contains information in other words there's life and you could say well life didn't come but something that has an information content a bit that is not associated with a mind in some sense. So computer program written by itself doesn't exist. A hieroglyphic on a cave doesn't exist without a person. But first, despite that long preamble, what is it from bit?

And then we'll get into the creation aspects of it. Yeah, well, it from bit is a slogan from the physicist, John Wheeler, who basically said maybe the fundamental level of reality, what there is is the difference between two different answers to a question. Yes or no, one or zero. And Wheeler took this in a certain direction towards the participatory universe where observation plays a central role. That kind of gets you into the role of the mind and the foundations of physics. But the direction that I want to take it, which a lot of people want to take it, is just the idea that at the bottom level of physics, there could be something digital like ones and zeros,

binary states. We have this wonderful example of this that so many people. know, which is John Conway's Game of Life, which is a cell editor automaton. It's basically a two-dimensional universe with a whole bunch of cells, each of which can be in one or two states. They can be on, they can be off. And we've got these very simple rules that a cell turns on if three of its neighbors are on and it turns off, if there are too few or too many neighbors. And from those basic laws of binary physics, of digital physics, we can generate all kinds of behavior. And from this, you know, many people have speculated that, okay, there's so many forms digital physics

could take. Maybe even our world has a level of digital physics underneath standard physics. And my sense of this is that most physicists don't take this terribly seriously as a working hypothesis. It goes way beyond any evidence we have now. There are people like Stephen Wolfram who are developing ideas in the vicinity. But I don't need it to be true. I just need it to be kind of a coherent hypothesis. Most important, Secondly, you know, these digital worlds, they're not worlds where nothing is real. They're not worlds where things are an illusion. They're worlds where things are ultimately digital, which is important for me because

it means at the very least that just because something is digital doesn't mean we should say it isn't real. A big central theme in this book is that digital objects can be real too. And if we're in a simulation, so if one thing I want to say is if we're in a simulation, we shouldn't say the objects around us aren't real. Rather, we should say they're digital, which is. to say, you know, if we're in our simulation, we've got tables, we've got chairs, we've got molecules, atoms, quarks, underneath them, and underneath those, we have bits. This is to say

that if we're in a simulation universe, we're really in a kind of it-from-bit universe, with some digital level underlying all these analog levels that we experience. Bits underneath physics. So that's how the, that's the it-from-bit idea and how it connects to the simulation idea. And then, yeah, the next question, of course, is how it connects to the creation idea. Right. So in the book, you say creation plus it from bit equals the simulation hypothesis. So where does that, how can we reconcile that statement, equation, so to speak? Yeah, it's interesting.

I was always a bit skeptical about creators and creation. Most of my life, I've considered myself an atheist, not particularly religious, don't believe in a God. The whole God idea seems somewhat supernatural to me. it. Why don't you start thinking about simulations in the simulation hypothesis, suddenly there's a root to thinking about gods, or at least creators, that no longer seems so supernatural, just an entirely natural way of thinking about gods, just as the creator of a simulation. You know, we can create simulations in our own world now without needing any supernatural powers. So we thought it'll maybe somebody in some universe created this universe as a simulation. If you have a simulation,

and it looks like you need a simulator. You need some kind of agent to create it. And then that simulators, well, I don't want to call the simulator of God or a God, but the simulator has at least some of the traditional properties of a God. They created this world. Presumably they can affect all kinds of things that are potentially all knowing about this world. Those are three of the central properties of a traditional god. Now, there are some other properties the simulator might not have.

being all good or all wise, no particular reason to think the simulator has to be especially good or especially wise. Maybe they didn't create the entire cosmos either. They created this bit of the cosmos. In some ways, that's more like what's sometimes called a demiurge in some religious traditions, like a sub-god that created at least this part of reality for us. They fashioned it as a kind of constructive creator. So most importantly, I would not recommend that we erect a religion around the simulator. I don't think this being is somebody we should worship. I don't think we should expect this being to have any special moral insight.

So for reasons like that, I'm uncomfortable describing the simulator. Even if I believed we were in a simulation, I'd be uncomfortable describing the simulator as a god. But nevertheless, it's a way of getting some of the properties of a traditional god. And yeah, and it combines the it from bit idea because, after all, how does the simulator set up the simulation by arranging a computer program, by arranging the bits? So what I want to say is if we're in a simulation, then, yeah, the it from bit hypothesis is true. It's all made of bits. And all that was put there by a simulator, by a creator. So basically the simulation hypothesis equals the it from bit hypothesis.

Hypothesis plus the creation hypothesis. Yeah. We're being called up the charm of the equation. Pretty simple. I like it. The elegance and simplicity. So to push back with a little respect, as you know, I have for you. If you think about, you know, a God as being worthy of worship, what would it take for such a simulator or Abrahamic deity?

In other words, what would it take for you personally to find such an entity worthy? or is there no combination of traits, properties, behaviors that would make any simulated or actual God worthy of David Charlemler's respect or worship, rather? I would greatly respect. There are beings I would greatly respect. I would have awe for, you know, whether it's a simulator or a non-simulator, someone who created the universe and understood so many things that they have knowledge and understanding and power that goes far beyond ours. I would have the most enormous respect, admirer. generation and all on, but I don't think I would worship them. Maybe, I guess I just don't really understand worship and why it's ever appropriate. This is an amazing being that, that I would find fully, fully awesome, but worshiping them is like a special, why would I worship them? Is it like, are they the source of all moral truth? I guess I don't really see why that would be true. It's a bit like my attitude to God would be like my attitude to Einstein, but

squared cube to the minute of power. So then it has to be something of benevolence. And there's oftentimes, you know, there is no commandment in the Hebrew Bible, which is the root of the Christian Bible. Obviously, you know that. There's no commandment to worship God. There's no commandment to believe in God. There's a commandment to love God.

And I always say, you know, you don't get commanded to do things that are natural. Like, I don't have to command you, you know, Dave, to eat whatever kind of burger, vegan or meat that you like because you like to do it. It's fun, right? it's a taste good for people that have children. I don't have to be commanded to love my kids. It's totally natural. In other hand, you have to be commanded to love something that's inherently unlovable. So in the biblical tradition, God recognizes that he is not lovable, that there are things about what he's doing and there is authority that by its nature is not, you know, it's anathema to

some aspects of creatures such as we who have free will. And I think that is the divine thing that we are endowed with free will, unlike animals. We're the only creatures. We're the only creatures. creatures that are mentioned, creations that are mentioned, that know good from evil, no life from death, know that we will die. In fact, that's what, you know, homo sapien really is referring to. So I think worship is the wrong word, and I think it's a misle. Let's like the translation, the, you know, the Sixth Commandment, most people say thou shalt not kill, right? I mean, you might, you might have been familiar with it. It's not, thou shalt not kill. Of course, God does a lot of killing. We're instructed to kill all sorts of people, but thou shalt not murder, which is

stealing of a life. So there's a lot of King James translations. One of them is worship. And I think the real term should be awe. And it's sort of, you know, in the way I have conceived of it as coming later in life back to Judaism, which was my roots, is that God should be visualized, kind of like a parent, a father, you know, unlike the graduate student girl, that's the hero of your book, the goddess of your book. But it's sort of like your father, something, a paternalistic figure that is worthy of awe, but a little bit of fear. So it's like your father's the king and you're close to him and he cares about you. And it is a personal God, which I have issues with, you know, just believing in the supernatural suspension of all the laws of physics for Brian Keating.

But there's something one of my rabbis once said, he's like, you know, well, if you were God, I'll ask you Dave, you know. But he said, you know, if you were God, what would you be doing? And I said, I don't know. I said, what would you be doing? He said, he said the exact same thing that's happening right now. You think I know more than God? You know, it's like things are behaving in a certain way. And for us, you know, it's a little bit of hubris.

And you don't have to respond to this if you don't want to. But the notion that, you know, we should believe in God, like, oh, wait, God's waiting for Brian Keating or David Chalmers, you know, if he exists, if it exists, you know, he really, no, it's more does God, you know, believe in us, so to speak. Are we acting in accordance with those principles that will create flourishing or minimize harm and crowds the greatest amount of pleasure? So I think that's the more term. And the word in Hebrew is cavade. It means to respect or make heavy. and that's also what you must do to your parents. So you can respond if you like.

I don't, you know, we don't need to continue on this main, though. I have attitudes of respect and awe to many human beings. I've got respect and awe for Einstein. I've got respect and awe for, you know, Nelson Mandela. Respect and awe for Beyonce. It's like, in very different ways. And it's like totally, if I believed that there was this God who created the universe and showed themselves to have amazing wisdom and amazing power,

then I would have respect and awe for that being, something short of worship. I mean, if our simulators turn out to have those properties, maybe they're going to turn out to be super intelligent AIs or something. I'd have respect and awe for that as well. But if they start behaving like, you know, they start raining down death and destruction, then I'm going to worry.

Actually, I got into this when I'm, I probably got into this idea when I came across my five-year-old nephew you playing a game like SimCity. And he would show me, he would build up these worlds and these cities and these forests, these everything. And then he said, now's the fun part.

He sets them all on fire. He kill all those people. They're gone. They're gone. I thought, you know, that is kind of curiously reminiscent of the Old Testament God. Right.

Yeah. So I thought, yeah, that's, so I'm thinking my God is not going to, if I'm going to, if I'm going to have respect at all for a, for a God. I hope that, yeah, they're going to be much more the way that you were describing the God.

They're not going to care whether we worship them. They're not going to rain down death and destruction, but they're going to try and set things up. Right. Yeah, exactly. So there's a lot of that. And also I often, you know, point out to people, they say, oh, you know, scientists should run the world. You know, scientists embody, you know, the best aspects of humanity. And, you know, you're so childlike.

And it's wonderful to see something. And I say, yeah, you know, scientists are like children. We're very inquisitive. We're very curious. We're very, you know, kind of, you know, egotistical in a certain good way. And we also don't play well with others. We don't share our toys with others. We compete. Like, you know, it's like we're selfish. We're jealous. We're petty. So it's like there's no single, I always say, there's no single edge swords out there. They're all double edge swords. So every blessing comes with a curse. And one of the curses of such a God is, yeah, is that temptation, that involuntary volition or otherwise, you know, like what happens if I, you know, take away, you know, this little tiny virus out there. And we'll see what happens. And we could talk The Odyssey some other time. But I want to get back to the book. You know, when I'm reading the book, I'm thinking, let's say there is the simulation.

I know my audience, and we're going to take audience questions. I have stored up from, you know, over 100 responses to you generously sharing your time of the podcast. But one thing that, you know, they want to know, and they're screaming out to know is why. And I know it's not a kosher scientific question. and ask why, but why would they simulate anything? In other words, you know, the Talmudic rabbis of 2000 years ago, you know, came to a conclusion that God shouldn't have created the world.

Like, it caused more problems than it solved. So why is there a simulation? Why? What is the, is there, you know, is there a teleological purpose for such a universe? I don't know. Why do we create simulations? I mean, we create a lot of simulations, like simulated worlds now already. I guess right now there are maybe two overartes.

purposes. One is entertainment, simulations of video game worlds and so on. There's a vast entertainment industry. And the other one is science. People like running scientists like running simulations now to understand the systems that they're working with, whether it's the cosmos or traffic patterns in the city or a waterfall. Everyone is doing science with simulations. It's easy to imagine that both reasons could be reasons why people might eventually create simulated universes. Do science run the simulation many times, pick up the statistics, see how often life develops. Hey, you want to study those Drake equation issues. Let's see how often life actually develops given certain laws of physics and how often intelligence develops and so on.

So that's one good reason. Yeah, entertainment is another or the third is predicting the future. We run simulations to predict what will happen, whether it's, say, a military simulation or a financial simulation or this, or that episode of Black Mirror where a couple is an app that a couple can run, run a whole bunch of simulations of their relationships to see if they're compatible. Actually, it's a sidebar on this. I've come to think that using simulation technology for predictive purposes is actually very difficult, especially for social predictive purposes, because. do you simulate the simulation technology or not? Like do the people in that Black Mirror episode, when they run and they go out on dates,

do they actually have. Simulations all the way down. I think you mentioned that too, the turtles are running, right? Yeah. But if you do that, then of course, you know, the simulations they're running are themselves going to have simulations.

And you're going to basically have an infinite regress. The alternative is at some point, just simulate worlds where they don't have simulation technology. That's going to be totally unberlialial. Two people can be totally compatible in a world without the technology, but the technology will ruin it for them. And more generally, I think, this is going to be a regress for any social uses of simulation technology for predictive. From origin to legend, I need to become a symbol.

I need to become the hero. Build the legacy. Let's get to work. You want to get nuts? Let's get nuts. I'm allergic to nuts. Lego Batman, Legacy of the Dark Night. This is my favorite part.

Available May 22nd on PlayStation 5, Xbox Series XS and PC. Rated E10 for ages 10 and up. Brings up a lot of little rabbit holes we can go down. Maybe I'll start with one. You know, there's the old joke. You know, if you can simulate being authentic, then you've got it made. You know, on the surface, he's deep, but, you know, superficially, he's very shallow, right? deep down his show all the it from bit you know there is a certain kernel of of um and as you mentioned

skepticism among the physics community likewise stephen wilfrum past guest multiple to guest on this show um there's skeptic great skepticism in other words there's just direct criticism and there's nothing that's practically relevant to theories of everything to unification um there these these things tend to cellular automata tend to be really good at simulating the properties of cellular automata, just in the same way that string theory is the best theory ever invented to study the properties of string theory. In other words, it's very much siloed. And I wonder, you know, is that because of this, you know, this substrate dependency versus, you know, you could get things irreducibly down to information? But again, all instantiations of computability, even if you claim the human brain is a computer, it's still running on a substrate.

How do we get the substrate? Where does the substrate come from? Or can you have a truly substrate free? I don't want to have any support. I hate when authors would or podcasts would have me on their show and say, describe your entire book so my audience doesn't have to buy. No, I'm not going to do that. But at the end of Reality Plus, you talk about the dust cloud and so forth and this

fundamental primitive computer. But the conclusion tends to be towards that, well, it can contain enough bits. And theoretically, that it could support a computational system of, you know, know, not infinite complexity, but some limited but large complexity. How do you get the substrate independence? We still need a substrate. Yeah, well, the substrate could just be something like bits, pure bits. I don't really actually need the it-from-bit hypothesis to make this idea around. I think it provides a very vivid illustration. Certainly I don't need the bits to be

serious physics required for a unification, for a grand unified theory. or it would be enough if we could have, say, a grand unified theory, physics. And then in the simulation, maybe the bits don't ever actually show up in the equations. Because, you know, when you run a simulation of Newtonian physics, then if it works well, you'll just see Newtonian physics. You won't see the bits. Right. But they're instigated in silicon or in a quip.

There still is a matter that has to exist to run the computer. So here we have two versions of the it from bit idea. One is what I call the pure it from bit idea where the basic level is bits. And you've got zeros and ones and there's nothing more basic. Yeah, the zeros and ones we all know about, they're made of something more basic. They're instantiated in voltages on transistors within circuits. So that was what we might call the bits from it idea. The bits that we actually have on our computers go into underlying it.

And that could be happen. You can combine the it from. bit idea of that. Then you get what I call the it from bit from it idea. Objects in our world made of bits. The underlying level, those bits are made of something more basic. Maybe, for example, there's a computer in the next universe up that's instantiating those bits with some analog of voltages in their world, and that'll be it from bit from it. And then the substrate will be the it's and that world. Okay, then we've got to ask the question of the it's in that world made of bits, which are made of further it's and then we'll get it from bit from it from bit from it and on you and on you go but also one thing

is also worth saying is you don't actually need classical bits to make this work you can make this work with with qubits you know the kind of the kind of the analog of bits you find in quantum computing and I think you can also make it work with with continuous bits of what I sometimes call reels in the book actually doesn't have to be binary physics it be continuous physics. The key idea is just kind of a level of information underneath everything, which could be binary information. It could be quantum information or it could even be a kind of continuous information.

I use bits because that's by far the simplest version of this idea, but I'm not sure that binaryness per se is so important. But the issue you raise of, is there a substrate beneath the information? That's super important because I think there's two very different ideas here. the world of pure information. It's a beautiful vision, but it's not totally clear. No, I agree. In fact, you know, so I like to steal man my own arguments from time to time.

So one of the things that I pointed, because I'm a, you know, more in the materialist that you do need a substrate. It's a fundamental issue. Even the qubits are actual matter, material particles, et cetera, and those in the forces and fields that interact between them. But I also will push back on my own thinking and say, well, you know, the first operating system, by definition, was written without an operating system. You know, assembly language is written without assembly language. You know, the first, the first modern, you know, editors

are written, word processors, written without word process. So, so there's proof, but, you know, at some level it is, you know, increasing the columnagora of complexity, et cetera, goes up, and you do bootstrap recursively. But, you know, it's, to me, it's, it's, it's difficult to envision without some fundamental layer of a substrate, which, you know, comes from physics. And the issue that I think presents a serious challenge that you do address in the book and you partially alluded to it now, but is the notion of what we believe at least is non-computable. So Turing's original 1950s or whatever paper or 48 paper, whatever, computable, it was on computable numbers. And it was on numbers that could be programmed and actually executed by a universal

machine, a generalized computing machine. And that excluded things. like transcendental numbers or irrational numbers. And you just mentioned continuous real numbers, but of course, all the numbers include transcendental numbers. And the two most important numbers seem to be, you know, numbers that computers have a very difficult time with. In other words, the number zero and the number infinity, if it's a number.

PIE and E, yes, no, I talk about that. But I always say, you know, the middle four digits of pi are my pin number. So don't tell anyone, because I could lose a lot in my University of California savings account. But, yeah, so those are, yeah, so they are irrational numbers. We can't, we can't, square root of two, et cetera. There's no program that can calculate them for transcendental numbers. And so they fail the a Turing type machine cannot produce them. But similarly, infinity cannot be really represented on a computer.

We can, you know, approximate it to arbitrary precision. But, you know, if you truly do things that the human mind is totally capable of finding, know, the logarithm of negative infinity. We can think about those numbers, and we can use those numbers even to construct calculus. But how can a computer of any kind capture this notion of infinity? Is that something that's uniquely relevant to this squishy wet supercomputer? Or do you think that notion will exist even in a simulated universe? And therefore, the continuum is true.

In other words, there'll be an infinite number of bits required to represent the actual universe. Yeah, well, we can, maybe we can, yeah, infinity is tricky, but we can axiomatize infinity. We can kind of get our grasp, we can try to articulate our grasp of infinity via axioms. And yeah, in traditional set theory, there is this axiom of infinity, which is a finite string of symbols that we lay out. Different levels of infinity. Yeah, exactly. Our understanding of infinity. Yeah.

But, of course, what we find, actually, is these axioms for infinity actually underdetermined. its nature. It's like how many levels of infinity are there? Is the continuum hypothesis true or false? Is the one after the integers the reels? Or is there a whole bunch in between? I mean, you can add axioms to try and determine these matters, but then there's always going to be stuff left underdetermined

for reasons and the foundations of mathematics. So it may be that we will never get a precise grasp of infinity through a finite set of axioms. On the other hand, who says that humans have such a precise grasp of infinity anyway? Maybe we can lay out enough axioms that capture the elements of our own understanding of infinity. It won't articulate everything, but maybe it will articulate everything in the human conception of infinity. So, switching gears again, maybe we'll go back to Steelmaning. Now I'm going to ask you to Steelman.

In part three of the book, the wonderful section, you end part three by saying, very carefully, you say, we can't know that we are not in a simulation. Why did you phrase it like that? And what's the best argument against that? In other words, against what would you say to what would you say on behalf of somebody who says, no, we can know that we are not in a simulation? How do they know? What is the evidence? They'll presumably try and come up with some evidence, they say, it's conclusive evidence that we can't be in a simulation. The basic reason, I think that's impossible. It looks like any evidence could itself be simulated.

If the evidence can be simulated, then we'll never know for sure. In principle, we can have simulations indistinguishable from unsimulated reality. So to defeat that, someone's going to need to come up with something. Maybe, you know, maybe who's to say, like a Penrose style argument? Look, human beings can do things that no classical computer could do. I mean, I'm dubious about that. But even if it were true, I'd think, okay, let's just move to quantum quantum. computers or quantum gravity computers and build our simulations, build our simulations like that.

I mean, maybe the biggest challenge is consciousness. Could consciousness be simulated? And many people say, you're crazy if you think a simulation could have consciousness. But I haven't thought about this a fair amount. And I don't see any bar to consciousness in principle being had by simulated creatures. I would at least say that we can't know that consciousness can't be simulated. So maybe think of that. That also justifies this negative way of putting it.

I'm lowering the bar myself. I don't want to say we know we are in a simulation because I think that's false. I don't think I know that for sure. I think there's some probability, but we might not be in a simulation. So I want to say, but we might, so the most important thesis for me is we might be. And when you translate, we might be into like epistemic logic, the not logic of knowledge. It just comes out as not K not. You can't know you're not in a simulation, but in older English, we might be in a simulation.

Is that any better? So I appreciate that. I think it's always important to describe these things. So I want to turn in the remaining 15 minutes if you can spare another 15 minutes there. Okay, great. So I'm going to ask you first. Would you take the red pill? It lets me out of the matrix.

It gives me access to all of reality. I think I would totally take the red pill because it's going to enable me to... I grew up in Australia and Australia was great. But then I discovered there was this whole wide world out there and I wanted to explore the whole wide world. If I grew up in the Matrix, I think in my view is the Matrix is real.

It's not like it needn't be a dystopia. It depends on how it's run. It needn't be an illusion. But if I discover that I'm in reality and out there is reality plus, if I get the opportunity to explore reality plus, then fantastic. Now, I'm going to want to be able to come back to the Matrix. All my family, all my friends are going to be in the matrix. I don't want to lose touch of them entirely.

And for me, the Matrix will be part of reality. But insofar as the Red Pill represents knowledge, understanding, exploration. Okay. Next, we have some questions from the audience. So what do you think is the ultimate and final machines that human beings create? This is from a man by the the name or a woman by the name of high reality sensorium, which I was going to call my second child. So what's the ultimate machine that humans, homo sapiens could create? Oh, that's easy. The ultimate

machines that Hobosamians creates is the first machine which is smarter than homo sapiens. After that, we will leave the creation. After that, we will leave the creation to the machines. Okay, next question from point text. I'd like to know if David Chalmers thinks the delayed choice quantum erasure experiment, results

support the simulation hypothesis by observing error correction codes in process, where the prior state of quantum entangled photon has changed to match its entangled counterpart. So does the delayed choice quantum erasure results, which are not fully definitively support, I suppose. Do you believe that that has any bearing, say, on the support of the simulation hypothesis?

Yeah, in general, I'm dubious. I haven't really worked carefully through those results. It's in general. There are various things in physics that people have pointed to as potential supports for simulation hypothesis, whether it's error correcting codes, possible glitches and approximations. Some people think that the collapse of the wave function might be potential support for the simulation hypothesis because it shows like a bit of rendering efficiency. Okay, well, in VR, people often say don't render the world until, you know, for the observed.

ever until you need to. So some people say in quantum mechanics, we're going to, yeah, for efficiency purposes, we'll never collapse the wave function until we need to. But as far as I can tell, the world with, yeah, the world with an uncollapsed wave function is just as hard to simulate as the one with a collapsed wave function. And, yeah, a world with a whole lot of error correcting codes,

that's going to add to the overhead. Next question is about a fellow philosopher by the name of Bernardo Castro, I guess, maybe. Do you have any thoughts or feelings on his work? I'm trying to eventually get him on the show because he has the same initials as me, BK. So, yeah, what do you think about his notion of analytic idealism? Oh, BK.

Yeah. This is a view with a long tradition in philosophy. Many of the ancient Indian philosophers were idealists. George Barclay in the 17th, 18th century was an idealist. In the 19th century, everyone was an idealist. Hegel, Kant was the kind of idealist, but then it got very unpopular for a long time. But lately it's had a bit of a comeback.

And yeah, in analytic philosophy, in the scientifically oriented philosophy, a few people are advocating the idea that underneath the world might be a level of mind. One way of getting there, actually, is by the it from bit idea. And then you say, what are the bits made of? We need a substrate.

Some people say, ah, consciousness. It from bit, from, consciousness. At the underlying level, reality is the interplay of consciousness. There are some people that panpsychists also put forward this idea. Anyway, Bernardo, I don't understand his view fully, but he certainly has the idea that underneath physics is a level of mind, perhaps a single cosmic mind. This corresponds to what sometimes called cosmopsychism. It's like the whole universe has a mind, and we're just aspects of that mind. It's extremely speculative, I think it's got many big issues that needs to address.

One big question is how does our mind emerge from the cosmic mind or from the bit minds? No one's answered that question yet. But I view him as someone that's doing really interesting speculative metaphysics. James asks me, is it true you told your kids that if it is still not solved in your lifetime, you are to put, quote, still hard on his tombstone. True or false? I'm hoping for, I'm hoping that if not, I don't have. kids but I'm hoping that maybe our AI successes will one day solve the hard problem for us.

It's like if it's too hard for, it's too hard for humans, let's just program super intelligent AIs to create ever more intelligent super intelligent AI's. Hopefully they'll be better. You don't know kids, but you have graduate students. And by the way, what was Douglas Hofsetter like as a graduate advisor? Oh boy, he's amazing. He was full of, you know, full of ideas and passions and he was interested in everything under the sun.

one weekend on humor, another one on creativity, another one on analogy, another one on sexist language, and so on. But he's also passionate about, one thing you don't quite get from his books. In his books, he's such an enthusiast for so many things. In the real world, he's actually a disenthusiast for most things. He hates 90% of AI, 90% of philosophy, 90% of cognitive science. And that comes out also a little bit in person. But the breadth of his interest is amazing.

I love being in his AI lab as a graduate student. So many smart other graduate students around from psychology and AI and full-off. Of course, his famous Gertel-Echer Bach is a work of art on the Pulitzer Prize. Just quickly on the notion, because I can't resist, like I asked you to, you know, if you're ACDC to play back in black, I can't resist you asking you about Popper and falsification. as I claim it's sort of a physicist version of Girdle's incompleteness theorem. It's kind of the best we have, but it's not quite as complete or good as what Girdle did for mathematicians. What do you make of Popperism?

Are we too inured to the paparazzi, as Leonard Suskin calls them? I don't know. Popper kind of went out of fashion and philosophy a while ago now. Most people who think about the philosophy of science and philosophy, the dominant tradition is Bayesian. We think about this in terms of probability. higher probabilities, lower probabilities, prior probabilities, and their interaction with evidence. But this total concentration on falsification, I mean, yeah, sure, falsification is important, but in general, updating of probabilities by evidence is the more important thing.

And by the way, yeah, some people worry about the simulation hypothesis that it's not falsifiable. Some versions of it might be falsifiable, imperfect simulation hypothesis. The perfect simulation hypothesis may be not falsifiable. Bible in principle. But that's just to say, okay, the perfect stimulation hypothesis, I'm happy to say it's not exactly a scientific hypothesis. It's still a philosophical hypothesis. Popper himself put forward many philosophical hypotheses that weren't scientific hypotheses. So I think it can still be meaningful. Okay, Dave, we come to the end when I ask my thrilling three existential questions about the meaning of life, your past world line, your future history, your gifts to the universe.

But to watch this, you're going to have to subscribe to my YouTube channel. Not you, Dave. you're going to watch it in real time because you're going to participate in it. But you have to subscribe to my mailing list, Brian Keating.com. And there I update and I provide all links to David's wonderful work and his book, of course, that you should buy. I listen to it and read it. I think it's just one of the,

it gets your best to date and I can't wait to see how successful it becomes. So if you want to hear Dave answer the thrilling three final questions, you'll have to subscribe to the channel and to my mailing list, Briankeeting.com. I'll send you the link. So for now, signing off of this main portion of the. episode of Into the Impossible with Dave Chalmers, NYU, and author of Reality Plus, a guide to philosophy's problems and the ultimate answer to the perhaps the most important questions

of the modern age. Dave, thank you so much. Ambition comes in all shapes and sizes. At First Citizens Bank, we roll with your goals because we're built for what you're building. Fit for your ambition.