AS290: Are We Living In A Simulation? With Phil Torres

Transcript

This transcript was generated automatically and may contain errors — the audio is the record.

0:00:29 Thomas: Hello, and welcome to Atheistically Speaking, this is episode 290, and I'm your host, Thomas Smith. All right, got a really cool episode for you today. Here's the thing, this is, again, one of those that I was actually going to do two parts, and then I got incredibly sensitive to the fact that you guys are really supportive, I value your time a lot, and I thought the second part would have been a little too repetitive, and um, so I bailed on that idea. I'm giving it to you all at once. This is a cost to me because I have to, I have to adjust things and make sure the schedule works out so I can provide great episodes for you on a two-episode-per-week basis. It's not easy, but I just wanted to let you guys know that I really, really do value your listenership, and I wanted to make sure that you wouldn't feel at all, I don't know, maybe unsatisfied with the part two.

0:01:26 Thomas: That being said, taken in whole, I really like this episode a ton. The first part we go through, the so the first half we go through the simulation argument. It's great. I love the argument. It's fascinating. Really cool topic. And in part two, which well was going to be part two in in the second half, we try to, I have a, what I think is an issue with the argument, I try to voice that and Phil tries to get back at me a little bit with that and and we go back and forth. So, with that being said, let's get to a nice, long episode with Phil Torres on the simulation argument, so interesting to me. Thank you.

0:02:13 Thomas: Joining me right now, back for what I think is the third time, all right.

0:02:18 Phil: I think that's right, good.

0:02:19 Thomas: Third time, the three-peat guest, Phil Torres, how you doing?

0:02:23 Phil: Good. How are you doing? Excellent.

0:02:25 Thomas: I'm Doing great, I first Time you're here to talk about your book, The End, What Science and Religion Tell Us About the Apocalypse. And you were Back to talk about your Skeptic article about the future of terrorism. And now I am really excited to talk to you about the simulation argument. I saw you post something about it and it gave me an indication that you have a more technical knowledge of it than I do. And I'm excited because I've been wanting to talk to somebody about this. So, really excited. But before we do that, why don't you take a minute and tell new listeners who maybe weren't back with us before a little bit about who you are and what you're doing.

0:02:27 Phil: Yeah. Sure.

0:03:02 Phil: Yeah, so as you just mentioned, uh I published a book earlier this year uh called The End. It's sort of about existential risks and apocalyptic terrorism. Um, shortly after that, I, I founded a still quite fledgling institute called the X-Risks Institute. Uh, and you know, the idea is just to better understand the nature of uh our evolving existential predicament in the universe, um, in part because of, you know, pretty new and significant phenomena like climate change, biodiversity loss, as well as emerging technologies like like nanotechnology and biotech. Another thing the Institute tries to do is understand not just the technologies, but the various agents who might want to use those technologies for harm. So this might include apocalyptic terrorists, would be an obvious instance.

0:03:50 Phil: It could be maybe future eco-terrorists or states. You know, maybe even, you know, there's a group of individuals who, which is a pretty small group, but it could potentially get larger in the future, who subscribe to a particular ethical theory called negative utilitarianism, according to which the ultimate aim of moral conduct is to reduce suffering in the world. It's not to reduce suffering and increase happiness or well-being or something. It's merely to reduce suffering. So it follows that you were to exterminate humanity, that would actually be an ethically good, commendable thing to do, because you'd be eliminating suffering. So we also focus on those aspects, which individuals, if they had access to a doomsday button, would actually push it either on purpose or or perhaps on accident. So yeah, I, I publish, I do a bit of academic stuff and then I try to, to engage in some outreach as well, like publishing for a motherboard and bulletin, the atomic scientists and, and, uh, you know, non-academic journals like that.

0:05:01 Thomas: Awesome. Yeah. And I remember it was interesting to talk about those existential risks, how maybe they're a low percentage thing, but the cost is so high that it's worth thinking about. But so people can go back and check out those episodes we did originally. So, but for today, are we living in a simulation? I went back, I'm so glad you sent me the website. I'll link it, but simulation-argument.com had everything I needed to know.

0:05:27 Phil: Mm-hmm.

0:05:28 Thomas: And for some reason, I guess I thought there was more history to the argument. I didn't realize that it virtually all spawned from the Nick Bostrom paper in 2003. And so I read that and I was like, oh, great. I have a good good basis and starting point. But why don't we go through it and I'll allow you to do the steering here.

0:05:49 Phil: Yeah, sure. So, I, the simulation hypothesis was first mentioned by a guy named Hans Um, But so it's, it's been around for a little bit, and it, it has some um, it has sort of echoes of ideas that have been around four, four millennia, like Plato's cave, although, although, this, so I'll clarify this in a moment, but like Pluto's cave, you know, where we're, you know, the ideas you might be living, you know, which are actually seared to shadows, it's not the real, it's a real reality, um. And, you know, Descartes famously had this idea of the evil demon that tricks you into, you know, believing that, you know, in my case that I'm sitting in a chair and in a particular room when actually that's that's not the case. It's just this very powerful genius that is has deceived me into thinking that.

0:05:55 Thomas: Oh, okay. Right, right.

0:06:40 Thomas: Correct me if I'm wrong, but isn't the only way he got out of that by assuming there's a God who wouldn't want him to be deceived like that? Wasn't that a part of his argument?

0:06:41 Phil: So, Yeah. Exactly. So so, you know, he was a a foundationalist so he tried to get these these foundational um beliefs that that couldn't be doubted and then to build up uh knowledge from that so he you know one of the foundation foundational beliefs is uh is you know I think therefore I am so that's something that can't be doubted another thing uh if if I remember correctly you're exactly right that you know it's

0:06:53 Thomas: Yeah, right, right? It's pretty easy when your foundational belief is actually not something that would depend, would need a lot of proving for me. But I feel like he cheated there. But anyway, I get, yeah, so you're right to say that totally makes sense to say this, this is very closely related to ideas that go back a very long time. And I didn't mean to say all original, originated. I just, this specific argument, I think did kind of come from just this paper.

0:07:11 Phil: Yes. Yeah. Yeah. Yes. You are exactly, exactly right about that. Um, so there's a difference between the simulation hypothesis and the simulation argument. So the argument is is pretty novel and that was proposed by Nick Bostrom at Oxford University, who runs the Future of Humanity Institute over there, um, and this and part of the simulation agreement is the simulation hypothesis.

0:07:41 Thomas: Hmm. Okay.

0:07:54 Phil: So basically, yeah, okay. So Bostrom argues is that, the whole of the argument is that one of the following three disjuncts must be true. Or actually, to be even clear about it, at least one. So it might be the case that all three are true, or perhaps just two of the three, but at least one of the following claims is true. Either our civilization self-destructs before we reach a post-human state, which is defined as just a state in we are technologically sophisticated enough to run simulations, vast vast simulations in which simulants or sims with minds, you know, something like ours exist.

0:08:45 Phil: So either we self-destruct before we reach that state, or we continue to develop as a civilization and reach a post-human state, but for whatever reason, we decide not to run simulations. And then the third option is we're living in a simulation. So those are the three options. And And he provides this argumentative apparatus that attempts to show that this is the case. At least one is true.

0:08:59 Thomas: Right. Yeah. Yeah, Yeah. And it's interesting because I think when I've heard it presented, I think I've heard it a little more lazily or quickly. And so I was under the misconception, I think maybe some other people might have been, that the simulation argument was an argument for us being in a simulation. But it's not actually that. It's saying that at least one of these three things is true. Andso we can go on to how it works from there. But just to clarify, that was something I learned. He doesn't believe we're in a simulation. He just thinks if this argument is sound, then one of these three things has to be true.

0:09:16 Phil: Yeah, yeah exactly. So he thinks we, I certainly don't want to speak for Bostrom, but my sense is that he, I mean he most definitely believes that one of these three is true and it may be that we're in a simulation. There's also reasonably high probability that there's some kind of existential catastrophe, a great filter of sorts towards which our civilization is rushing and will catastrophically collide with. In fact, in Bostrom's 2014 book called Superintelligence, he argues that we should recognize the default outcome of a superintelligent machine to be doom. In other words, there's an extremely high probability that if we do create super intelligent machines, something really bad is going to happen.

0:10:38 Phil: I think a lot of people, probably Bostrom himself, think that it is also very likely that we will create these super intelligent machines. The first disjunct, which is the self-annihilation scenario, definitely quite plausible.

0:10:55 Thomas: Is the first one a self-annihilation thing or is it just annihilation of any kind?

0:10:57 Phil: Um, it would be annihilation of any kind, but I think a lot of like existential risk scholars think that if we do, if we, do, uh, if, you know, civilization collapses, it will probably be because of our own actions. You know, like the probability of an asteroid hitting is, is pretty improbable. I've used species-destroying asteroid hits earth every once every, uh, 500,000 years. If I remember correctly. Super volcanic eruptions are very rare. Climate change is going to be a pretty major catastrophe that's going to last the next 10,000 years or so. But I don't think many people think that that's gonna result in human extinction. It's gonna results in a lot of problems.

0:11:15 Thomas: Right, right. Yeah.

0:11:45 Phil: But so I think the thought is that, yeah, if we do go extinct, it'll probably be because of a nuclear war or some kind of engineered pandemic or something of that sort, uh, or, you know, super intelligence takeover, something like that.

0:11:55 Thomas: Okay. Yeah, yeah. Well, we'll have to talk about that another time.

0:12:02 Phil: So one way to think about this is Boston himself has put it, this argument essentially is a way of constraining and shaping the sort of space of possible futures. You know, so, you know, perhaps we reach this advanced post-human state and for ethical reasons, maybe we decide not to run simulations in which creatures with minds like ours exist, because perhaps the suffering would be too great. In fact, one of the biggest, the most compelling arguments a lot of atheists believe against the existence of God is the argument from evil, which is essentially that there seems to be an incompatibility, probabilistic or logical, between a God who's omnibenevolent, so perfectly moral, and omnipotent, can do anything that is logically possible. And then the world we happen to find ourselves in, in which, you know, suffering and evil are absolutely pervasive, including, yeah,

0:12:40 Thomas: Yeah, I'm certainly sympathetic to that argument. I think that's true and I would would even go so far as to say, you wouldn't even have to be omnibenevolent. You just have to be pretty good. You're just going to be moderately decent as a being to say maybe not put a bunch of people through really horrendous things. But yeah.

0:13:23 Phil: Well, That's a good point. And it underlines exactly how powerful the argument from evil is, because the traditional conception of God that you find in the Abrahamic traditions, why not, is one who is morally perfect. So it's a very strong claim. Still have an argument if you said God is just generally good, but a flawed being. Well, maybe you'd expect the world to be at least a little bit better than it actually is. So Perhaps

0:13:55 Thomas: But I think it is funny. It makes the argument even stronger to consider that one day humans, let's say we do gain this capacity to create these experiments, we would, I do think we would say, I think we would stop and we'd think, you know what, if we make a simulation that has conscious feeling beings in it, we better do it right. Like we better, we'd better not make it like this, the real one. I mean, how many people had to suffer? I think it's hilarious in a way that we would totally do that, but we can't expect that of God. Like somehow God wouldn't have made that calculation.

0:13:55 Phil: Yeah. Yeah. Totally. Yeah. So this is exactly my point, the same line of reasoning that underlies the argument from evil could also perhaps motivate, you know, post humans to not want, excuse me, to not run simulations. They might just see it as utterly unethical. Then again, they might not. And if they do, then when that, so, you know, if we develop into the future, it's 100 years from now, and perhaps, you know, simulations are ubiquitous, you know, every middle school kid, you know, is able to run a couple universes on her or his computer, then Bostrom's argument says that we should take that as very strong evidence that we ourselves are in a simulation. So there's, so moment, um, but I think first actually it's really worth mentioning just a few things that this argument presupposes. One is that you can create conscious beings uh within a simulation.

0:14:44 Thomas: Yeah.

0:15:33 Thomas: Yeah. Yeah, I was going to ask about that very thing and see What your thoughts are, because it's interesting, like I totally can grasp and understand the idea that we could create like a robot or something that is completely conscious and is more or less like us. It's a little harder for me to quite, you know, grasp onto the idea that just within some sort of software, we could make consciousness. I think I get how it's possible, but it's a little hard to grasp. Do you have any thoughts on that?

0:15:33 Phil: Right. Yeah.

0:16:03 Phil: So it, it definitely is, uh, it's an odd thought, but I think most philosophers, philosophers of mind in particular, uh, don't really see a problem with it. So, so the idea is you, there's a a long-time question in philosophy of what the nature of the mind is, and many theories have have been proposed. I mean, there was, you know, behaviorism, which dominated part of the 20th century, um, which more or less wanted to eliminate any talk of mental states or anything like that, and and sort of reduced all of that, those phenomena, to just behavioral actions and dispositions and so on. And there was another theory at the same time that was called the identity theory, which said that the mind just is the brain, you know, so any given, like, type of mental state is reducible to a type of brain state. And then people thought close, more closely about that, and decided that that doesn't really make sense, because you and I can be in the the same mental state.

0:17:01 Phil: We can both believe, for example, that Obama is president right now. But it's almost certain that if you were to open up the skull and look at the microstructure of all the neurons that are residing in your cranium, they would have a different structure. So they would be different. So you would expect the brain states to be identical, the belief states are identical. Does that make sense? So,

0:17:25 Thomas: I Don't know. I mean, I think this is a really complex topic, but I wouldn't expect just because we share one common belief on a proposition that that would mean our brain states would be the same. I mean, the brains are so much bigger and more complex than that, that I think it would be very difficult to track down where everything in my brain would need to align in order to have that true belief. And then everything in your brain would needs to align. I guess maybe I'm misunderstanding.

0:17:56 Phil: No, no, no. I think you're getting at exactly the reason it was abandoned. At first, it seemed, I mean, there's a historical perspective. If you sort of immerse yourself in the literature of the time, sometimes theories which seem silly in retrospect seem quite plausible at the time, given what people know, or the sort of thought currents that people were sort of wading in. So yeah, there was a time when people just tried to reduce the mind to the brain. We'll just say that the mind is the brain, and then they thought a bit harder about that and realized that there are all sorts of problems.

So that led to, right, so this led to exactly the next theory, which is called functionalism. And this is a theory that is widely accepted by cognitive scientists and philosophers of mind today, and it's precisely functionalism that makes plausible the idea that you could create actually conscious beings within a completely simulated world, and whose brains and minds themselves are being simulated. So functionalism is basically just the idea that what makes a mind what it is isn't the stuff that it's made out of. It's not like the brain itself.

0:18:01 Thomas: Hmm. Oh, okay.

0:18:34 Thomas: Well, but the mind is still the brain. Maybe that's what I'm getting hung up on. I mean, sure. Go ahead. Okay.

0:19:19 Thomas: Sure.

0:19:20 Phil: Rather, it's what it does. So there are other examples of this in the world. For example, what makes gold gold? It's that it has a particular number of protons. Hydrogen has one proton. That's the essence of what it is to be hydrogen. If you have one proton, then you're definitely hydrogen. If you have more than one proton, then you are absolutely not a hydrogen. There are many other things, like poison is an example. The properties that make poison what it is have nothing to do with the substance that the poison is made out of.

0:20:00 Thomas: Right. It's how it affects us, I guess. Or maybe another creature. Yeah.

0:20:02 Phil: It's how it it affects exactly. So it's exactly, so it's, yeah, it's how it's you know, it's it's harmed. It's defined by its harmful effects to biological systems. So I mean you could have, you know water could be poison, you know there's such thing as water intoxication, you know that you know if you have enough of it it's going to affect you in a particular way. Yeah and so on for all the other you know toxins out there like you know lead and arsenic and so on. So maybe it's the case that minds are defined not by what they are physically, but by what they do. So therefore, it follows that if you have a system, it shouldn't matter what the system is made of. If the system does what mental states do and what minds do as a whole, then the result should be a conscious state. So one of the ways to think about this is, ultimately mental states consciousness a product of a certain certain kinds of functional organization. That's the term that's that's often used among of mind.

0:20:19 Thomas: Okay.

0:21:08 Phil: So functional organization is the key. It follows that if you were to create a computer system that exhibited the same abstract functional organization as your brain, then you should expect it to be conscious no less than you are.

0:21:23 Thomas: Right. Let me, uh, let me try it. Cause this, this actually does bother me about the argument. I'm willing to accept it to go on, but I'm curious just maybe if I can gain a better understanding from asking you about this. Because I, again, I totally get it when it's a robot brain where the brain, you know, or roughly speaking, the robot has programming to do whatever we could eventually make it complex enough to where it's functioning, in a way and it has sort of cognition, it has information processing and memory and the things that would lead to consciousness. It's a little less clear to me how that works inside of software.

Because when you compile this code, let's say we've got one of the giant processors, giant computers that he describes in the argument that Bostrom describes, it's maybe the size of a planet or something. Ultimately, we're still running code. We're still running ones and zeros, and it will compile in a way. I don't get how... Let's say you have... I guess to draw an analogy, if you have a car racing video game, sure, you have cars in the game, and you could look at it and say, well, this is a car. It's driving. But you've You've programmed the graphics, you've programmed what it should look like, what it should feel like, and all that.

0:21:53 Phil: Yeah.

0:22:42 Thomas: That doesn't really make it a car when you break down the coding of it. In the end, it's ones and zeros. It's a little hard for me to grasp. I'm wondering if you can provide any more insight, or if I just kind of need to accept that it's something that's hard to grasp?

0:22:57 Phil: So, yeah, I think you're right about the car situation. In the case of minds, I think the situation of minds is different, though. Um, within a computer, perhaps the the best thing to do is just to to reiterate a bit, or to maybe to go into a bit more detail about the the whole idea of Uh, you know, these, you know, types of mental states that are being identified with types of functional states, and where those functional states are independent of whatever, uh, the material substrate is. So again, the analogy of poison is perhaps quite helpful. Of being a poison is identified with a certain functional state. Namely, it deleteriously affects biological systems. Now, you can swap out the material substrate with, again, with lead or arsenic, or whatever, you can swap that out, and the higher level property of of its being a poison will remain the exact same. So the idea is, you know, with with respect to the to the human nervous system, this system is, you, it's super complex, you know, it's

0:23:09 Thomas: Okay.

0:23:18 Thomas: All right.

0:24:06 Phil: The most complex known object in the universe, with, you know, 200 billion neurons, each of which makes 10,000 connections or more. So it's just a massively complex information processing system. So if you sort of abstract away from the physicality of it, you can imagine information, in the form of whatever form it would take, being shuffled around from one side to the other. And it's that process that results in ability to solve math problems, our capacity to, or to dream. And it also – it appears to be related to this very high-level property known as consciousness, right?

0:24:51 Phil: So one of the things that follows from functionalism is what is called multiple realizability. And that's exactly what I was talking about before. The material substrate unique about biological matter that would give rise. So, you know, you could imagine, for example, you inject some, you know, some nanobots into your brain and they go around to each cell and they they destroy the cell and then build up a fake, an artificial cell in its place that has basically the same structure. It has a little extension that sends signals called the axon, and then it has this little extension that receives signals called dentrites. And so imagine if you replaced one cell.

It seems unlikely that that would influence, it would affect your consciousness. Right. I mean, you'd still be because Everything would be functioning the exact same. So then you can imagine doing this incrementally, and, you know, you replace 100 cells at a time. There is definitely some debate about this issue. But to me, it seems fairly clear that there's no point at which I suddenly lose consciousness. Consciousness, um, nor is there any reason to think that I would, with a fourth of my brain being, you know, artificial, you know, hardware essentially, uh, that I'd be any less conscious. So then you could imagine going through the whole brain.

0:25:06 Thomas: Sure. Right. Sure.

0:26:17 Phil: Until it's it's all just, you know, these little nanobots that are sending signals exactly the way my neurons would. It's hooked up to my peripheral nervous system, which, you know, controls movement and stuff.

You You would expect my behavior to be the exact same, because the causal link between the signals that are impinging upon my ears right now and going into my brain, and then that information is just a bunch of signals that have been rearranged and then sent out to my tongue to produce a sound, all of that would be intact. Um, along those lines, philosophers have often thought that it seems to be the case that you would be, you would retain your consciousness, uh, as as well.

So then it's it's, of course, only one, uh, small step between, say, you know, saying, "I'm going to replace the the, you know, white and gray matter in my brain," with some sort of artificial material, and then saying, like, "I'm I'm going to take the abstract functional organization of my brain and just implement it on a computer," and have some really, really sophisticated software run the computer behind my eyes on the, you know, as

0:27:15 Thomas: Yeah. Yeah.

0:27:25 Phil: A virtual computer on the whatever supercomputer you're running it on. Does that help a little bit? Does that make sense?

0:27:25 Thomas: Yeah. Yeah, it does. I think that ultimately it's something I don't know that I feel like makes sense, but I, you know, but because it's so interesting to think about a simulation and the computer, I totally get the substrate independence concept. I get that. And it makes sense that, you Know, we could, again, we Could have a robot brain or we could have our brain that was replaced with all artificial material. And I don't believe that that makes a difference. It's not the substrate exactly, but it is a little weird to think about it being represented in a piece of software. And furthermore, not just one brain, like all, like every brain, you know, let's assume we're in that simulation right now. Somehow we're all sharing the hardware and we're also sharing, you're sharing the software. So I think that's you know it's amazing to me that that could be possible and that we would be unique. But I think I'll accept it for the sake of argument and just say that

0:27:36 Phil: Yeah. Yeah. Yeah. Yes. So

0:28:23 Thomas: It sounds amazing.

0:28:24 Phil: I'm wondering if your objection isn't more technical than philosophical. It's about the technical feasibility of creating a simulation where you have a bunch of these really complex systems within a single program that are interacting and are being stimulated. You know their, their nerve endings, sit, you know, they're, simulated nerve endings are being stimulated in the very same way that our nerve endings are stimulated, you know, in our world, when we touch the desk or you wander through the wind, you know, Or something, um.

0:28:56 Thomas: Yeah. Yeah, sure, yeah. So I mean, we can take it for granted, it's just interesting to talk more about that though, I think it makes sense, right?

0:28:58 Phil: Yeah. Well, I think this actually gets at two key, um, assumptions that Bostrom's argument makes. So, so these assumptions aren't, aren't really part of the argument itself, but they are sort of the background, propositions that the whole argument is predicated upon. And that is that functionalism is true. So That if you recreate the right functional organization in a computer, it doesn't matter, doesn't matter where. I mean, you could do it with, you know, jars with pennies that you pull in and out. It doesn't matter what it is, but if you genuinely recreate that functional organization, you will get consciousness. And then the other one is the capacity to, the possibility of having the computational power necessary to create a huge simulation where perhaps you have billions of independent conscious, you know, conscious minds wandering around, interacting with each other, uh, climbing mountains, sitting in front of their computers, and perhaps even, even creating their own simulations within the simulation. Since computers, just like minds, are functional things, you should be able to create a simulation within the simulation, and so on, perhaps, resulting in this huge stack, this huge hierarchy of nested simulations.

0:29:13 Thomas: Sure. Yeah.

0:30:11 Thomas: Which, that, oh man, I, I don't, uh, I think I would need to study that and think more about it, but when I read that part of the argument, I was amazed that it would make sense that we'd be able to do that, because it seems like you could go infinite. And maybe you would just eventually crash the system or something.

0:30:19 Phil: Yeah, yeah.

0:30:29 Phil: Yeah, it's, I Bostrom actually mentions that in arguments against this hierarchy, hypothesis is that costs are inherited upwards, you know, so if you have a on level B a Civilization that's running a billion simulations on a lower level C the level a above But both B and C is gonna have to have the computational power not only to run Simulation B, but all of the other simulations on level C. So so it could be the case that due to the you know, the computational costs of running simulations, right as we begin to run our own simulations, our simulators look down. Exactly. They say, this is, I'm not going to pay for this. This is ridiculous.

0:30:56 Thomas: Yeah.

0:31:08 Thomas: We get the blue screen. Yeah, yeah. We give them we get the memory dump like we have to restart Windows and everybody. Yeah man. Interesting. OK, so we'll take those two assumptions on board, though. Well, I'm happy to assume that what you said about functionalism and that it can be done in a a computer or in a program, and also that we could one day develop that computer. I'm happy to take those on board.

0:31:19 Phil: Yeah. Yeah.

0:31:37 Phil: And just to be clear, I mean, not all philosophers or scientists or, you know whoever, accept both of them. I mean some believe, believe, you know, Moore's law, and, you know, we're going to hit certain limits of computing power. Others are not, other philosophers are not functionalists. They believe that, John Searle is a very famous example of this. There's something special, something, some special unknown property of biological matter that enables it to produce consciousness. So even if you got the right functional organization in a computer, wouldn't be conscious. It wouldn't understand the way we do and so on. So the argument is most definitely based on these two assumptions. Think one thing a lot of people find confusing when they first come across the argument, is how the third disjunct is couched, right? So the first one is that we go extinct before we reach a post-human state. The second one is we reach a post-human state, but we decide not to run a large number of simulations. The third one is that we're in a simulation, right?

0:31:43 Thomas: Mm-hmm.

0:31:56 Thomas: Mm-mm.

0:32:42 Phil: So something doesn't seem right about that. So there is, as I mentioned earlier, this sort of argumentative apparatus that Boston provides for the third disjunct. So if you think about it logically, if the first two disjunctions are false, then it It follows automatically that we reach a post-human state and we simulate a large number of universes. And these aren't just universes some arbitrary physical constraints or laws of nature or something. These are universes in which the design is such that beings with minds like ours exist. So that's a really crucial part of the argument is, know, is that we run, Bostrom calls them ancestral simulations, just simulations of our evolutionary history. So, and the reason that's important is that, know, imagine again, this is like a hundred years from now.

And we've, you Know, every middle school kid has, uh, per his own simulation, which, you know, maybe they've got like a hundred simulations on every computer and there's a billion people in each one. Ultimately you get a huge number of, of Sims existing in the world. Now, if you take a step back and think about what this might mean, there's a question you have to ask. You know if you're simulated or not? Because you could look down at the simulations you're running, in which these conscious creatures are having rich experiences, that is, if the simulation is sufficiently high resolution. They're having rich experiences, they're climbing to the top of mountains, going on dates. They're feeling all the sorts of emotions that humans feel and capable of having all the thoughts that humans have as well. There's nothing in their environment that they can look at or point to or study more to figure out whether they're in a simulation.

0:33:06 Thomas: Yeah. Mm-hm. Right.

0:34:37 Phil: Simulation, right? And so the exact same, that's exactly our situation as well. If we were in a simulation, we would expect things, if we were in a high resolution simulation, let me put it that way, we wouldn't expect things to look exactly the way they are. So as soon as you recognize that fact, then you should start to wonder about the probability of you being one, one of the very few who's not simulated versus one of the very many who is simulated. So this brings in something called the indifference principle.

Indifference principle basically says that when you have a bunch of options, you don't have any independent reason for thinking one is more probable than the other. You should distribute your probability evenly, right? So we're in this situation, again, imagine it's a hard years, there are 10 billion humans living on earth And there are 100 trillion sims within the very simulations that we're running. So what we're indifferent to is whether or not we exist in vivo as actual real biological things, or what you might say, in machina. You know, whether we're just conscious minds that are ultimately ones and zeros. So you're indifferentto that. Therefore, you should assign your probability evenly.

0:35:06 Thomas: Right. Mm-hum.

0:35:53 Phil: And since there are way more sims than non-sims, you should think that you're a sim. So that's the crux of the third disjunct, which is the simulation hypothesis. Does that make sense?

0:36:07 Thomas: Yeah. Yeah. It makes sense. I have an objection, but I think we might talk about it later. I want to say, so first of all, when you say that, like I mentioned earlier, it doesn't mean that this argument proves we're in a simulation. I just want to emphasize that because I actually saw, I know you didn't want to speak for Nick Bostrom. That's, that's totally fine. But I did see he was asked that question and he said, no, it just means, you this doesn't prove that it means one of these things has to be true. At least one, as you put it. Um, and so that's interesting, but it, but it is still worth, I think what's so interesting about the argument is that it seems to just from, you philosophizing a bit, it seems to tell us something about the nature of even our universe or existence in general, which seems very powerful for just a philosophy argument to do, you know?

0:36:23 Phil: Exactly. Yeah. Yeah. Yeah.

0:37:00 Phil: Yeah, absolutely. I mean, that gets at exactly what I was mentioning earlier, that this sort of provides a framework for thinking about future, and it's it's a framework with very deep metaphysical implications, you know, either we're in a simulation, we can be pretty sure that we either die before we reach an advanced state or we reach advanced state, decide not to run simulations. So I mean, that's a very, that's a very interesting, uh, and strong claim about the sort of space of uh possible or probable futures creatures for our civilization.

0:37:34 Phil: But I think Bostrom would say, just to also be clear, so the argument itself not conclude that we're in a simulation. But if we advance to some state where we can run a whole bunch of simulations, then I think Boston would say we have really good evidence for thinking we are in a stimulation for exactly the reasons I mentioned and before, having to do with the indifference principle and the incredibly skewed ratio of sims to non-sims.

0:37:41 Thomas: Right.

0:38:02 Thomas: Right. Right. So now I think I want to try to talk about what I don't think works about this argument, if you don't mind. So when I first talked about this back in, I don't remember, it was months ago, but I'd heard, I think, I just heard Sam Harris give a summary of it. And my problem with it was that it seems circular. My problem was that, it feels like we are taking a bunch of facts about our current world that we're in, and then we're using that to conclude things about reality in general. And if we are indeed in a simulation, there's going to be a whole bunch of contingent stuff that we don't know. We don't know anything about the outside world. We don't have access to all these different facts that could change how we look at this.

0:38:04 Phil: Yeah, yeah, sure.

0:39:01 Thomas: And me like it was circular. And it's the same thing when you look at that claim that you just made, that if we do, as a species, end up creating simulations, that therefore that somehow increases the likelihood that we are in a simulation. That seems incredibly circular and weird to me. Now I'm fully ready to realize what mistake I'm making, but that was what sort got of sparked my suspicion. Now, What I did was, on this website you gave me, which is, again, really awesome, simulation-argument.com, there's an FAQ that Bostrom did, and he actually addressed this claim. And so I thought it was so interesting. Maybe, I don't know, what do you think about even just reading his answer? Because I think it may be a good basis. I think there's something wrong with it. We can do that, or do you want to respond first? What do you think?

0:39:27 Phil: Yeah, I I mean, I'm happy to just summarize what Bostrom writes here. Yeah, he essentially says, so if we're living in a simulation, there's one thing we can know for sure about our simulators. Actually, I believe there are two things if we're in a simulation. One, we can know that our simulators are not omnibenevolent and omnipotent. That's essentially just the argument for people. So I feel very confident you can infer that our stimulators are going to be morally mediocre entities. Entities.

0:40:16 Thomas: Yeah. Yeah. I agree with you on that, yeah. Yeah, yeah.

0:40:31 Phil: So, and the other thing you can infer is that the universe above ours is such that they can create simulations. So if, so if that's the case, as Bostrom, you know, argues in his then response, that means that third disjunct is true. We're living in a simulation, right? So in other words, his argument, which is that at least one of the three is true, true, is true.

0:40:39 Thomas: Yes.

0:40:49 Thomas: I agree with that. Right, right.

0:40:55 Phil: That doesn't vitiate or any way undercut the argument. Then on the other hand, if we're not in a simulation, then all of the evidence that we were talking about before, like creating these massively powerful computers and then running a whole bunch of simulations and, and you know, going from there, um, that is perfectly compatible with, uh, one of the three disjuncts being true as well, so that also doesn't seem to to be problem for the argument

0:41:24 Thomas: I think I may need to try to read that because I do have a specific, I think what's happening is I do really think this is a bit of a logic trick, because I, A is totally, I think you represented that great. It's totally true. So he said he had that response. He said, you know, this argument doesn't work. The objection that because we're in a simulation, it undermines the whole reasoning to begin with. It doesn't work because look at the possibilities.

Either A, as you said, we're in a simulation, therefore three is true. Easy. I get that. I totally grant that. Or B, we are not in a simulation. And he says, then the empirical evidence noted in the simulation argument is veridical, taken at face value, suggesting that a technologically mature civilization would have the ability to create a vast number of simulations. And consequently, by the simulation argument, there's a very high probability that at least one of the three disjuncts is true. What I think doesn't make sense there is that in that sentence, you have to eliminate three because we're not in a simulation. So when you look at it that way, it actually becomes the most, I don't know, like kind of a meaningless statement. Either A, we're in a simulation and therefore we're in a stimulation, or B, we are not in a simulation and, therefore, only one or two is possibly true.

But that doesn't make any sense because if we are NOT in a simulation, we we actually have no reason to expect either that we do create a lot of simulations or that we don't. It completely undermines all the reasoning. So it's... I think he snuck 3 into B, but it shouldn't be there because he already granted that we're not in a simulation.

0:42:15 Phil: Well, I don't, I actually don't think that that would make a difference. Also, I, I'm, um, I wonder how easy it is for listeners to follow along. I mean, if you go to the website, you can actually sort of read.

0:43:05 Thomas: Yeah, yeah, sure. That's why I tried to, I tried to re-say what you said just in case it would help a little bit.

0:43:13 Phil: Sure. Yeah, yeah. Yeah. So I mean I think, I actually, I don't think that that would affect it because I mean he could have rewritten B by saying that either one or two is true.

0:43:24 Thomas: Yeah, but that wouldn't, I don't believe that would work because okay, let's picture just B because we've, he's separated these two things, right? So he's done just B, we're not in the simulation. How can you go from that to then then concluding that either, and I'll reread one and two just to make sure people know what it is. So picture it this way. B, we're not in a simulation. And then he says, therefore, we have to look at the empirical evidence noted in the simulation argument. And therefore, one or two is true. One is the human species is very likely to go extinct before reaching a post-human stage. Or two, any post- human civilization is extremely unlikely to run a significant number of simulations of their evolutionary history.

0:44:08 Thomas: Why would those two things be true? Why would either of those, why would only those two things be truth if we're granting that we're not in a simulation?

0:44:08 Phil: Yeah I think because if both of those are false, then it follows that we reach a post-human state and run a large number of simulations. That's just logic.

0:44:23 Thomas: Mm-hmm. That's fine. What's wrong with that?

0:44:26 Phil: And if that's the case, then we are in a simulation.

0:44:28 Thomas: Ah, but we're not. We know we're Not. I seriously think this is a logic knot that he tied. Because that's the whole basis of this argument. If we go with B, we are not in assimilation, There's no, just picture that. We're not in a simulation. Let's pretend hypothetically you and I know that somehow, like we just know, hypothetically.

0:44:48 Phil: Yeah.

0:44:49 Thomas: Why would we then be able to conclude, ah, therefore we know that humans won't one day create a vast number of simulations? We don't. I think the argument fails because of that.

0:45:00 Phil: No, okay, okay. Let me try to address that. The way he sets it up here, I think the way he articulates it here might be throwing us off a bit. Because the idea is that it's It's ultimately an epistemological issue. What reason do we have for believing that we are in a simulation or that we're not in a stimulation? Let's just go with that. That. So reason for thinking that we're in a simulation, there are various reasons that people might propose, but the one he puts forward, which I think is by far the most robust reason, which is that we run a bunch of simulations in which things like ours exist.

0:45:03 Thomas: Sure. Okay. Yeah. It's probability. Yeah.

0:45:43 Phil: Right. So if we don't do that, then there are two options. We either reach a post-human state and decide not to run the simulations or we die out. So it's it's not that, I mean, the way he puts it here, if we are not a simulation, then, uh, you know, there's, you know, goes through some options, then concludes there's a very high probability that at least one of the disjuncts is true. Well, that, the way she stated it there, is is absolutely consistent with us not living in a simulation because it could be the case that three is false, right, and one and two is true, but ultimately the, what the argument is attempting to do is, well,

0:46:16 Thomas: No, but hold on. I think that misses the fact that I think where the error is, is by saying if we're not in a simulation, then the empirical evidence still suggests that only one or two could be true, because I think there could be a four or five, six, seven to infinity of we do actually.

0:46:33 Phil: Yeah, if we're not a simulation, then then if we're almost certainly not a stimulation, then three is almost certainly false.

0:46:39 Thomas: Right. That's fine.

0:46:40 Phil: Right. So that leaves that leaves one or two. And it's still.

0:46:43 Thomas: Yeah, but I'm saying it's inaccurate to leave one or deux because there's no reason then to conclude that. Okay, so 1 and 2, if you go back to what they are, and again, I'm totally happy to be wrong here. I just want to make sure I actually am. So 1 and 3, the reason they exist, essentially, to put it in kind of a simplistic terms and maybe help people follow this a little bit in my mind, is there ways to explain why there aren't a lot of simulations, right? Can we agree on that summary? Like the reason he comes up with one or two is that they are, are explanations for why there wouldn't be vastly more Sims than humans, because they're, they're both geared around, we don't get to the stage or for whatever reason, we decide not to run these simulations, right?

0:46:53 Phil: Yeah, yeah. Sure, sure.

0:47:35 Phil: Yeah. So just to to nitpick briefly, I wouldn't say they're explanations, but I think I think the I I wouldn't say they're explanations. I would say that they're alternative scenarios. And again, the ultimate idea is that thinking as carefully as possible about this issue, it seems like there really are, or so Bostrom would argue, there really only three possibilities for the future. Nature. Either we die out, or we reach a post-human state and don't run simulations, or reach a post human state and run a bunch of simulations, in which case we're in one, probably in one.

0:47:38 Thomas: Sure. Really? Okay. Sure. Sure.

0:48:14 Thomas: I get that. Yeah.

0:48:15 Phil: Right. So it's not that they're just alternatives. I don't think they're they explain.

0:48:19 Thomas: They are alternatives. The only reason that they're the only three options is if you're taking the possibility that we're in a simulation. If we knew we were not in a stimulation, there's no reason for those to, like, again, if we knew we Were Not in a Simulation, there is no reason to contain it to only those three because we're granting it. I think this is where the circular nature comes out.

0:48:21 Phil: Yeah. Right.

0:48:41 Phil: But how would we know that? Right. Okay, So maybe we've pinpointed part of the the issue because you shouldn't be granting it, I mean, Bostrom does in his response here, but it's it's very for a very specific reason to to respond to this, but but ultimately the issue is you don't know whether you're in a simulation or not and so so how can I weigh the, you know, how can I be a rational person and weigh the probability that I'm a sim and this is all all, you know, some, uh, just ones and zeros all around me, or that, what are the other options, that we go extinct, or we reach a post-human state and just, you know, are, are so ethical or whatever that we decide not to run a bunch of simulations.

So then that's where this argument comes in to help us to, to determine right now, we don't, we don't really have strong evidence. I mean, I think there's fairly strong evidence for disjunct one, that we, some we, sort of existential catastrophe happens, unclear what sort of independent evidence we'd have for two. And for three, there seems to be some evidence that this may very well be the case because our computational systems are getting more and more powerful, humans are morally flawed.

0:48:55 Thomas: Mm-hmm. Right. Sure.

0:49:56 Phil: So we probably will want to simulate some universes in the future. So independent evidence aside, you should assign a 33% probability to each of these disjuncts being true. Now, if we advance into the future and we run a whole bunch of simulations, we should significantly increase the probability that the third disjunction is true. In fact, if you run hundreds of trillions of simulations that probably should go basically up to one. We should be 100% certain, more or less 100% that we're in a simulation. The idea is that we don't know, and the argument is sort of constraining the way we can think about the future. The simulation hypothesis, in particular, is saying that in the event that we run a bunch of simulations, that will constitute evidence for thinking that we are probably simulated, given the indifference principle, the fact that we, you know, don't Well, no. Reality would look the exact same if we're simulated or if we were non-simulated, given a high enough resolution simulation.

0:50:52 Thomas: Yeah. Right. Yeah. Well, unfortunately, I think maybe I won't be able to restate my objection in a way that works because I think, you just restated the argument again, which makes me worry that I did. So I think what I worry about what's happening like on a macro level to try to describe it, it's like I said before, if we're in a simulation, we can't conclude things things about reality. I mean, we can't, and I think it's even weirder for on in this simulation that we may or may not be in, if we get new evidence that that somehow tells us even more about reality, even though we have no direct access to it.

0:51:15 Phil: Yeah. Yeah.

0:51:34 Thomas: I means, if you're, imagine, okay, sure.

0:51:34 Phil: Well, let me just add real fast. Just real fast, so if we are in a simulation, then the simulation argument works, Because the simulation is just saying that one of the three is true. So if you assume we're in a situation that the third is true, so the simulation argument is sound.

0:51:46 Thomas: Sure, sure, sure. Yeah, yeah. Absolutely. So, But if we're not in a simulation, I don't believe it's sound because if you're allowed, and that's that's why I'm so frustrated here, because if you are allowed to say if we're in a simulation then three is true and that is totally fine, but we are looking at both the possibilities.

The other possibility is if we are not in a simulation, and then you claim, and he claims, that the simulation argument still works, I say no, because you can't use the fact that we are in a simulation to say the simulation works on one thing, and then then you say, if we're not in a simulation, it still works, even though we're ignoring the fact that one and two aren't the only possibilities in that case. If we're no longer constrained to one and two, we could advance, we could create a zillion simulations, and that's fine because we knew we weren't in a simulation all along. I'm not sure I'll be able to restate it any better. But yeah. Do you get that little trick, though? Like you were allowed for A to say, oh, we're in a simulation, so that's fine. Therefore, three is true. But for B, when I say, Oh, if we're not in a stimulation, I wasn't being allowed to conclude that therefore one, we are not constrained to one and two,

0:51:50 Phil: Okay, right. Yeah, think the idea is that if we're not in a simulation, then three must be false. Because if three is true, then it would contradict the assumption that we've made that we're not in simulation, right?

0:53:09 Thomas: That's fine. Of course. Yeah.

0:53:19 Phil: So that leaves one one or two.

0:53:18 Thomas: Yeah, but we're we're not, I, but that's assuming the argument works, but it doesn't work if we're granting that we're in a simulation. We don't just leave one and two. I think that's an error there. But I mean, if, if if I'm wrong, I'm right. I don't know. You know, it sounds like we're at a bit of an impasse, but sure.

0:53:34 Phil: Well, okay, let me see this. What could you imagine is not a version of either one or 2? So let's assume we're not in a simulation, but imagine some future that that, uh, imagine some future that is a version of disjunct one or just on two.

0:53:46 Thomas: Oh, easily. I can imagine

0:53:56 Phil: Well, I'm sorry.

0:53:55 Thomas: A future that is.

0:53:57 Phil: Imagine one that is, is is, not just a variation of disjunked one and disjunced two. Sorry.

0:54:02 Thomas: Oh easy. The only reason he limited us to disjunct one and to in the first place is to is to account for the fact that the, the odds should be that we're in a simulation, but if we're granting that we are not, then how about disjunct 7 million, we live to a certain extent, and then we create a trillion simulations. That's fine.

0:54:23 Phil: Well we can't create simulations because we're assuming,

0:54:25 Thomas: No, we can. We're only assuming that we aren't in one. We're not assuming that can't create one.

0:54:30 Phil: Right. But, but, But OK, so this gets at the issue. If we, if we, no, I don't think it is. I genuinely don't it is, I mean, the idea is there are two claims that are incompatible. One is that we run a lot of simulations, and the other is that were not in a simulation.

0:54:35 Thomas: I'm telling you, it's circular. It's circular! Okay, okay, okay. Right. Right.

0:54:48 Phil: So if you're assuming, so these are, this is a disjunct, right? It's one or the other, is true, not both. So if you're assuming that the first one is true, that we're not in a simulation, then you have to assume that the second, namely that we run a bunch of simulations, is false. That's the whole essence of the argument.

0:55:10 Thomas: And what you just said there is why I don't think it works, because that logic is totally flawed. There's no reason if we're not in a simulation to assume that only one in two is true. There is plenty of reason to assume we will develop simulations. I think that actually shows why it's a trick.

0:55:25 Phil: But if we do, then that suggests yes, we are in a simulation. So, so then there's, there's a paradox. Totally get how it's confusing, but I mean, I would just emphasize that, so if you're looking at, you know, question four B the assumption is we're not in a stimulation that is incompatible with us creating a lot of simulations. So you're automatically saying that disjunct three is false.

0:55:50 Phil: And then I would say that There is literally no future scenario that I can imagine that isn't some variation of one or some variation or two. Maybe we live a billion years and then we go extinct, and maybe in that time we never reach an advanced state. Maybe we reach an advance state Trump's not elected and so we don't have an existential catastrophe.

0:55:48 Thomas: If you're saying that, then you're... Sure. Go ahead. Sorry.

0:56:02 Thomas: Yeah.

0:56:11 Thomas: So my My opinion is you just begged the question. You begged the questions by granting the argument. So, we're looking at whether the argument's true, but you said the fact that we have not No,

0:56:13 Phil: Yeah.

0:56:22 Phil: No, no, there are two different arguments, right? Right. So there is an argument for disjunct three. And then there is the simulation argument, which is that one of the three is true. And in both of these cases, one of them, at least one of three is going to be true, whether you assume X hypothesized that we're in a simulation or whether we're not a simulation and then the mechanism.

0:56:42 Thomas: No, no. Hold on, hold on. I've got to stop you because the only reason you're limiting us to one through three in scenario B in which we're not in a simulation is because you've assumed that the argument still works. You've assumed that if we're not in a simulation, that that's evidence that we didn't create a bunch of simulations later on, and I think that's circular.

0:56:45 Phil: OK, no, I'm not saying it's evidence. I'm saying it it's it's incompatible.

0:57:06 Thomas: You said it's incompatible. Yeah, How would you know that without granting the argument that we're evaluating? Evaluating argument in itself, it proves that the argument that we're even looking at is uh, is what proves that claim that you just made. So I don't believe we can grant that in order to evaluate it, but it's a, it's okay if we're at an impact, like seriously, I mean, I, I I would, if I were, if I were a betting person, I would bet that there's a way that I'm wrong, but like I'm, I'm personally not getting it, but I apologize if that's frustrating, I'm sorry about that.

0:57:08 Phil: Yes. Yes. Yeah.

0:57:40 Phil: No,

0:57:41 Thomas: But yeah,

0:57:41 Phil: No, no. Honestly, let say also for listeners, I mean, the simulation argument is difficult to wrap your head I around. Mean honestly I've been reading about this stuff for a long time and even recently after being somewhat of a veteran of these issues and still sort of of like seeing it in new ways and sort of understanding the logic. I mean, I can also say that there have been a lot of philosophers who have been skeptical of this argument and looked at it. And to date, this is true. I think nobody has published a refutation of it.

0:58:19 Thomas: I totally agree. Strong evidence that I'm wrong. I guess I'm not grasping how I'm wrong. And on a macro level, though, doesn't it seem... I would have expected a better response than this number for on the simulationargument.com slash faq.html that I'll link. I Would, would have expected a better response indeed I were wrong. Because don't you see how on a macro level, we're using facts about the world that we're in to conclude whether or not we're in the real world, like to isolate only those three options. Again, I know it's not proving that we are in a simulation, but it is narrowing it down to those three option based on contingent facts, you know, like I just think about like what it, let's, let's say, you know, Super Mario was a simulation in which it was complex enough that those, those characters actually had, you know, actual experiences and

0:58:37 Phil: Well, let

0:59:15 Thomas: Consciousness, what would they conclude about the nature of reality? They would conclude that you jump up and hit a little box and then a mushroom comes out, like they would have totally wrong conclusions about reality, they wouldn't be able to reason about the, the world, world outside of it. And that's where I think it becomes a problem. And it makes me really skeptical about it. But who knows? I don't know. You can take another stab at it if you want. Sure.

0:59:41 Phil: Me just Yeah, sure. I don't think for the argument to hold, you don't need to reason about the outside world. Right. So, I mean, you know,

0:59:49 Thomas: I don' t know. Because he says here, here' s what he says right here. The claim that we cannot have any information about the underlying reality if we are in a simulation is false. It's true. We We can know that simulations exist if we're in a simulation, but we can't know that for sure because we don't know if we are in one or not. You know what I mean?

1:00:10 Phil: Yeah. So while I'm saying the argument itself isn't based on you simply don't play into the argument itself. So once you conclude that disjunct three is true, there are certain things you must infer about the higher level. Mainly the most significant is that they must have the capacity to simulate universes like yours. But the argument, itself, isn't drawing from that higher level, it's only saying that imagine imagine the future. We create a bunch of simulations. Then given the truth of functionalism and the indifference principle, then we would have to conclude for probabilistic reasons that we must be, we're much more likely, if everybody, imagine, you know, world, there's 100 trillion people in it, conscious beings and only 1 billion, uh, are on our level, if everybody bet that they're on our label, let's say top reality, just about everybody would get it wrong.

1:00:21 Thomas: Hmm. Sure. Sure. Right. See,

1:01:24 Phil: But if well

1:01:24 Thomas: That's the step where we would be concluding things about the reality above us based on our reality. Because if we make a bunch of simulations and we say, okay, let's just deal with reality. Let's say we're reality B or something, and we don't know if there's a reality A or not.

1:01:31 Phil: We would, yeah. Yeah

1:01:40 Thomas: We don't now. If we're in reality B and we create reality C that has a gajillion humans or whatever, sentient beings and conscious creatures, whatever, sims, we're not going to say, well, we might be one of them. We know we're not one of them. So you're using that fact to conclude, conclude that there's something, that reality A has a bunch of simulations, and therefore we might expect that we're actually sims. And that feels like it must be wrong.

1:01:40 Phil: Yeah Yeah.

1:02:11 Phil: Right. So I agree that it feels weird.

1:02:18 Thomas: What? Tell me where I made my mistake, man. Tell me.

1:02:20 Phil: Here's another way to think about it. There are some good arguments out there that result in very counterintuitive conclusions, right? Of the features of a good argument,

1:02:21 Thomas: Sure. Sure, sure. I don't believe I'm only relying on intuition just so, I mean, sure, absolutely my logic could be wrong, but I do do believe I'm using logic, not just intuition. I was just trying to appeal to the macro level of what I'm saying rather than like the detailed problem with our A, B, you know, one through three thing. That's the only reason I was appealing to intuition.

1:02:33 Phil: Oh.

1:02:53 Phil: Yeah, no, no, no, I mean, most of philosophy is based on intuitions, you know, um, you know, what are your intuitions about the, you know, the, how, how this concept can be analyzed, uh, you know, that, that's that sort of, you know, has been at least a foundational question in a lake philosophy. It's very intuition based, so I don't think there's anything, most of moral philosophy is about intuitions, you know.

1:02:58 Thomas: Right. Yeah.

1:03:14 Phil: Um, and trying to navigate. Yeah.

1:03:17 Thomas: I know. I just didn't, I didn't want the straw man to be, I was only saying this is counterintuitive, therefore it's wrong because I'm not. But OK. Yeah.

1:03:18 Phil: Yeah. Yeah. Okay, so, so one way to think about, so one of the, the features of good arguments is that sometimes you will get conclusions that seem, they just seem wrong. I mean, you could say same thing about like the ontological argument or you know that, that argument, like, just seems wrong, you know, I'm not sure where it goes wrong, but I just feel like something's not right. I think a lot of people get the same, uh, have the same reaction to the simulation argument, but here's the, here's interesting part.

And I think the reason why philosophers take it seriously, when you zoom in to all of the little gears and springs that are in the argument and you examine each one, doesn't seem to be anything wrong wrong with any of these little steps. And yet you get to the end. I mean, by analogy, maybe this is kind of a weird analogy, but maybe it works. Like, you know, there's something called the heap paradox, right? Which is that, you Know, if I have a grain of salt, and then I pick up another grain, put it next to it, do I have the heap? No, obviously not. Adding one grain of salt does not create a heap. So let me add another one. Does that create a heap?

No. No. It seems like there's no point at which adding a grain of salt results in a heap. And yet at some point you have a heap, you know, that's the strangeness of the situation. So I feel like there's a somewhat similar thing going on with the simulation. When you zoom in, you look at, here's the principle of indifference. Okay. That's totally robust. Like nobody's going to argue that that's wrong.

1:04:11 Thomas: Yeah, right.

1:04:58 Phil: Here's the, you know, what are we indifferent about, whether or not we exist in vivo or in machina, is that right? Well, yeah, that seems totally right. Um, you know, if, if functionalism is true, and there was a sufficiently high resolution, uh, simulation, which of course might not be the case, that's part of the, the, uh, the way the argument could be falsified. Um, so you sort of look at all of these, the intricacies of it, and you can't spot, you know, know there's no obvious error, and when you just, when you take all those ingredients and you stuff them together, you end up with this, this totally weird conclusion, you know, which is that one of the three are true, where one of the disjuncts says you're almost certainly in the simulation.

1:05:39 Phil: So I, so I, I hope that helps that I to explain why it is philosophers take it seriously.

1:05:39 Thomas: Yeah yeah. I think I would need I will I will need to pinpoint where where my objection is in the original springs and gears. All right. Well, thanks for taking us through that for one for sure. And also thanks for trying to answer my, my questions. I'm, I will go back to the drawing board, try to get a more technical version of them. Uh, and then maybe, maybe not, maybe it will just occur to me, uh, that I'm wrong. And I would love to, that wouldn't be the first time. So, so either way, but thanks for being a good sport and fielding those. I appreciate it a lot.

1:05:45 Phil: Okay.

1:06:15 Phil: Yeah, absolutely. It's a real pleasure. I also would welcome emails from listeners who have questions. You know, like I said before, the argument is not, it's genuinely not easy to grasp and to navigate. So I think struggling is totally normal for everybody who comes across it the first time, because it has a super perplexing conclusion and a lot of like kind of logical acrobatics that happen.

1:06:44 Thomas: Well, thank you. I, I appreciate it. Uh, Phil Torres. And again, go, go ahead and give away your website one more time if you'd like to.

1:06:51 Phil: Yeah. So you can you can check out publications and stuff on xrisksinstitute.com.

1:06:58 Thomas: Awesome. All right. Phil Torres, it was a good one. Thank you very much.

1:06:59 Phil: Yeah. Yeah. Thanks a lot. Really appreciate it.

1:07:29 Thomas: Cool stuff. I hope that you enjoyed it as much I did. Reminds me of when I tried to argue with Sean Carroll about Boltzmann brains, that one, it turned out, I figured out the mistake I was making, and I was wrong, and, I, I, I figured I was, I just, need, it, there's, you can't just take things on authority, or, you know, you have to actually understand things, in order to be able to speak about them intelligently, and in this case I'm going to do the same thing. I do think I'm not convinced that there isn't a circular issue with this argument. I'm actually going to try to work on it and see if I can figure out where my issue is and I'll probably end up being wrong and that's cool. I don't mind that at all but I really do want to actually understand why it's wrong.

1:08:18 Thomas: So if anyone wants to help me out with that, if you have a thought, or if maybe you agree with me, that, probably not, but that's okay, um, feel free to comment, send me an email or something. Now it's time to thank new patrons, got some new patrons this week, thank you so much, Martin Jonsson, Dustin Clouse, Chris Lukowski, and Denys Zakhozhyy. Thank you guys for being new patrons. I really, really do appreciate it, seriously, it means a ton to me, and now, I gotta thank my all-time greats, and they are, Hunter Ash, Sakashite Fukasami, Jesse, Yalong Blackcock, The Inciting Incident Podcast, Kaylee Reiser, Jonathan Moyet, Jay Aldenwalt, Ian Rosano, Robert Shepis, Blackhawks, Erosign Sharks, Sammy, Steve Sigman, Pretheist, Roger Metcalf, Adrian Borschoff, Peter Skelton, Brian Gareford, Dale, and Dale, thank you guys for being my top patrons, you are the best.

1:09:10 Thomas: That's it for this week, thank you so much for listening to the show, assuming you made it this far, and I will see you guys next week. Week.