Welcome to the June 2025 Ask Me Anything episode of Mindscape! These monthly excursions are funded by Patreon supporters (who are also the ones asking the questions). We take questions asked by Patreons, whittle them down to a more manageable number -- based primarily on whether I have anything interesting to say about them, not whether the questions themselves are good -- and sometimes group them together if they are about a similar topic. Enjoy!
Support Mindscape on Patreon.
Click to Show AMA Questions Only
Click to Show Full Transcript
0:00:01.6 Sean Carroll: Hello, everyone. Welcome to the June 2025, Ask Me Anything edition of the Mindscape podcast. I'm your host, Sean Carroll. Recording this, a week before the AMA goes live because we have a busy week here today in Baltimore this week. This is the week of the first inaugural ever Natural Philosophy Symposium to be hosted here in Baltimore, organized by Jenann Ismael and myself, the two of whom constitute the Natural Philosophy Forum at Johns Hopkins. So it's a three-day conference. It'll already be over by the time you're listening to this. Hopefully it wasn't a disaster. I'm hoping for great things. If you're not able to go, then that's okay. We're gonna record everything. It'll be on YouTube at some point. We have an excellent crew of young Hopkins students, film students, who are doing that work, so it should be a very good record of what happened during the symposium. We're trying to make natural philosophy a thing. In other words, we're trying to let people understand what it means when we say natural philosophy, that we're not doing history. We're not talking about Galileo and Newton or whatever, even though we use the same label.
0:01:12.6 Sean Carroll: We're doing what they would have done. We are trying to move our understanding of the universe forward, in ways that are scientific, that are based on data, but that are also philosophical. That are reasoned very carefully through, that are patient with all of the underlying questions. And it's not just philosophy of physics that we're doing here. We have people in consciousness, in AI, in biology, in social sciences, in complexity. It's quite a number of areas of science that have significant overlap with philosophy, and we have an all-star lineup coming to talk about them, some really great scientists, some really great philosophers. Of course, we've chosen the ones who we think can get along, can get interested in the questions that each other think are interesting, so hopefully it'll go well. We'll let you know how it goes. One tiny optimistic note, is that Nima Arkani-Hamed, who is one of the speakers, told me that his title will be, Two Cheers for "Shut Up and Calculate." Nima, of course, is a hard-nosed physicist, but the good news is that he has previously given a talk called Three Cheers for Shut Up and Calculate, but he tells me that his stance has softened. [laughter]
0:02:26.0 Sean Carroll: So we'll see exactly what that means. It makes me optimistic for what's going on. The other news, of course, I wanna tell you is that I did appear on the Piers Morgan Uncensored show on YouTube. It's not on TV anymore, it's on YouTube, but that's perfectly okay. Talking with Eric Weinstein. Not a productive conversation, no real reason for you to ever watch it. I'm just letting you know because it's out there. My goal was to reach out to people who are wondering whether or not it's true, that the physics establishment is ossified and unable to listen to new ideas. So I sort of tried to spell out, on the one hand, why physicists think the theories that they think are interesting are interesting, based on physics arguments. And also, if you did, hypothetically, want to make an impact on physics as an outsider, what it is you would have to do. So I tried to be constructive in that way. Halfway through, Eric just went completely nonlinear and went on an unhinged rant of personal insults against me, so I was remembering why I don't usually do this kind of thing.
0:03:31.0 Sean Carroll: But it is important to occasionally do this kind of thing. I'm not gonna do exactly that ever again, but there's a reason why I did it. I wanted to reach audiences that I wouldn't ordinarily reach. And as you know, if you listen to Mindscape, science is in a terrible position right now, here in the United States and throughout the world. The funding is completely being cut. And it's not Eric's fault in particular, but the movement of which he is a part, which lowers people's respect for and understanding of academics. It's not just science. It's all of academia and all of higher education. If you start to think of it as a conspiracy, to keep out the really interesting ideas, then you're not gonna mind when it's brutally cut. So I wanted to try to reach those people in some way if I could. Who knows whether I actually succeeded or not. I don't think I did an especially good job. There's two ways to be in that kind of mode. You can either just be totally enamored with the fighting, and the pugilism, and the shouting of invective and things like that, and that's never gonna be my style.
0:04:39.4 Sean Carroll: I'm interested in talking about the substance. But if that's gonna be your style, then you should try to be super-duper diplomatic and peacemaking, and I'm not good at that either. I let my disdain show through sometimes in sarcastic or snarky comments. So, I gave it a shot. There you go. That's the best any of us can do.
0:04:58.5 Sean Carroll: Finally, a little tiny piece of news is, that I had the pleasure and the unusual thing of giving the commencement address for the master's degree ceremony here at Johns Hopkins. That was a lot of fun. It was very nice to be invited, and it's also a challenging gig I realized, when I thought about it. You wanna say some platitudes to the graduates, et cetera. But also we are in a weird time right now. You can't just let the collapse of higher education and respect for science and things like that go unmentioned when you're giving a commencement address, but also I don't think you want to raise a fuss. It's not like you're a student. There's these stories now of students who are talking about Gaza and the Palestinians and are being punished by their universities if they mention that in their commencement addresses, which is completely terrible and embarrassing for those universities. They should feel bad.
0:05:53.0 Sean Carroll: Good for the students. I'm totally on their side. But I'm not a student. I'm there as a guest, and the focus should be on the students, and I'm not there to raise a ruckus or upset people in the audience, even if it would be good for them. But I could sort of glancingly allude to the situation we're in, while sort of exhorting the students to try to make things better. Sadly for me, the same week here nearby, not at Johns Hopkins, but at the University of Maryland, the commencement speaker was Kermit the Frog. And Kermit kicked my ass in terms of being a good commencement speaker, I gotta say. Kermit was both inspirational, but also funny and interesting, and that's okay. Kermit is a pro. He's been around. He has really polished his delivery in these situations, so I can't really get upset if Kermit did a better job than me. I can only aspire to be as good at this as Kermit the Frog. And finally, of course, this is an AMA episode, so thank you as always to the Patreon supporters of the Mindscape podcast. They're the ones who make it possible. If you wanna ask AMA questions, you could be a Patreon supporter of the Mindscape podcast. Go to patreon.com/seanmcarroll, kick in a few bucks, join the group, and you can get your name read out loud by me, if I happen to pick your question. Wouldn't that be fun? So with that, let's go.
[music]
0:07:28.1 Sean Carroll: David Wright says, "Your interview with Branden Fitelson triggered a lot of great insights into the nature of what we believe is true, and how we confirm our hypotheses. My question is, how do we distinguish scientific truth from common sense? Both are based on socialized priors. Common sense is not filtered for relevance bias, and other fallacies mentioned in the interview. Is this a good way to look at this distinction?" Honestly, I'm not fond of invoking common sense at all in these discussions. People try to do it. I've known very, very smart people who say that science is really just common sense made a little bit more rigorous and careful and objective. I don't think that's the right framing for this way of talking, because common sense has good aspects and bad aspects. Common sense has good aspects in that, it's sense. It's trying to be at least reasonable about what the universe is, but it has bad aspects in that it's common.
0:08:23.7 Sean Carroll: It's what, is right in front of your face, and what is right in front of your face is not always the best scientific way of thinking about the world. Common sense might lead you to Aristotelian physics, but it's not gonna lead you to Newtonian physics, even with extra layers of being careful and testing hypotheses and things like that. So I do think that it's better. For me, I prefer to think in terms of the manifest image, scientific image version of these things, that is to say, or the folk physics versus scientific physics version of thinking of these things. That is to say, all of us have some model of the world. Judea Pearl, back when we talked to him, and Alison Gopnik, when we talked to her, emphasized how children, little babies, are constantly modeling the world. They're trying to figure out the causal map of reality. If I do this, what happens? And we all carry that in our minds, in our brains, when we go through life. We have expectations for what will happen next, features of how the world fits together in different parts, and so forth.
0:09:34.6 Sean Carroll: And is that common sense? I don't know, but I think it's a little bit better than common sense. I think the way to think of it is not as common sense as the sort of epistemological flavor, a way of thinking. Rather than that way of thinking, I like to think of the physical model that we have in the manifest image of the world, because common sense can mislead us in all sorts of ways. Common sense does not tell us that the Earth goes around the Sun. [chuckle] It tells us the opposite thing. Science is not just common sense plus hypothesis testing, precisely because science remains open about the hypotheses until they are tested. Science says, even if a hypothesis seems to go against common sense, let's think very, very hard about what it's predictions are, and then compare them to what we observe. There's a famous story. I've told it already. I'm gonna tell it again.
0:10:35.3 Sean Carroll: Everettians love this story about Ludwig Wittgenstein and Elizabeth Anscombe. Elizabeth Anscombe was one of Wittgenstein's students, and she meets Wittgenstein on the grounds of, I think, Cambridge University, and Wittgenstein is staring up into the sky, and Anscombe says, "What are you looking at?" And he says, "Well, I'm wondering why people ever thought, that it made more sense to say the Sun goes around the Earth, than the Earth rotates." And Anscombe says, "Well, I suppose it's because it looks like the sun goes around the earth." And Wittgenstein says, "What would it have looked like if the earth were rotating?" [laughter] It would have looked the same thing. But it kind of fits in less well with our common sense. Even though the predictions are the same. So to me, that's more science-y. The science aspect of things is really taking seriously what the predictions of a theory are and using them to compare with the data rather than some consonance with our commonness, our common sense. The reason why Everettians love this, of course, is because when you say the world is just a vector in Hilbert space, people say it doesn't seem like that. And you can say, well, what would it seem like if it were? And then you can make an argument that it really is. So I don't really think that there's a direct line from common sense to science. I think that they're two different things.
0:12:01.8 Sean Carroll: Jonathan Peretz says, "It seems like most working theoretical physicists are sort of okay with loss of information as a result of quantum measurement. However, when it comes to black hole information loss, it's considered a bigger puzzle for some reason, if I get the right impression. Why is that? And isn't there a sense in which the formation of Hawking radiation due to virtual particle pairs and all that, constitute a quantum measurement also?" Not really. That last part is not really right. If you take a quantum mechanics course, you'll get a version of what I'm about to say. If you read a real book, a careful book like John von Neumann's book about the mathematical foundations of quantum mechanics, you'll get a clear version of this. In traditional textbook quantum mechanics, there are two different ways for the world, the quantum system, to evolve. One is when you're doing a measurement, and one is when you're not. When you're not doing a measurement, you're just obeying the Schrodinger equation. It's what's called unitary evolution. When you are doing a measurement, the wave function collapses. You get a measurement outcome.
0:13:03.8 Sean Carroll: There's a probability given by the Born rule. And so, why I'm rehearsing this, many of you have heard this before, but in textbook quantum mechanics, there is a sharp distinction between what the system does when you're not measuring it, and what it does when a measurement is performed. So Hawking radiation is not a measurement. That is something that happens according to the unitary laws of physics even when you're not measuring it. Now we can argue about what a measurement is. The good news is, that in realistic situations, in a laboratory or something like that, it doesn't matter. We all agree. If you want to really, really care about what measurements are, that's gonna depend on what your particular favorite formulation of quantum mechanics is. And that is going to affect whether or not, or I should say how much you care about the possibility that information is lost. In the traditional textbook version of quantum mechanics, information absolutely is lost as a result of quantum measurement, and it is not when you're not doing quantum measurement. Okay? So the problem with the black hole, is that if you take seriously sort of the best version of a guess as to what happens. I should say guess is not the right word.
0:14:20.7 Sean Carroll: If you take seriously Stephen Hawking's calculation of what happens when black holes evaporate, then information is lost even without any collapse of the wave function or measurement being performed. So that is one reason why physicists are a little bit bothered by that. It's one thing to say that information is lost when you do a measurement. It's another thing to say, the laws of physics lose information even without a measurement being done, because we have the measurement problem. Even if we don't sit here advocating our favorite solution to the measurement problem, we sort of circle around measurement and say, we don't completely understand that, so maybe information is lost or maybe it's not, because under certain formulations of quantum mechanics, like Bohmian mechanics or many worlds, it's just not lost. Under other formulations, like spontaneous collapse models, it is truly lost. And so, people are comfortable doing that and then worried about why, even without a measurement, information seems to be lost in black hole evaporation. Of course, I'm actually a little, I have the opinion that people are cheating a little bit here for exactly the reason I think, that you're asking the question in the first place.
0:15:32.0 Sean Carroll: Like, if you allow information loss in one kind of category, even if you say, well, we don't understand it perfectly, we'll understand it later, why not in the other category also? But none of these arguments is airtight. These are all suggestive. Some way, sometimes how it's put is quantum mechanics by itself without including the measurement, just the ordinary unitary evolution, preserves information, and general relativity by itself preserves information. So it would be weird, if the combination of them to make quantum gravity didn't preserve information, didn't conserve information. That's a bad argument actually, because it's completely plausible, that quantum gravity does not conserve information, but the separate limits that we're commonly looking at of quantum mechanics of small objects and classical gravity of big objects do conserve information. So I think this is all sort of suggestive and hand-wavy. It does go beyond that. I should give more credit. People have tried to be very explicit about what the consequences would be, if black holes destroyed information. So there's a famous paper, I'm gonna get it not exactly right. Maybe Bankstein, Peskin, Susskind, some subset of those people, wrote a paper claiming that if black holes destroyed information, then tiny Planck-scale virtual black holes inside Feynman diagrams, would lead to macroscopic destruction of information.
0:17:02.4 Sean Carroll: And you would have noticed that a long time ago. I think that's controversial. People don't quite agree. But it's of a piece with a general expectation that when you break something fundamental about your theory, there might be all sorts of other unanticipated consequences. So it's safer to imagine that our cherished beliefs are not broken, unless we absolutely are forced to accept them.
0:17:26.8 Sean Carroll: Mark Sleight says, "I love your talk with James Ladyman and now proudly wear the structural realist label and talk about it with everyone I meet. Not really. You seem to want to leave room for some kind of non-structural existence or properties, perhaps even leaning in the direction that is probably something like that, if I don't misread you. Am I correct to assume that you would have to commit to being an epiphenomenologist, I don't know how to pronounce this, epiphenomenalist about this? It seems to me that I'll talk about the non-structural supervenes on the structural relationships and how they unfold. Therefore, if you respond to this, your response will be exactly the same whether the non-structural exists or it doesn't. Does that make sense?"
0:18:06.6 Sean Carroll: I don't know whether that makes sense or not. Even though I am happily talking about it frequently, I'm still not 100% sure how to distinguish between what exists and what doesn't exist. I have some half-hearted ways of thinking about that, but I'm not wedded to them. But you're right, I do sort of both want to be a structural realist and a kind of more materialist realist at the same time. And the reconciliation is simply that, I think we don't know the fundamental material structure of reality. We're trying to learn about it. We're trying to come closer and closer to it. But part of the motivation for being a structural realist was the idea that if our theories in fundamental physics, keep getting better, and we keep completely overturning the previous ontology along the way, from Aristotle to Newton to relativity to quantum mechanics, how can you be a realist about anything if you're gonna throw it away later when a better theory comes along? What is it that you're realist about? And the structural realist answer to that, is that there are commonalities shared by quantum mechanics, relativity, Newtonian mechanics, even Aristotle. We're all predicting the motion of the planets in the sky more or less the same way.
0:19:22.6 Sean Carroll: That is a pattern, a structure that is preserved as you go past these different theoretical improvements. And so, the structural realist says, even if I don't know the once-and-for-all final ontology of reality, there's still something that I know and that something that I know is real. And I completely buy that. I'm not even sure if I'm saying it the way James Ladyman or a real structural realist would say it. This is my way of saying it. So yes, I completely buy that.
0:19:49.9 Sean Carroll: I also believe, that there is a bottom, that there is some one true theory of reality, even if we don't know it yet, and we might be very, very, very far from knowing it yet. I have no idea. Without that one true theory of reality, I don't know why we would be able to be so confident that these structures exist, or at least not confident, but I don't think we would have a good explanation for why these structures exist so robustly. To me, this story of relativity, quantum mechanics, Newton, et cetera, is true because these are different patterns within the one underlying reality.
0:20:28.1 Sean Carroll: So I think that there is a world that exists, and that world exhibits structures, and it's those structures that we have a more direct grasp on along the way to getting the once-and-for-all final theory. I'm gonna group two questions together.
0:20:43.0 Sean Carroll: Paul Soldera says, "In your recent episode with Karen Lloyd, both of you talked briefly about the possibility of life evolving more complex cellular structures while in this underground biosphere. If this did happen and complex cellular structures formed underground first, what does this do to our estimates of life forming on other planets? Karen seemed to say that life is likely to be much more prevalent than we first thought. Do you agree? And Robo says, "The Karen Lloyd episode is definitely in my top five. The passive life forms that she described just hanging around waiting for some chemical reaction to happen nearby, strains my worldview, resulting in an update of my priors for life being found in other worlds, maybe during my lifetime. Did her presentation have a similar effect on you?" On the one hand, I wanna sort of prevaricate about this question. On the one hand, yes, I think that the evidence that we get from studying these life forms in extreme conditions, adds a little bit of hope to thinking that life can exist on all sorts of different conditions, even if maybe flourishing would be an exaggeration.
0:21:47.2 Sean Carroll: Karen was very clear that life could right now presently exist on Mars, even if there's not abundant running water or things like that, because the timescales for this kind of underground life, these intra-terrestrials, they can just persist a long time but move and metabolize very, very slowly, just leeching off of the tiny energy gradients and the occasional stray electron, and nutrients that fall down from the surface and things like that. It adds to our catalog of conditions under which life can exist. But on the other hand, I think that our catalog of the conditions under which life can exist has no reason to be complete at all. I never thought that we knew where life could exist and where life could not exist. I was always very suspicious of this discourse about Goldilocks zones and things like that, that always seemed to be very, very heavily oriented toward life as we know it, which is obviously a little distortive in the space of all possible life. So on the one hand, this absolutely is a new and interesting way for life to exist, new in some sense. It's been around for a long time. But it doesn't dramatically change my credences that life could exist elsewhere. It's just an interesting specific place that life could exist.
0:23:09.7 Sean Carroll: Okay. Again, I'm going to group two questions together. Sean Bentley says, "Any update on science funding in the US? For example, NIH research grants, et cetera. In your estimation, has the situation gotten any better, any worse, or about the same since your solo episode a few months ago? And Kyle Stevens says, "The DOGE cuts to government grants seem to have fallen out of the news cycle. What direct effects are we seeing now that a month or two has passed?" So it is very depressing to me to hear someone say that the DOGE cuts have fallen out of the news cycle. They've not fallen out of my news cycle because they affect my life rather directly and dramatically. I'm very well aware of them. So the fact that to someone else they have fallen out means that the real world out there is not being kept up to date on these things. To Sean's question, it's become much, much worse than my solo episode a couple months ago, because the solo episode was prompted by the idea that they were going to lower the overhead rate on grants from the National Institute of Health, from the NIH.
0:24:10.4 Sean Carroll: So that's clearly it was prefiguring much worse things to come, but it was one relatively localized thing. So the grants themselves, the actual grants that had been given out, might imagine to become the same or to stay the same level, but the overhead, the indirect costs, which go to the universities to keep up the infrastructure and the buildings and things, were going to be cut. So that would have been disastrous all by itself, but it was just the NIH and it was just indirect cuts. Since then, the cuts have spread to all the other agencies, and it's not just indirect costs or overhead costs. It's everything. It's the very basic things. So the New York Times, just a couple days ago, had a very nice graphic, an article with some graphics that explored the cuts to the National Science Foundation, the NSF. So there's many different government departments that give science funding. NIH is one of them, National Institutes of Health, but NSF, National Science Foundation, is another one. There's also Department of Energy, et cetera, et cetera. The NSF, the National Science Foundation, is the most broad-based one. It gives money to all sorts of different scientific Disciplines, and so it's a good test case for what the current government thinks about this, and it's being devastated.
0:25:24.7 Sean Carroll: The overall grant so far funded for the National Science Foundation from this year versus last year, they're comparing what is actually happening. So these aren't promises. This is what is actually happening in 2025 versus what the average spending was between 2015 and 2024. So an almost 10-year average before. The NSF funding is down 51%, so they're cutting half of the money spent on science by the National Science Foundation. If you want some specific examples, it's not just woke DEI initiatives. Here's physics. Physics is down 85% in the National Science Foundation. Graduate education, in particular funding for graduate students, is down 100%. It was cut to zero in the most recent budget. Of course, they're cutting things like equity and things like that, but chemistry down 57%, math down 72%, materials research down 63%, biological infrastructure down 68%, chemical engineering down 71%.
0:26:31.9 Sean Carroll: It just goes on and on and on. The Office of Polar Programs that does science at the South Pole is down 88%. Earth sciences is down 80%. It's a complete destruction of American science. I don't know how to state it nearly as loudly and as alarmingly as it could be. It's not even the kind of thing that is going to be judged in it's effects over the course of weeks or months. It's going to be judged over the course of decades because talented students, both from the United States and elsewhere, will not want to stick around the United States in this situation. Even if they wanted to, there's not money to support them. Talented or promising programs in health and materials and chemistry and whatever, are going to be shut down. They are being shut down. Test animals in biological laboratories are being euthanized, because they can't afford to feed them and to keep them. There's a real danger that the Roman Space Telescope is not going to be launched. Just over and over and over again, this is gonna be completely devastating.
0:27:39.8 Sean Carroll: Say goodbye to the idea that the United States is the world leader in science for decades to come. Even if Trump only lasts four years and they immediately replace him with, 20 Democrats in a row, doing their best to refund science, it will take that long to rebuild trust, to rebuild infrastructure, to rebuild the knowledge base, and that optimistic scenario is unlikely to come true. So that's the world in which we live right now. In terms of me personally, what is it affecting? I have no friends who have had grants cuts. The university overall is suffering because we have less money and therefore searches for new faculty members have been cut, postdoc searches have been cut, the number of graduate students we can have have been cut. It's devastating. What can I say? I can't yell about it enough. Therefore, I'm gonna move on. [laughter]
0:28:28.7 Sean Carroll: Albin asks a priority question. Remember that Patreon supporters can ask one priority question during their lifetimes, and I have to trust them that they're only gonna pick one, and I will do my best to answer it. So Albin says, "I've been really into integrated information theory after your podcast with Christof Koch. From what I understand, the consciousness we experience comes from the phi max, which is the most integrated and complex information. Each subsystem of the brain can also have it's own value of phi. So here's what I'm thinking. Could this dominant phi simply be the result of evolution and natural selection? Maybe evolution shaped our brains in such a way that the best-suited phi controls the body and remains dominant, while all the other phi values from subsystems are suppressed through the same evolutionary process. Assuming our brains are made up of many overlapping and non-overlapping sets of subsystems, is it possible that there are multiple consciousnesses inside us? And the one we currently experience, just happens to be the phi max, with nothing inherently special about it. It would be similar to the many worlds idea but applied to consciousness."
0:29:31.1 Sean Carroll: I don't think it's at all similar to the many worlds idea. Sorry about that, Alvin. But otherwise, I'm on your wavelength a little bit here. I would tweak the nomenclature a little bit. I think it's pretty respectable to believe, that within our brains there are subsystems, there are modules that are doing different tasks, and somehow they come together to make our highest level conscious experience. Of course, that's all the work. Figuring out how they come together at that higher level. Daniel Dennett talked about this in our discussion with him a while ago, and we've had other people talk about similar things.
0:30:12.8 Sean Carroll: Whether or not you want to call those subsystems kind of lesser consciousnesses, I'm not so sure about that. My impression, and I truly don't claim to be an expert here, it's not something I looked into beyond the occasional podcast episode. My impression is that the point of IIT is not just that consciousness is phi max. For those of you who did not listen to the episode, phi is this number you're supposed to be able to calculate given some interacting system with information flowing back and forth between it, and you can calculate it for different subsystems of the system and for the system itself, and also for supersystems. Like bigger systems of which the system you care about is embedded inside. And the argument is that consciousness is located at the peak value because this phi measures the integrated information, and in many different people, for example, there's a lot of information, but it's not that integrated. In different subsystems of your brain, there's not that much information. It's in a single human brain, that you have the most integrated information, and therefore, that's what we identify as a single individual conscious agent. So my impression is that, the claim is supposed to be, that that is the conscious agent, that the IIT is not saying even if there are other subsystems of the brain for which there is a phi, the fact that they're embedded in a system with a larger phi means that they themselves are not conscious.
0:31:41.1 Sean Carroll: It's only the top level that is conscious. You can argue about that. I'm not saying that's true, but that's my impression about what it says. But I think that in that case, that's probably the right way to think. I wouldn't want to, if you have a car and it has an engine, it has wheels and a chassis and things like that, all these things are not mini cars. They're all parts that contribute to the carness of the whole. I think that's probably a closer analogy to what is going on in your brain, than to think that there's a whole bunch of like mini things getting together in kind of a parliamentary thing and then taking votes. But I'm not an expert on this, so that might be completely wrong. Or there might be some intermediate sense in which there is something like consciousness in the subsystems. That's something I just don't know too much about.
0:32:31.4 Sean Carroll: David Sotolongo says, "I'm curious about your political philosophy. One facet of that would be, assuming you're not an anarchist, is what do you think the aims of government ought to be? For example, a classical liberal might say, the only legitimate aim of government is to protect people's natural rights, while the utilitarian would say the aim of government is simply to improve people's lives. What do you think?
0:32:52.0 Sean Carroll: Yeah, I think perfectly obviously, good question. I am not either one of those, a classical liberal nor a utilitarian. I certainly think that, my political philosophy is, that government is the organized expression of our collective will. That is to say, there are things that a society wants to have done, that are naturally group level in some sense. Like you want roads, okay? I, as an individual, am not gonna go around building roads in my neighborhood. Number one, I don't have the capacity to do that. Number two, there's a lot of other people who will benefit from it. So we come together to cooperate and say, as a society, let's build some roads. But in the real world, you don't wanna have direct democracy for all these questions. You don't want every detail about the roads to be voted on by everyone in your community. So you pick some representative leadership, voila, you have a government. I think that is a perfectly legitimate thing to have, a government that does the things that you want done collectively. Whether it is national defense or education or health or social welfare things, many things that the government can actively do over and above the sort of classical liberal view of simply the night watchman state where you're just protecting people's natural rights.
0:34:18.0 Sean Carroll: Now the goal of that, of specifically what the government is supposed to do, I would not phrase in utilitarian terms. Because I do think that it is best, the best kind of government will acknowledge the existence of absolute rights that individuals have, and that is very anti-utilitarian. Any utilitarian worth their salt, thinks that there aren't any natural rights because whatever an individual might want, it is always possible that the greater good is gonna deny them that. So I think that having, this is exactly the argument that the people in the United States had, when they were writing the Constitution, and people pointed out, well, you need a Bill of Rights in the Constitution, 'cause all you're doing in the main text of the Constitution is saying what the government can do, and how the government does it. You're not saying what the government can't do, and I think that Jefferson and others were absolutely correct to say, "You need a list of things the government is not allowed to do, and those are rights that are kept by the people." So that is not a utilitarian way of thinking.
0:35:29.0 Sean Carroll: The utilitarian would just say, "There aren't any rights. There's whatever gets the greatest good for the greatest number, even if some individuals have to be sacrificed along the way." You can argue whether in practice, governments live up to this goal, but I think that's the political philosophy that you were asking about. I'm sorry I didn't actually give you a full-blown political philosophy, but maybe I don't have a full-blown political philosophy. Those are the hints in the direction that might help you figure it out.
0:35:53.1 Sean Carroll: Julio Cantillo says, "Has your fame as a podcaster/physicist benefited Villanova?" No. [laughter] To the extent that I have any such fame, I doubt that anyone at Villanova or many people at Villanova even know about it. We now have a pope who went to Villanova. So I am not high up there on the list of people that Villanova benefits from. There's various sports figures. [laughter] The NBA conference finals right now, has the New York Knicks in it, who has half their players are from Villanova's basketball team. So I am not high up there on the list of famous people that Villanova benefits from.
0:36:32.5 Sean Carroll: The new pope, Pope Leo, was an undergraduate at Villanova, a math major as well as something else. He was majoring in something else. He's identified with Chicago 'cause he lived there and was the archbishop there or whatever, but Villanova was his undergraduate education. He's an Augustinian. The Augustinian order is the monastic order that governs and runs Villanova. There's only like two Augustinian universities in the United States. There's many, many Catholic universities that are run by the Jesuits. So the Augustinians have always had like a younger brother kind of syndrome vis-a-vis the Jesuits and education. So Augustinians are super-duper proud that they have given rise to the new pope. So we'll see how that actually turns out. I have no prior knowledge of who this guy is.
0:37:21.7 Sean Carroll: Elias says, "Do you feel like the personality of your guest influences your interviewing? Things like, do you get along on a personal level? Did you know them before, et cetera?" Inevitably, yes, I'm sure it does, but I don't quite think that personality is the dominant axis along which to think about this. I've had really good podcast interviews with people who I have known before, and with people I've never met before and vice versa. There's almost no correlation that I notice between whether I know someone from before and how well the podcast goes. Generally when I'm saying this, I should mention I think generally the podcasts go well. So I don't think I've had any truly, really bad disastrous duds. But I do think that knowing someone is helpful because you have a gauge of how good of a speaker they are, and what they know and things like that.
0:38:11.8 Sean Carroll: It's always kind of a crapshoot when you go in with people who you don't know at all. But it's not about their personality so much as their interviewing style or conversational style. And I know that's related to personality of course, but it's a different thing. Some people just give really short, clipped answers, and that's the worst honestly, even if they're super clear. It's a podcast. You have over an hour to fill. You shouldn't think of it as simply filling, but I guess what I mean is, you have the room to expand and elaborate on things in a podcast. That's one of the nice things about it.
0:38:51.3 Sean Carroll: I think a lot of people, even those that have a little bit of media expertise or experience, are used to be on the radio, and they only have five minutes, at most. Five minutes is long for a radio segment, so they gotta bang out their answers right away. And sometimes they don't adjust to the more leisurely pace of the podcast world, where they can elaborate on things a little bit. Other people, as you know, elaborate at great lengths and it's harder to have an actual conversation because they're giving their potted little mini lectures. It's all good. It's all different ways of being successful. But I do think that some people sort of vibe with the podcast spirit a little bit better. And I don't think it's quite a personality thing, but it's definitely a thing.
0:39:34.0 Sean Carroll: Michael Bright says, "I'm picturing a single photon fired from a laser in the vacuum of space. I understand that you can't see the photon, unless it's pointed directly at you or bounces off of matter, but that's not my question. My question is, is this single photon traveling as a wave that is spreading out in the electromagnetic field in all directions? And if so, does that mean the single photon, if you were to observe it, can be anywhere in a sphere that is expanding in all directions at the speed of light?"
0:40:02.9 Sean Carroll: This is a question that comes off as pretty straightforward, but is actually quite subtle for the following reason, quantum field theory. Quantum field theory is the reason this is a subtle question. You start with the electromagnetic field, you quantize it, and just like starting with the electron or the quark or whatever, what comes out in terms of the structure of the quantum mechanical wave function of the electromagnetic field, for small variations, small fluctuations, small field values near the vacuum state, is a collection of particles. This is something that any reader of Quanta and Fields, Volume 2 of The Big Ideas in the Universe, will be very familiar with. When you quantize a field, the low-lying states look like collections of particles. So you have to take both of those seriously. On the one hand, it's a field that you're quantizing, and in situations that are not very close to the vacuum state, the fieldiness might be super-duper important, like the Higgs field, filling all of space. You need to take it seriously as a field. But if you're just thinking about single photons, then you can really think of them as single photons.
0:41:13.8 Sean Carroll: You can think of them as particles, and you can ask about the quantum mechanics of those particles. You don't need to think about the electromagnetic field. You can think about a quantum particle or a collection of quantum particles. And so, rather than thinking of, of course it is related to the electromagnetic field. Don't get me wrong. We've not erased the electromagnetic field. But when it's just a couple of particles, it's better to think about a set of photons than it is about the electromagnetic field. And so, what you're really asking is, when the photon wave function is emitted by, I don't know, a hydrogen atom undergoing a transition or something like that, does it spread out in a spherical pattern until it's observed? Yes, generally it does. Not always. You can make other cases, like if you make a laser, you can make very highly collimated, pointed photons. But a laser is also in the classical regime in some sense. But at a single photon level, typically, if you solve the Schrodinger equation, it would be emitted in all directions. And therefore, yes, when your telescope detects a photon that was emitted by a quasar billions of light years away, until you detected it, it's wave function was spread out all throughout the universe.
0:42:33.0 Sean Carroll: And I know that's weird, but there's many, many, many, many other photons whose wave functions were spread out all throughout the universe that you didn't detect and they were detected somewhere else. So the story kind of hangs together, but it's just very different than how we actually think about things, because we think about things in ways that are false but good enough for government work.
0:42:54.5 Sean Carroll: Jake Rigby says, "Does the past hypothesis describe a fine-tuning problem? And if so, how is it or could it be related to the classical cosmological fine-tuning problems?" Yeah, I think very much. It is exactly a fine-tuning problem. In fact, I think it's the correct fine-tuning problem. It's the one that matters. I'm not exactly sure what you mean by the classical cosmological fine-tuning problems, but there's three that were identified by Alan Guth in his original paper on inflation. By identified, I don't mean he identified them. He talked about them. They had previously been de-identified. The horizon problem, the flatness problem, and the monopole problem. The monopole problem is a special problem that only exists if you already believe in grand unified theories.
0:43:37.3 Sean Carroll: Let's put that one aside, because maybe grand unified theories aren't right. The horizon problem and the flatness problem were identified by Bob Dickey and Jim Peebles, and I think that Guth, as a postdoc at Cornell, went to a talk by Dickey, where he heard about these problems and that got him thinking about them. So the flatness problem says, the overall geometry of the spatial slices of the universe could be zero or it could be positively curved or it could be negatively curved. If there is any small curvature away from zero at early times, that curvature grows relative to the matter and radiation parts of the universe, and therefore why isn't it huge right now? Who set the flatness of the universe to be so close to zero, or the curvature of the universe so close to zero at early times?
0:44:27.9 Sean Carroll: The horizon problem is a little bit trickier, but it says if I look back at the cosmic microwave background radiation, and I say in the conventional Big Bang story without inflation, I look at a point in one direction, I get it's temperature, it's 2.7 degrees, Kelvin, I look at a point on the opposite side of the sky, also 2.7 degrees Kelvin, but if I look at the past light cone of those two points that I'm observing, those two points have non-overlapping past light cones, that their past light cones only go backward to the Big Bang, and then they hit the singularity and then they stop. So there is no common causal past to those two points, and therefore how did they know to be at the same temperature? There's no influence from that one side of the universe to the other, until we observed it with the same temperature. So how did it know? Of course, it's kind of not surprising, it's not at all surprising that the local temperature at the moment of recombination when you make the microwave background is the same, but you don't observe the local temperature, you observe the local temperature filtered through expansion and redshift and all that stuff. So really what you're asking is, why did the universe start expanding at exactly the same time in two regions that were never in causal contact? That's the horizon problem. I think that both of these problems are sort of ill-posed in the usual way of talking about them.
0:45:53.9 Sean Carroll: The flatness problem is I think just a fake problem. If you look more carefully at the set of all possible universes, almost all of them are spatially flat, so that's not really a problem at all. The horizon problem is definitely a problem, but it's not the problem they say it is. What it really is, is just a smoothness problem. Why is the universe so smooth even if it's not given time to equilibrate? But that's a bad way of putting it because even if it were given time to equilibrate, it wouldn't be smooth either. It would get lumpy because gravity is important. What it's really a problem about, is why is the entropy low? That's the problem. That is actually the fine-tuning problem that the early universe faces. So Roger Penrose and some other people, including myself, have been trying to make this point for a long time, but no one will ever listen to us. I'm used to that.
0:46:41.9 Sean Carroll: Eric Dovigi says, "Do you think that there might be biological factors that determine whether or not a given life form can develop mathematics, just as there are certain biological factors that determine whether or not a life form can develop language? Or do you think the same set of variables would determine both?" Oh man, I don't know. I really have no idea. I suspect that there is not a unique set of biological factors. There obviously are biological factors in the sense that human beings can develop math and bacteria don't. There's a biological difference between us. So in some cheap sense, sure, there are biological factors. But I tend to think, that if you generally become smart enough, you will develop mathematics in some very vague sense. So you might become smart in the way that an octopus is smart or the way that a human is smart. It's not necessary that you will develop mathematics. I don't think that octopuses have advanced mathematics. Maybe they would get there if we gave them another billion years to think about it. But maybe they just don't need it. Maybe their evolutionary pressures do not push them in the direction of developing mathematics.
0:47:50.6 Sean Carroll: So I think that there is some kind of threshold you pass when you become, I don't know, Turing complete or something. I honestly don't know how to say it precisely. But you become a logical thinker and you're able to do a whole bunch of things. I'm sure that there's more to be said about it than this, but these are just my uneducated guesses. I think that it would be very, very interesting. I'm trying to do the opposite of poo-poo your question. I'm literally saying, I don't know the answer. For language, there's just a much closer connection between the abstract capacity and our biological physiological capabilities.
0:48:33.1 Sean Carroll: The shape of our throats and our tongues and things like that, not to mention the capacities of our brains, and maybe even things like the capacity of our hands to grasp objects. All of these could come into developing language. For developing math, I think it's much less clear how that happens. So I'm skeptical that there's anything specifically biological, but I'm very open to the possibility.
0:48:56.3 Sean Carroll: Rob Greiber says, "On a recent popular podcast, Jan 11 talked about the featurelessness of black holes, how they can be fully described by just three factors, their spin, their charge, and their mass. In that sense, she went on, black holes are just like electrons, which can also be fully described by just a few factors. And therefore, you can't say this is my electron, but rather they are all the same. Setting aside the question of location, this got me thinking about the comparison. Are black holes really comparable to, or can be thought of as, objects like electrons? Or due to Hawking radiation and quantum mechanics, are black holes more like a process and less like a particle?" Ooh, that question got very deep there at the end, because the naive answer, sorry, not even naive answer, the simple answer is, yeah, black holes are kind of like particles. I even mentioned just a few minutes ago, that you can imagine including black holes as virtual particles in Feynman diagrams.
0:49:50.3 Sean Carroll: Usually, the effects would be negligibly small of doing that, because they're so massive. The smallest mass black hole is the Planck mass, which is huge compared to particle physics mass scales. But if you have some very delicate thing like information conservation that you're gonna ruin by including these black holes, then they could show up, their effects could show up, even though they're very massive. So I think that particle physicists tend to think that black holes are, yes, some kind of particle-like thing. But now you're saying, are they more like a process because they can decay?
0:50:24.6 Sean Carroll: Well, the longer story is that, yes, there's a very real sense in which that's true, but there's a very real sense for which that's true for any unstable particle. In quantum field theory, unstable particles don't actually have their own part of Hilbert space to live in. They're basically made out of other particles. You should start your quantum field theory with your set of stable particles. And then, if I say that a muon can decay into an electron, a neutrino, and an antineutrino, that's saying that the quantum state of the muon can be built out of the quantum states of the electron, neutrino, and antineutrino. So I don't need a special quantum state just for the muon itself or for any other unstable particle. This is something that, particle physicists know when they take quantum field theory, but something like the muon that lives long enough, can be treated as very particle-like. We're thinking of building a muon collider, for goodness sake. So even though it's sort of not a completely stable particle, it's not sort of one of the basis elements in Hilbert space, it's particle-like for all intents and purposes. A neutron is the same way, but it's lifespan is measured in minutes, not microseconds like for a muon.
0:51:50.3 Sean Carroll: So when you get to particles that have very short lifespans, like a top quark or a Higgs boson, or more commonly in particle accelerators, you get all sorts of briefly lived combinations of quarks. Unstable mesons and baryons and things like that. And these are known as resonances, because they appear in particle physics calculations. They give a little bit of extra lifetime to a certain particular combination of quarks, not ever long enough to be seen in a particle physics detector, but long enough to affect the probability of other things being seen. So in a very real sense, a black hole is like a neutron or a muon or one of these unstable mesons. It is a semi-stable, but not perfectly stable macroscopic configuration that can be treated very much like a particle to the same extent that a muon or a neutron can.
0:52:50.3 Sean Carroll: Matt Haberland says, "You mentioned that you would prefer a simple majority vote to the Electoral College system. What else about the Constitution would you change? That is, for what other aspects would you have preferred an alternative?" Well, I certainly don't have a fully blown out, fully thought out, I don't know why, yeah, maybe the sort of violence of the present political moment in time is making me think of blowing things up rather than thinking them through. I don't have a fully thought out view of what the Constitution should be like. I'm very impressed that the Constitution that was made 250 years ago or a little bit less has done so well in the sense of lasting so long. Clearly, the opportunity to change the Constitution over time has helped quite a bit, but also clearly it's not perfect.
0:53:42.4 Sean Carroll: I would be very, very much against anything like a constitutional convention, as I've said before, for the simple reason that I don't trust the people who might end up being in that constitutional convention. If you really had the right people doing it, you could probably do a much better job than the Constitution we have right now, but I would place close to zero credence that we would have the right people in there. Politics has changed quite a bit since the American Revolution, but I can imagine certain changes I would like to make. I think that a lot of them, for very natural reasons, came out of the historical moment that they were proposed in.
0:54:22.4 Sean Carroll: There were worries about dominance of the bigger states versus the smaller states, and the slave-holding states wanted to protect their ability to keep doing that, 'cause they knew that there were abolitionist sentiments in the North and so forth, and so many compromises were made that we don't need now, and in fact, are actively harmful right now. The Electoral College is one of them. The Electoral College is clearly a failure. Nobody in their right mind, if they're thinking carefully, can think the Electoral College as it is currently set up is a good idea. The whole idea originally of the Electoral College, was that you would vote for some smart people who would then pick the president. That's clearly not what happens. It's nowhere close to the actual current system, which is just that states have a usually majority winner-takes-all allocation of a certain number of electoral votes. No one had that system in mind. The ex post facto justification for it, is that you want small states to have a voice just like big states. That has completely failed, just to be super clear here, because what matters is not whether you're smart or big in the Electoral College, but whether you're close to 50/50 between one party and another, whether you're a purple state.
0:55:37.5 Sean Carroll: Red states don't matter, whether they are big or small. Blue states don't matter, whether they are big or small, and there are examples of all of those. This is a dumb system. There's no reason why Pennsylvania and Ohio should get that much attention while California and Texas get almost no attention, and not to mention Idaho or Kansas. That just is a bad way of doing things. So I would absolutely get rid of the Electoral College. I would get rid of the Senate or at least dramatically change the Senate. The states, what constitutes a state in the United States is more or less entirely arbitrary, and there's just no reason why a more or less empty state should have as many Electoral College votes as a highly populated state. Sorry, the same representation in the Senate as a highly populated state. So that's the same kind of thing.
0:56:29.4 Sean Carroll: Otherwise, I do think that the Constitution does a pretty good job. I think that it could be much more clearly spelled out. The American Constitution, for better or for worse, is really short. [laughter] I remember I was at a summer school. I was teaching at a summer school in Europe in the early 2000s, and it was weird that I was there. I was there with Rob Wald, who at the time I was a professor at the University of Chicago, and so we were talking about relativity and cosmology. But most of the school was dedicated to politics and international relations. And there was a guest lecture by Stephen Breyer, the American Supreme Court justice, and he was chiding the Europeans, and I think for good reason, because they at the time were thinking about the European Constitution. And the European Constitution was this, the proposed one, I don't know what it's current state is, but it was this completely impenetrable massive document where they tried to figure out everything that could ever possibly happen in the world and give you a ruling on it. And Breyer was making the point, which I completely agree with, that's not what a constitution should look like. That's what the code of laws should look like. The constitution should be lean and mean. The constitution should give you just some general principles. Okay, completely on board with that.
0:57:50.1 Sean Carroll: Nevertheless, I think the American constitution is way too vague on all sorts of things about the relationship between the judicial, legislative, and executive branches. And what happens as a result of that, is that the power struggle between these two branches, these three branches, shifts back and forth, and to and fro depending on the particular political moment that we're in. I think a lot more explicitness about what the executive branch is allowed to do, what Congress and the courts are allowed to do, and what they're supposed to do would be very, very helpful. So I wouldn't necessarily be in favor of a dramatic overhaul of the constitution. I think you could tweak it a little bit. The electoral systems definitely could be overhauled.
0:58:35.9 Sean Carroll: I would love to see a very different electoral system, maybe involving something like ranked choice voting or something like that. I guess the big question that I have ideas about, but I haven't really completely thought through so you shouldn't trust those ideas, is this whole experiment that the United States is doing with a separately elected head of state, the president, a good idea, versus a more parliamentary system where you have members of parliament choose the leader of the party and then the leader of the world if that party gets into power, leader of the country if that party gets into power.
0:59:15.1 Sean Carroll: I do think that the fact that we have a single presidential system separately elected, is a huge reason why the United States is and always will be dominated by a two-party system. It's almost impossible to get a third party in there. And I suspect, I believe, that this has a lot to do with political polarization, especially the polarization in our representatives in government. The difference between Republicans and Democrats in Congress, is way more clear and stark, than the difference between Republicans and Democrats in the United States. I think that's a flaw of the system. It's just that I don't quite know how to fix it. I haven't thought about it. I think a real political scientist would know better than me.
1:00:00.2 Sean Carroll: Alexander Kodronsky says, "I think working with audio has given me an intuitive understanding of the Heisenberg uncertainty principle, something that is fundamental to all waves in general. In audio, we can visualize sounds as wave forms, equivalent to the position space wave function, or snapshots of the frequency spectrum on an analyzer, equivalent to the momentum space wave function. An instantaneous sound like the transient of a drum hit occurs as a sudden peak at a specific moment on the waveform, and will have a broad range of frequencies in the analyzer. A steady sine wave will extend through time on the waveform but have a single peak in the analyzer. There's a natural trade-off between the two extremes, and every audio engineer has an intuitive grasp of this. Is this the Heisenberg uncertainty principle in action in sound waves?"
1:00:42.9 Sean Carroll: Almost it is. Yes, it is very, very close to the Heisenberg uncertainty principle, because what you're calling the position representation and the frequency representation, is exactly like you say, just the position representation and the momentum representation in quantum mechanics. So once you know, that that is a true fact about the relationship between position and momentum, then the Heisenberg uncertainty principle is simply that fact that you mentioned about waves, that a wave that is spread out in position, can be isolated and localized in momentum and vice versa, but you can't isolate and make something precise and localized in both at the same time. However, I don't want people to think that Heisenberg was cheap about this. The really important step is to say, that momentum is kind of like the frequency domain. Like, where did that come from? It was sort of implicit maybe in what Louis de Broglie did with matter waves, and indeed it sort of grew out of that, but that's the big conceptual step.
1:01:50.2 Sean Carroll: In classical physics, position and momentum are just two completely independent things. They're not two different aspects of the same underlying wave. That's the real big leap that gets you to the uncertainty principle.
1:02:02.8 Sean Carroll: Miles Jenkins says, "There used to be a place somewhere in the world where you could go and see the official standard meter or kilogram locked in a glass case. Now, of course, all the standard measures are derived from physical constants. But what about the laws and equations of physics currently accepted as authoritative? Is there anything like an official central register that can be referenced? Or is it like the English language, where cat being spelled C-A-T and not K-A-T isn't chiseled in granite in some great hall of truth, it's just overwhelming preponderance of usage? It is much more like the latter. There is no compendium of currently accepted laws of physics because, what would it be? Who is currently doing the accepting? Is the second law of thermodynamics? Does that count? Or is it only particle physics and gravity and fundamental physics that count?" You can find the currently accepted equation of the core theory in various places that I've written, including the most recent book, Quanta and Fields. But you don't need an official central register for that to be kept.
1:03:05.8 Sean Carroll: You can get it on a T-shirt if you want it. But there are other laws. There are laws at higher levels. Is F equals ma, a currently accepted law of physics? We teach it to all of our students, but we also know it doesn't apply in certain regimes. Even Einstein's general theory of relativity isn't quantum mechanical, so there's some breaking down of it. You see where you would get into trouble, basically, like that. Instead, what you have is the actual belief systems of working physicists. That's the closest you have, and that's more or less analogous to the usage of language in the English-speaking world. Unlike, I take it that in French or other languages, there is an academy that is in charge of saying what counts and what doesn't. We don't have that in English or in physics.
1:03:49.0 Sean Carroll: PT Milo says, "In explaining your lack of enthusiasm for idealism, you point to the unexplained coincidence of all this widespread agreement by subjects, right down to the equations. Does this mean you are unmoved by an anthropic-favored explanation of all this consilience? I.e., the most likely world in which complex questions can be asked is also a world where the mereology includes a hefty minimum dose of useful universal consistency accessible across subjects? Couldn't it be, that fundamentally relational worlds don't achieve the requisite consistency and stability to make these sorts of inquiries in the first place, unless they are rooted in a bed of sufficient but emergent universality?" So if I understand what you're saying, I think I'm on mostly the same wavelength as you here. I do think that, well, let's put it this way.
1:04:39.0 Sean Carroll: Maybe it's too strong to say that I think it, but it makes sense to me, to think that some amount of coherence and consistency in the behavior of the world, is a prerequisite for the existence of intelligent life and therefore is anthropically demanded. But remember, to make the anthropic principle be something non-trivial, it's not enough to say without this feature, we wouldn't be here. You also have to say, there are different parts of the world where that feature doesn't hold. If you say that there is a multiverse in some very expansive sense, in some places there is consistency and coherence of the laws of physics, in other places there's not, then I would certainly say that, okay, the explanation for our observed consistency is something like the anthropic principle. But if you don't have a theory, where those other parts of the universe exist, then you just have a brute fact about our universe that you do have this consistency, and that requires some different kind of explanation that is not anthropic. Then it's just lucky that we have that feature of the world to allow us to exist, but it's not truly an anthropic explanation for it.
1:05:48.8 Sean Carroll: Also, I'm not quite sure what this has to do with idealism. Idealism, I'm not an idealism expert, so I'm sure that idealism experts have different views on this, but to me the most straightforward version of idealism, is our minds or consciousness come first and the apparent physical world comes second, and I would think that different people would have different versions of the physical world that they would conjure into existence, and there might be some minimal overlap for people to be able to talk to each other, but there wouldn't be the exquisite, perfect, precise overlap that we get from doing experiments in physics. The results of which most people don't even know about. So how do they come up with them in their idealistic projection of reality from their minds? I don't know. Again, I'm sure the real experts in idealism have ways of dealing with this. I'm not that interested because I don't give a lot of credence to idealism being on the right track.
1:06:45.7 Sean Carroll: Peter Bamber says, "In an answer to an earlier question, you said that the typical velocity of a dark matter particle is about 300 kilometers per second. We don't know what kind of particles they are. We don't know what their mass. We don't know their mass, though we have an expected mass range, and some alternative theories posit they don't even exist. How can we estimate their typical speed?" That's a great question, actually, in part because it's an insightful question, in part because I know the answer to it. It comes down, actually, to the principle of equivalence, if you want to think about it that way. It's true that we don't know the mass of the dark matter particles. There's a huge range of possible masses. It turns out the mass doesn't matter for figuring out the answer to the question, what is the typical velocity? It's only the typical velocity. There's obviously a range of different velocities. Basically, you imagine that it's cold dark matter. It's not warm or hot dark matter because warm or hot dark matter would have had some non-trivial velocity in the early universe, but cold dark matter is essentially starting at rest, and then it just moves in the gravitational field of whatever structure forms in the universe after time. So you have something like a galaxy or a cluster of galaxies, and you can apply the virial theorem.
1:07:58.7 Sean Carroll: The virial theorem says that, roughly speaking, it says that in distribution, particles moving under their mutual gravitational attraction, there is a constant of proportionality relating the average kinetic energy to the average potential energy. So basically, you have approximately equal values of kinetic energy and potential energy for a system that is roughly in equilibrium, and the galaxy counts as roughly in equilibrium. So you might think, okay, there's a given potential energy, but the kinetic energy depends on the mass of the particle, one-half MV squared. The point is that the potential energy, and also depends on the mass of the particle in exactly the same way. They're both proportional to the mass.
1:08:41.1 Sean Carroll: So that fact cancels out. There's no dependence on the mass, and if you know the gravitational potential in which the particle is moving, you will know it's average velocity. This is not actually very surprising when you think about it. The Earth moves around the Sun at a typical velocity, if the Earth's mass were different, it would still have the same velocity moving around the Sun. So that's where we get this number, approximately 300 kilometers per second. That has to do with the Virial velocity of a typical particle in the gravitational potential of the galaxy.
1:09:14.8 Sean Carroll: Friedrich says, "During my physics education, I was surprised by how unwilling physics and math professors were to use modern proven methods to increase the usefulness of lectures and seminars. Very little interactivity, antiquated ways of presenting a complex topic, and no idea of how much we students retained. We have an excellent department of physics education within the same building, but especially, the theoretical physics professors stonewalled against all kinds of constructive criticism and relied on their simplistic model of the brain and how learning works. Does that match your experience? How much potential is left untapped due to antiquated teaching methods?" Well, I have to say, I think there's two things going on here that kind of point in opposite directions. One is, I'm a little skeptical of the methods advocated by education departments, because in my experience, they're often imagining a certain kind of learning experience which may or may not be relevant to what the particular class has in mind. I remember I taught a course when I was still a postdoc at MIT. I taught a course in general relativity out of which my general relativity textbook eventually grew years later, and I think the course is very good.
1:10:25.2 Sean Carroll: The students seem to learn a lot. They seem to like it. Many of them have gone on to become professional physicists, but it was a big lecture course. It was like 50 students in the class. There's not that much more I could do than stand at the board and lecture and give them problem sets to do. And I told this to someone who was in the physics education side of things, and they said, I remember, "That's not even education at all. You weren't teaching them anything." [laughter] And they said, well, it should be a flipped classroom. It should be more interactive, and that's just not appropriate for a large lecture course in general relativity. They need to learn the symmetries of the Riemann tensor more than anything else.
1:11:00.4 Sean Carroll: And, of course, that's not all you need to do. The lecture is never enough. You need interactivity, but the interactivity in a large course like that, takes the form of doing problem sets, which I encourage students to do with their classmates and talk about it. And, of course, there's office hours and recitations and the whole bit. So, on the other hand, I do think that there are ways to teach better, and I do think that sometimes, the education specialists really do put their fingers on ways to teach better. And then, guess what? Physicists just, or not even physicists especially. Professors don't wanna be bothered.
1:11:37.8 Sean Carroll: Professors are not chosen for their dedication to teaching. That's not why they get hired. And this is a bit of an issue because students, of course, want the professors to be dedicated to teaching, but students don't wanna go to the colleges where professors are chosen to be good teachers. They wanna go to the colleges where the professors are famous researchers. That is a tension that you have to learn to live with. If you want to go to the universities where the faculty is chosen for their scholarly expertise, most citations, most influence on the field or whatever it is, then you can't be surprised. You can be upset if you want, but you can't be surprised when they're not really dedicated to teaching. Some will be. Some will absolutely be open-minded in thinking about what are the best ways to teach. I've seen all sorts of innovative attempts tried. Sometimes they work, sometimes they don't. But there's controversy over what the best way to teach is in any particular system.
1:12:35.8 Sean Carroll: So I think it's a combination of those two things. I think that there's overclaiming sometimes on the part of the people who are specialists in physics education, and I think there's underappreciation at times on the part of those who are actually given the task of doing that physics teaching.
1:12:52.1 Sean Carroll: Anonymous asks a priority question. "In the block universe view, where all moments in time are equally real, there still seems to be an additional degree of freedom, the present we experience. Why is now now? What distinguishes this particular slice of space time today from all others?" I've heard questions like this before, and I truly don't get it. I think that there must be some, I don't wanna read anything into the questioner's mindset, but there must be some underlying belief in some essence of personhood, of soul or something like that, that travels through time and experiences different moments at different times. To me, the answer is obvious. The me, who's experiencing the world at noon on Saturday, could only be experiencing the world at noon on Saturday 'cause that's who I am. There was a me that was experiencing the world at 11:00 AM on Saturday, and that person experienced it then.
1:13:46.7 Sean Carroll: It's like saying, why is the number seven seven? Why is the number seven not eight? I just don't quite understand how that could be a question. There's nothing that distinguishes this particular slice of space-time from all the others, other than the people who are in that slice of space-time.
1:14:03.8 Sean Carroll: Sam Hartzog says, "As physical theories progress from emergent to fundamental, the amount of information needed to specify a state grows astronomically. Given that we're progressing from a broad, generalized approximation to a precise solution, this burgeoning information requirement is in some sense trivial, but it nonetheless brings to mind parallels with algorithmic compression. Is there a more than trivial connection here? Have past or present collaborations between data/computer science experts and theoretical physicists yielded anything interesting from the perspective of the physicists? Or has this type of collaboration traditionally yielded one-way insights, namely data scientists taking inspiration from physical theories to develop more efficient compression methods?" I think it's probably more of the latter, but I'm not sufficiently embedded in those communities to give you a very accurate point of view. You're absolutely right, that one of the features of emergent phenomena is, you're able to make predictions or at least make statements relevant to the theoretical context in which you're working on the basis of much less information than you would at a more microscopic level.
1:15:05.8 Sean Carroll: And I do think that there's been a lot of work on figuring out when that is possible, the kind of different ways in which it can be possible, et cetera. The computer scientists, the data scientists, are much more down to earth. They're like, give me a big data set. I will find some sub-manifold in which all the action is taking place. The simplest example is principal component analysis, where you have a vector space and you have a distribution of points. And you can say, well, most of these points are embedded or aligned along some direction. They're not uniformly scattered through the space. So we can simply parameterize them by where they are within that subspace, not in the whole space. But you want something much more rich than that, thinking like a physicist. You want not only to know what variables matter, but then how those variables enter into dynamical equations or other sort of physical relations. My impression is, that that is an underdeveloped area right now. I've been talking to people here at Hopkins who are interested in it. So we're thinking about it, but I don't think that there's any body of work that you point to. And here's the textbook that tells you how to do that. I don't think we're at that point quite right now.
1:16:16.9 Sean Carroll: Ilya Lvov says, "In your previous AMA, you discussed Norton's Dome and asserted that we cannot get such a non-deterministic solution in quantum mechanics, due to the linearity of the Schrodinger equation. Does this mean that there cannot in principle be a quantum system whose classical limit is precisely Norton's Dome? If true, might this be interesting in it's own right?" Well, for those of you who don't remember, Norton's Dome is a specific example of a classical mechanical system highlighted by philosopher John Norton, where it's a dome with a very specific shape of the rate at which the dome curves with the property that if you imagine a particle, a perfectly frictionless spherical ball obviously, sitting at exactly the top of the dome, then there's a solution to the equation that says the particle sits there forever. There's also a solution to the equation that says, at some arbitrary time the particle starts slowly moving in some direction, and that direction is arbitrary and that time is arbitrary, so the system is non-deterministic in the usual sense of doing things.
1:17:21.3 Sean Carroll: I pointed out that this is not going to be. There's no equivalent to this in quantum mechanics because quantum mechanics is linear. All of the solutions are completely deterministic according to the Schrodinger equation. So the question is, does that mean there cannot in principle be a quantum mechanical system whose classical limit is precisely the Norton's Dome? That's not exactly what I would say. I would say there cannot be a quantum mechanical system that stays in the classical limit and acts like the particle on Norton's Dome. So if you put some wave function on the top of the dome, the wave function already has some spread. If it had zero spread, then it would start spreading right away. So more typically, it would have some spread to begin with, and then the whole wave function would start spreading down the dome. But what that would mean is, that the wave function would no longer be localized at the top after some period of time. It wouldn't be localized at any particular direction around the dome. It would be spread all over the place. So what happens is, the quantum system stops behaving classically. But to some extent, that's not dramatically different than any other quantum system.
1:18:25.2 Sean Carroll: There's nothing really special about Norton's Dome here. If you just have a wave function moving in a potential, they very often spread out all over the place. So I don't think it's teaching us anything dramatically new about quantum mechanics. I'm gonna group two questions together.
1:18:41.3 Sean Carroll: One is by Scott Collins, who says, "What do you think is behind the anthropomorphic language around artificial intelligence? Journal articles use terms like deception, faking, honest, helpful. Even terms like goals, alignment, rewards, and motivation are freighted with a lot of misleading human meaning. Is it just how humans communicate? Or maybe AI companies see conjuring the threat of AGI and conscious AI as lucrative? And then Fred Brunner says, "What kind of criteria or concrete tests would convince you, that an AI system has reached AGI? Could you share an example or two of what such a test might look like?
1:19:16.7 Sean Carroll: So maybe you can see the connection between these two, or maybe not. For the first one, why I think that the anthropomorphic language is so popular is something I would be loath to speculate on. At least, I'm happy to speculate on. I shouldn't say it that way. I would be loath to decide what the right answer was, 'cause I don't think I have enough data. But you can imagine various different reasons why it would be possible. Certainly, the fact that these models are being created by companies that eventually want to make money, and think that one of the ways to make money is to cast their AI models as human-esque in various ways, means that they're developing models to have the ability to sound human. That's not surprising at all. But on the other side, that's sort of the supply side version. The demand side version is, people kind of naturally put things they experience into the boxes that they're used to using to organize their experiences.
1:20:21.2 Sean Carroll: So when we see human beings acting in certain ways, and we assign terms like deception or motivation or whatever to those actions, then when we see other systems acting similarly using the same words, et cetera, it's very, very difficult to resist using exactly the same existing words. It might be more philosophically rigorous and correct to invent a new vocabulary to describe the same words coming from a large language model as when they come from a human being, but very few people are gonna do the intellectual effort to do that.
1:20:54.3 Sean Carroll: So I don't think it's actually conjuring the threat of AGI, even if the companies just wanna sell you, I don't know, a toaster that is intelligent enough to stop toasting when the toast is done, it still might help them to make your toaster sound human. I really don't know about that. To Fred's question about the criteria or concrete tests for reaching AGI, I don't think that there is such a thing as AGI. I think that that's the wrong way to think about it. Again, there's this idea, since we're human beings, that there is a level of intelligence that human beings have, and computers will reach it or won't reach it. But that's just the wrong way of thinking. It's not that computers don't have the capacity or ability that human beings can, it's that those capacities and abilities are very, very multifaceted.
1:21:43.2 Sean Carroll: Computers might become much, much better than us at some things, like playing chess, while much, much less good at us than other things. Like right now, it's still difficult to subtract two three-digit numbers with a decimal point and get the right answer in certain language models. So I just think that looking for AGI is the wrong kind of thing. You should appreciate, and I just say this over and over again, so I don't know what good it does, but you should appreciate the large language models or any other form of computer program for what they are good at, rather than trying to test them against human beings.
1:22:18.8 Sean Carroll: Which is not to say that they won't someday get better than human beings at everything, that's fine, but still recognize and understand what they are good at, and how good they are at it. And that will help a little bit prevent you from anthropomorphizing these programs overly.
1:22:37.2 Sean Carroll: Brian Ravitch says, "I've been learning about Feynman's path integral, and while I don't fully understand it, one of the concepts sparked another question in my mind. If a photon travels all possible paths, but most destructively interfere, leaving just the direct route we see, couldn't this concept apply to other particles? Specifically, is it possible that phase cancellation of particles could be an alternative to the many worlds interpretation? Instead of branches of reality, could it be that the other potential realities phase cancel out instead of manifesting in another universe?"
1:23:07.8 Sean Carroll: It might be possible, but it is not true. That's not actually what happens. The language used here is very similar, so I know why you naturally draw a connection between them. The many paths or many histories or many trajectories that a quantum system takes, certainly sounds a lot like the many worlds of quantum mechanics. But in fact, they mean very, very different things. Remember earlier, when we talked about the difference between unitary evolution, the evolution that a quantum system does when you are not measuring it, versus the outcome of measurements. The Feynman path integral is all about unitary evolution. It's just a way of rewriting the Schrodinger equation. It's just saying, if I have a quantum state and I let it evolve according to the unmeasured evolution equations, then I will get a certain answer at the end of the day. It's colorful to think of it in terms of a sum over all the paths, and there's nothing wrong with that, but those paths have nothing to do with separate worlds or anything like that. Why? Because they don't decohere from each other. In the many worlds interpretation, decoherence, that is to say becoming entangled with the environment around you, is the thing that we, in common language, refer to as a measurement.
1:24:22.2 Sean Carroll: So the splitting of the worlds happens under a measurement or under when the system becomes decoherent. But the Feynman path integral is typically used precisely when you're not becoming decoherent, precisely when you're just keeping all of the information in the quantum system and following it's evolution as an isolated system. So the words sound the same, but they're put to use in very different contexts.
1:24:47.7 Sean Carroll: Eli Reams says, "In your Does Time Exist podcast episode, you mentioned that if the observable universe is finite, then according to Poincare's recurrence theorem, after a very long time the universe will return to the same state as it was before. But I'm curious that, if the galaxies continue to accelerate apart through the expansion of the universe due to dark energy, how could this recurrence happen?" Well, this is a very subtle question actually, so I'm gonna give you the answer, but it may or may not be a satisfying answer. It has to do with the dark energy. It has to do with the vacuum energy. If there were simply a universe expanding forever but no dark energy, then you'd be right. Then the universe would just cool off, it would settle down, and there's no reason to ever think that there'd be any Poincare recurrence.
1:25:32.0 Sean Carroll: And classically, if there were no such thing as quantum mechanics, the same thing would be true even with dark energy. Even with the cosmological constant, there's the cosmic no-hair theorem, you would empty out and become cool and cold and dark and desolate, and that would be it forever. But you have two things. You have dark energy, which let us imagine is vacuum energy, is the cosmological constant. And then you have quantum mechanics. The reason why the cosmological constant is important, is because even empty space has an un-trivial energy density with the cosmological constant. And this gives rise to curvature of space-time, which gives rise to a horizon. So there's something called the de Sitter horizon, because de Sitter space is that ultimate answer when you have nothing in the universe other than the vacuum energy. We are approaching an emptier and emptier universe. It will take many billions of years to get there. Don't worry. But at some point, you have a horizon around you, and otherwise empty space. But Stephen Hawking taught us back in the 1970s, in collaboration with Gary Gibbons, he said, "Just like black holes have horizons and radiate, de Sitter horizons also radiate. They have a temperature." And you interpret that temperature as being part of an indication that there's an entropy.
1:26:48.4 Sean Carroll: That entropy, in turn, indicates that there are a finite number of degrees of freedom inside the horizon of the universe. Now, are there extra degrees of freedom outside the horizon? That we don't know. Classically, the answer is obviously yes. In quantum gravity, we don't know enough to say right now. So plausibly, the universe only has a finite number of degrees of freedom quantum mechanically, even though classically it just expands forever and cools off. And that would mean, that quantum mechanically there is a recurrence theorem that says it could come back. Now, if you detected some fuzziness, some lack of complete confidence there, yes, there is a lack of complete confidence there. These are things that we don't fully understand, but we would like to understand what the implications are of all the different possibilities, until we actually do understand what quantum gravity really says.
1:27:41.6 Sean Carroll: Sean Miller says, "In your paper on emergence, you incorporate historical influences into the emergence relation by assuming Markovian evolution where they are encoded within the present state. While this approach offers structural clarity, I wonder if it sets aside a crucial aspect of how emergence unfolds, particularly involving systems like those found in biology. Specifically, in life, key emergent events are not just characterizable by the static properties of a map between established levels, but by biological systems developing and stabilizing new modes of compressing information over the course of evolution. So my question is, do you view the evolution of these novel compressive regimes, and the corresponding redefinition of informational relevance they entail, as a separate issue concerning the origins or dynamics of emergence distinct from your classification of the resulting structural relationships? Or could this perspective potentially enhance your classification scheme, perhaps by providing a mechanism for understanding the stability or functional significance of certain emergence types like Type 1B or Type 2?
1:28:43.3 Sean Carroll: Honestly, I just don't think that this, what you say is completely true. Biological systems through evolution certainly do develop new ways of informationally compressing their knowledge of the world around them, and their own activities. I don't think that's a challenge to anything that we say in the classification system. Even physical systems undergo phase transitions and they themselves can be subject to different methods of compressing their relevant information. So that's fine. That's just saying that your theory of a higher level emergent thing changes over time.
1:29:16.3 Sean Carroll: The actual entities that you have, that might be relevant to describing the world at any one time might be different after such a phase transition than they were before such a phase transition. So I think that that's fine. I think it fits in perfectly well. It might not be the most convenient way to think about things. It might be that you gain insight by thinking about the past history of something. But given that the laws of physics are actually local in time as well as in space and everything else is based on the laws of physics, if you have some system in the current universe, and you wanna say, what is the relevant information about that system to predict it's future behavior, that relevant information has to be located in the system at the present moment. It can be a memory or a sort of a fossil or an implication of things that have happened in the past. But if there's something that happened in the past that leaves no memory, no trace on the present instance, then it won't help you understand the dynamics or the behavior of the present system. So that's why I don't think you need to go beyond the Markovian paradigm, even though it might be a useful way to think informally.
1:30:31.1 Sean Carroll: Jamie says, "You often say about anthropic reasoning that we need to think more carefully about the possibility of being a typical observer. And that rather than modestly assuming we are probably typical, we might be immodest to assume the typical observers are like us. Can you please do a slightly deeper dive than usual, and show us how to approach thinking more carefully about this issue of typicality, when we can use it and when and how it might mislead us? Love to hear connections to Boltzmann brains too.
1:30:56.8 Sean Carroll: Well, I'm sad to say I don't completely know. I don't really have a fully-blown... I'm working at it. I would like to have a fully-blown theory of how to correctly reason anthropically. My tentative place is that, you should only assume you're typical within the set of observers who look exactly like you. Now, the word look is doing work there. I don't mean that they are exactly like you. They might be located elsewhere in the universe. They might have different microstates inside them, but their macroscopic setup is exactly the same.
1:31:26.6 Sean Carroll: This is completely true, for example, in many worlds. When you have two different branches of the wave function, the two different versions of you on the different branches are, at least at the start, going to be completely macroscopically indistinguishable. It's also true for Boltzmann brains. I don't need to compare myself to a disembodied brain floating in the empty, desolate cosmos. I can compare myself to other copies of things that have fluctuated into existence that look like they're in the room I'm in right now. Those will fluctuate into existence also. So among those sets of macroscopically identical observers, I do think it's perfectly okay to think of yourself as typical. But I don't think it's okay to pretend to forget who you are in the universe. I think that you should only be able to help yourself to empirically verifiable data. I don't think that you're allowed to say, I'm gonna assume I'm this observer and not that one, even though from their own points of view, they're perfectly identical. I think that that sort of God's eye view is cheating a little bit in cosmology.
1:32:31.0 Sean Carroll: I'm going to group the next two questions together. Edward Crump says, "Loosely stated, Alex Vilenkin says, that particles from vacuum energy overcame energy barriers to initiate the Big Bang. This would suggest something present before the Big Bang. What is your opinion? And Thomas Scabbard says, "If our universe arose from nothing or quantum fluctuations in the quantum foam in the bulk, then does this not mean the time existed before the Big Bang? Even virtual particles exist for a brief time. Therefore, there is a passage of time without there being a universe. Am I wrong in this assumption?" So look, these are questions about which we have no complete answers. We have no definitive agreement between physicists on what's going on. We're talking about the quantum mechanics of the universe that involves gravity, but also we're talking about deep conceptual questions about the emergence of time.
1:33:23.8 Sean Carroll: If you're Alex Vilenkin, who is a very, very respected cosmologist, one of my favorite scientists out there who's done amazing work, he has a theory of quantum cosmology going back to the '80s, that was sort of one of the rivals of, there was a proposal of course, from Stephen Hawking and James Hartle, the Hartle-Hawking no boundary wave function of the universe. There was a separate proposal by Andrei Linde and a proposal by Alex Vilenkin. I think Vilenkin and Linde were more or less similar in their outcomes. But all of these are working within what you might call traditional quantum cosmology.
1:34:01.1 Sean Carroll: And traditional quantum cosmology says, you take gravity, you consider the gravitational theory of a closed universe, so spatially closed, like a sphere or something like that, simplifies some of the math. You do quantum mechanics to it and you end up with the Wheeler-DeWitt equation. And the Wheeler-DeWitt equation says, the universe, the quantum state of the universe, the wave function of the universe, does not evolve with time. Okay? The argument between Vilenkin and Linde on one side and Hartle and Hawking on the other, was about what the wave function is, what it's mathematical expression looks like. But the fact that it doesn't evolve with time was common to them. And that's a problem if you think that the universe evolves with time, which most of us do. So the answer is of course, the time is supposed to be emergent and having that actually work out is tricky, as I've explained otherwise, as I've explained recently in the Emergence of Time episode.
1:34:56.1 Sean Carroll: So I can't tell you the truth, because the truth is not agreed upon or maybe even known about this. All I can tell you is what I think. What I think, is that all these folks are talking about are wrong. I think that when Alex Vilenkin says that there was something, some quantum state of the universe that then tunneled, he calls it the tunneling wave function, that then tunneled into the universe that we see, I don't think that's a very sensible picture at the end of the day. When did it tunnel? When did it decide to tunnel? I think you're right, and Edward is right to wonder, how could this be possible? That needs to be something before the Big Bang, even if it's just a quantum state. And I guess Thomas is saying the same thing.
1:35:39.2 Sean Carroll: So maybe there is some correct understanding here where these words, like there was some state of the universe that was not directly recognizable in terms of conventional space-time, et cetera, did preexist the Big Bang and then there was a moment at which it underwent a phase transition into the expanding hot, dense universe that we know about or then to an inflationary universe, perhaps. Maybe those words work, but they would only work as an approximation. I don't think that they're the deep thing going on. So I think that we very much need a better understanding of what it means to say the time emerges before we can give the real answer to these things.
1:36:24.7 Sean Carroll: Graham Snowden says, "On a 19th of May BBC NewsHour program, hosted by Martin Griffiths and in firm opposition to the UK government's recently passed legislation to exempt tech firms from copyright laws, Sir Elton John made the comment that he felt betrayed that government MPs had rejected proposals to force AI tech companies to disclose what materials they were using to develop their programs. He stated that a machine is incapable of writing anything with a soul in it, and that this development is George Orwell times a thousand. What is your position on these two statements? Well, I don't love any of those two, either one of those two statements really, although I do kind of sympathize with the conclusion that Sir Elton reaches here, which is that the AI programs are absolutely profiting off of the creative efforts of other people who are not getting compensated for their efforts.
1:37:16.3 Sean Carroll: So, I don't think it's true that a machine is incapable of writing anything with a soul in it, because I think that we are machines basically, and I think there's not that much of a conceptual difference between a human being and a machine. The kinds of machines that we have right now are very different than the kinds of machines that human beings are, but I don't think that it's fair to say that a machine is incapable of this or that. Machines are capable of lots of things, and we should be open-minded about that. And I'm not sure why he says this development is George Orwell times a thousand, other than George Orwell said a bunch of warnings that we need to pay attention to, and maybe this is one of them. I don't know exactly what he had in mind there. I do think that, it's completely obvious that AI companies make money or plan to make even more money off of the work of other people without paying them back. And basically, their argument for it is, but paying them back would be really expensive. [laughter] Well, yeah okay, maybe, that's very possible.
1:38:18.4 Sean Carroll: But when I see these video generation things, and just for the last couple of days, there's been this discourse on social media about replacing screenwriters and directors and directors of photography and things like that, because it can all be AI. But clearly, you framed these AIs on works done by actual human directors and screenwriters and actors and so forth. So it's just hiding all of those contributions under layers of technological mixing and matching, but it's still absolutely there. And I think that that's perfectly legit for those creators, those artistic creators especially, whose creative works are being used to train the AI to ask for some slice of the pie, such as it is. Now, you can be subtle about this. I don't think you need to be like over-emotional about it. There are other things that are free out there in the public domain, that I think is perfectly okay if AI trains on, and that's okay. But yeah, I do think that Sir Elton has a point overall.
1:39:30.2 Sean Carroll: Paul Kent says, "Which sci-fi novels shaped or ignited your early or later scientific interests and thinking?" Honestly, my scientific interests early on when I was a kid were ignited by popular science books, not by science fiction books. I did read a whole bunch of science fiction, and maybe it had some effect. I don't keep track of all the effects that my early reading had, and I tend to doubt other people who claim to have had epiphanies from reading something, and it's probably more of a story that seems to be dramatic that they like telling later on, so I don't wanna go down that road. I've read a lot of stuff, so who knows what inspired me to do one thing or another? But I certainly read a lot of nonfiction books about physics and cosmology, and that's really what got me excited. The science fiction I like, there's a bunch of sort of classic science fiction authors.
1:40:20.9 Sean Carroll: Robert Heinlein was always my favorite, but I read a lot of Anne McCaffrey, Ursula Le Guin, Roger Zelazny, a bunch of different people. And most of them were not trying to do anything scientifically dramatically different. Heinlein sort of tried to do usually things that stuck within scientific plausibility, but he certainly had fun writing stories about time travel or whatever that were brain twisters and were fun to read. The one exception, I guess, which I've certainly mentioned before, is by Robert Forward, who was an engineer of some sort. He had a scientific background, but he wrote a couple of novels, and my favorite one was Dragon's Egg, which was about a life on a neutron star.
1:41:12.0 Sean Carroll: It's about human Spacefarers find this neutron star, and it was a long time ago, but don't ask me to remember all the details, but there's little primitive life forms on it, but because there's also gravitational time dilation, and everything is made of very, very tiny things, so the neutron star life, they're not made of neutrons. The crust of a neutron star is not made of neutrons. In the center, it's neutrons, but on the edge, there's other things going on. I don't know whether life on a neutron star is plausible, but it would be very different in very obvious ways if you knew the basic physics of it, and Robert Forward did, and that's what he wrote a story about. So the primitive life forms developed super-duper quickly, compared to ordinary human beings. So I thought that that was thought-provoking in a way that was directly related to science. I really liked that.
1:42:04.3 Sean Carroll: Taylor Gray says, "At what point, if ever, have you considered yourself an expert in physics? I am quite experienced in my field, but I'm not sure I'd ever call myself an expert. There always seems to be more to learn, room to grow, et cetera. I'm curious on your thoughts." Yeah, I think, I don't know if there's a sensible answer to this question. I don't know if it matters whether you can call yourself an expert. What does it mean to call yourself an expert? I think that you're an expert if you can help other people learn more. So there's various degrees to being an expert. A high school algebra teacher is an expert in high school-level algebra. They might not be an expert in what a professional PhD mathematician would call algebra. I do remember once when I was either late in grad school or early as a postdoc, and still, despite the fact that I should have, by that point in time, I wasn't yet sure how to think about my own scientific field of expertise. 'Cause I had written papers on general relativity, on modified electromagnetism, on topological defects, some things on magnetic fields in the early universe, like a bunch of things that in the general bucket of field theory, cosmology, and gravitation. But that's too wide a bucket to have as your specialty.
1:43:19.1 Sean Carroll: And I was at the summer school, and the students of course, sit around at night after dinner just like chit-chatting, and someone asked a question about Big Bang nucleosynthesis or something like that, and I answered it. And later, one of my friends at the school came up and said, "That was really impressive, your speech on Big Bang nucleosynthesis. That was like, very clear and helpful, and you clearly know your stuff." And I thought to myself, oh, I guess maybe I'm a theoretical cosmologist. Maybe that's what I am. I backtracked on that later, 'cause there's other things to do, but I think that sometimes you find that you're an expert just because someone else says, "Wow, you seem to have demonstrated some expertise. Thanks for doing that."
1:44:01.2 Sean Carroll: Chris A says, "Why don't cloud chamber tracks violate the uncertainty principle? So the question here is because when you see a particle go through a cloud chamber, it leaves a track. And it leaves a track that looks pretty narrow, and is also bending 'cause you put the magnetic field in the cloud chamber. And so, if you know what kind of particle it is, and therefore you know the mass of the particle, by knowing how much it bends, you can figure out more or less the velocity and the momentum of the particle. So it seems like, you are measuring both the position and momentum of the particle." Now, I'll confess, I had an answer to this question when it came up, but I did have the thought flick across my brain. This is the kind of question that maybe you could ask a large language model. It's like exactly as well posed enough to be maybe they'll give you the right answer, but also slightly vague enough that maybe they'll stumble. So I asked, I don't know, one of them, I forget. I think it's the OpenAI one, the ChatGPT. I think I have access to the... I believe I have access to the most advanced model. Thanks to a very nice Mindscape listener who granted me access. So I can use that. And indeed, interestingly, ChatGPT-03, I guess, or whatever it was, kind of got it halfway right. It didn't get it completely right. For one thing, it gave like four different answers. And I told it, I'm like, why do you need four different answers if there's one answer that is right?
1:45:28.9 Sean Carroll: And of course, you'll notice if you ever use these things, it was like, oh yes, you're right, I should have been more specific. Anyway, I think the real answer basically, is that it's a crude measurement. The uncertainty principle isn't pinning you down to being completely unable to measure positions and momenta. It's just that there's some uncertainty in it, and the track in a cloud chamber is at least as uncertain as the width of the little track, which is not trivial on subatomic scales. It's a big macroscopic thing. So I think that as a matter of practice, you're not pinning it down nearly enough to actually be in danger of violating the uncertainty principle. The interesting point that ChatGPT made was that, you're also not at any one moment, measuring both position and momentum at the same time. You're measuring position over and over again at different times, and you can keep track of the time, by keeping track of the curvature of the path, et cetera. So you can infer what the momentum was, but you're not actually strictly speaking measuring it. That one, I don't know whether that one is right or not.
1:46:39.5 Sean Carroll: I think the words all sound good, but I worry that it's weaseling out. I'm not sure that that's the right answer. I think I like my answer better. And then I gave a couple of other answers dealing with decoherence and things that are completely irrelevant. So don't trust these things. Use them, but don't trust them. Anyway, the uncertainty principle is not gonna get in your way of doing macroscopic uncertainty level measurements.
1:47:02.0 Sean Carroll: Kent Linkletter says, "I understand that the entropy of a black hole is based on the area of it's event horizon, and this would be directly related to it's mass. Does it matter if the things that fall into it have high or low entropy?" No, it does not matter. That's one of the joyous things about black hole entropy. For one thing, even if it mattered, it wouldn't matter in the following sense. The entropy of the black hole is way bigger than the entropy of whatever you're throwing into it. So if you have some Avogadro's number of particles and you're throwing it into a solar mass black hole, that Avogadro's number collection of particles might be in a zero entropy state or a maximum entropy state. It's not gonna affect the black hole entropy very much at all.
1:47:48.1 Sean Carroll: All that matters, as we talked about before, for the black hole is mass, charge, and spin. So from that gravitational point of view, the entropy of the thing you're throwing into it, is much, much less than what it's entropy would be if you turned it into a black hole. And so, by turning it into a black hole, all you're doing is equilibrating it. You're raising it to it's actually most highest entropy state. It doesn't matter when it fell in, it started with low entropy or high entropy. It will eventually go up to that equilibrium value.
1:48:20.2 Sean Carroll: Tim Gianitsos says, "Einstein was apparently influenced by Mach's principle, which said that a spinning object in space would feel a centrifugal effect only because it spins against a background of other matter in the universe. If there were no other matter, it would be impossible to spin. However, Einstein seems to revert his view on this during a 1920 Leiden lecture, where he suggests that it is possible to spin against a background of space with no other matter. He even goes as far as to resurrect the term ether, though it has different properties to the ether that relativity originally disproved. What is the modern view on Mach's principle? Is this informed at all by the CMB rest frame?"
1:48:55.8 Sean Carroll: Well, the modern view of Mach's principle is that, it may have had some historical use to Albert Einstein in inventing general relativity, but it's not actually part of general relativity. In fact, one of the things I was proud of is when I wrote my textbook on general relativity that was published, I guess, in 2005, it might have been the first complete textbook on general relativity to not even mention Mach's principle. [laughter] Because I think that Mach's principle confuses people. And the reason why it confuses people is 'cause it was certainly part of Einstein's thinking, when he was inventing general relativity, but it truly doesn't show up in general relativity.
1:49:33.3 Sean Carroll: But on the third hand, there are certain special cases where something like Mach's principle is true. It absolutely is possible to measure the spin of something with no other matter around. That's just true in general relativity. Newton's bucket experiment shows you that. It has nothing to do with the existence of matter far away, but there are specific cases like the CMB where nevertheless there is matter in the universe, and you can measure some things like your motion relative to that. The microwave background gives you a rest frame. Is that Mach's principle?
1:50:08.7 Sean Carroll: No, not really, but it's kind of Mach adjacent, and so sometimes people talk about it that way. I don't advocate talking about it that way. I advocate just forgetting about Mach's principle if you're a physicist. Take it very seriously if you are a historian of physics.
1:50:24.4 Sean Carroll: Michael Long asks, "Assuming humans have a measure of agency over some of the energy that flows through them, what is your hope for the energy you direct into the microphone?" I suppose this is a way of asking, what are my aspirations or reasons for podcasting or otherwise talking to microphones at various times? Yeah, that's a very good question. I'm sure that I've talked about it before. I have lots of different levels of aspiration for talking into the microphone. Mostly, if I'm very honest, it's 'cause it's fun. I enjoy it. Having a deadline every week can be a little oppressive sometimes. There's not always a lot of time to do it, but I get enjoyment out of talking to the different people that I have on the podcast.
1:51:10.9 Sean Carroll: That's the major motivation for me doing this. The fact that other people enjoy it, is something that actually gives me the motivation to take it to the accomplished extreme. Everyone would have a good time talking to different people, but I get to actually force myself to do it by having the podcast and coax people into being on the podcast by having an audience for it. So I think that that works pretty well. I do hope that people get something out of it, not just my enjoyment out of listening to people and learning things, but listening to me and to them in conversation or me here just in the AMAs. I hope that people learn something. And this is all sort of obvious, so I'm not dwelling on it. I also think that it's...
1:52:02.6 Sean Carroll: The higher level aspiration is to model a certain way of thinking and being. This is gonna sound grandiose, but recent discourse in the media and social media has been about whether or not the left side of the political spectrum in the United States needs it's own Joe Rogan, because Joe Rogan has been very successful in gathering a huge audience and having a lot of influence, especially young people, especially young men. And he endorsed Donald Trump, and a lot of them are conservative-coded in various ways. And I think it's a silly discourse very much because the actual Joe Rogan is not the conservative Joe Rogan. He's not conservative by any ordinary sense. His political views are a weird mix of things.
1:52:51.1 Sean Carroll: The last person he endorsed before Donald Trump was Bernie Sanders. So I don't know. These people, these consultants, political experts, just insist on seeing things through a one-dimensional lens of liberal versus conservative, left versus right, and the real world is much more complex and multifaceted than that. And Joe Rogan is a certain way of living with pluses and minuses, a certain way of approaching the world and dealing with it and struggling with it and asking questions about it. And there are other ways to do it. And this wasn't my goal when I started the podcast. Sorry, let me say it this way. The goal was not and never has been specifically in contrast with either Joe Rogan or anybody else. But I do think that as a scientist, as an academic, as someone who is open-minded I think, and interested in many different things, I want to let people know by example, that this is a way of living, that loving science is a way of living, but also loving other areas of knowledge is a way of living even if science is your main thing, talking to different people respectfully while not talking to any old person.
1:54:09.9 Sean Carroll: The fact that I try to have people on my podcast who I really think have something to offer, not that I just want to debate with them, is an important aspect of it. Respectful disagreements and things like that, but also respectful agreements. I've mostly had people on the podcast who I wanna learn something from. And I think that, this goes back. I've had experiences this way, like when I debated William Lane Craig about religion. The location of the debate was the Southern Baptist Theological Seminary or whatever it was called. It was not an audience that was predisposed to be on my side. But just by being up there and not being scary, it has a much bigger impact than any of the actual intellectual arguments that you give.
1:54:56.7 Sean Carroll: A lot of the people in the audience, have grown up thinking that atheists are just bad people. And just going up there, and being sort of friendly and nice and holding your ground in terms of making an argument about the substance is important, but also you don't need to be a jerk about it. You can just sort of give your point of view and let other people think about it. And I think that maybe there is something to be said for being that kind of person. I don't always measure up to it, heaven knows, but I aspire to do it. And, that's a good thing to sort of try to be in front of a large public audience where people can learn that that's a way to be. And also, of course, I'm not, like I said, I'm not always good at it, so I can learn to be better at it. I can learn to be a better podcaster, a better questioner, a better thinker, and all of those things. And letting people see that in motion, maybe does some good for the world. As I've said many, many times, I don't think that every scientist should have a public profile. I think there's plenty of scientists who are fine just doing their research.
1:56:00.9 Sean Carroll: I think that the field, needs to have a public profile. So some scientists need to be out there talking to the public. And also there need to be people whose profession is science communication, science journalism or whatever it is. That's a slightly different thing, than being a scientist who is talking to the public. Like as a scientist, when I talk to the public, I get to talk about what I think is interesting. But as a professional journalist, you have to be a little bit more objective and step back and think about what is the most interesting thing in the world, what is the most important thing, and translate it into a way that people can understand. So I want a very rich ecosystem of different scientists, some not talking to the public, some talking to the public, some talking to journalists who then talk to the public. All of that is important. So I play my tiny little role in that giant ecosystem.
1:56:52.9 Sean Carroll: Albin says, "Two new theories have been in the spotlight recently, the dark photon theory and one named gravity from four one-dimensional unitary gauge symmetries and the standard model. Do you think there's any real meat in those ideas?" I have no idea what the second idea is, gravity from four one-dimensional unitary gauge symmetries. That sounds wrong, but maybe there's something missing in the description that doesn't come clear to my brain. So who knows? Maybe. The first one, the dark photon theory, is one I'm very fond of. I wrote one of the early papers, wasn't a co-author, on one of the early papers on dark photons in cosmology. The idea of dark photons, is that you have ordinary photons, which come in quantum field theory from a certain gauge symmetry and very well understood, giving rise to electromagnetism. But what if you have another gauge symmetry that does not directly interact with ordinary matter but only interacts with dark matter? So you'd have a whole separate electromagnetism, dark electromagnetism, and separate particles associated with those, dark photons, et cetera. This is an old idea. It goes back to at least the 1980s. Nothing new there, but there's various problems with it right away. The biggest problem is that the dark photon and the ordinary photon tend to mix with each other. So you get right in immediately in trouble, if you don't sort of turn off that mixing.
1:58:16.3 Sean Carroll: So usually what people do, is to give a mass to the dark photon so that it's allowed to mix with the ordinary photon, but you wouldn't notice if it's too heavy. That's something you can absolutely do, and I think that it rubs me a little bit the wrong way, that nomenclature, because the photon's masslessness is really important to what the photon is. [chuckle] I think once you give it a mass, you should call it something like a dark Z boson or something like that. Rather than the dark photon. The paper I wrote with Marc Kamionkowski and other people was on real dark photons, dark radiation, massless dark photons, and we just turned off the mixing by hand. And the interesting thing was, that the idea that dark matter interacts with a whole new kind of electromagnetism is much less constrained than you might think.
1:59:02.7 Sean Carroll: You might think that you would have noticed by now, there would be dark atoms and dark chemistry and all that stuff. It is easy to get rid of that, but it's also easy to allow it to exist. And that's again, full employment for particle physicists. You get to imagine models. What can be accommodated in current knowledge of the universe and what is already ruled out, stuff like that. It's a lot of fun.
1:59:26.2 Sean Carroll: Why not have a new force interacting with dark matter? It's a little baroque, and it's not the most promising thing in the world, but it's a possibility, and you don't know how to look for it until you do the theoretical work trying to understand what it's effects would be. So people are doing that. I love it. I'm glad people are doing it, because that helps guide our experimental and observational searches for new things.
1:59:50.7 Sean Carroll: Peter Schuler says, "When one says that quantum gravity as we understand it breaks down at extreme gravities, is it analogous to how Newtonian physics breaks down compared to relativity? It's almost always wrong with the error so small as to be practically irrelevant until you get to the extremes. Or is there somehow a sharp dividing line where quantum gravity as we know it, is completely right until you reach some threshold?" Well, I think it's the other way around. Quantum gravity is always right. The question is, where is classical gravity right? And the answer is the former of your two options there. Classical gravity is always an approximation. It's never exactly right.
2:00:27.4 Sean Carroll: The Earth moving around the Sun is described by a wave function. There is some uncertainty in the position and momentum of the Earth. It's just that that uncertainty is so incredibly low that you don't notice it. So there's no threshold at which the quantumness of gravity is supposed to turn on.
2:00:44.8 Sean Carroll: AJ says, "There have been some recent books by Ross Douthat and I don't know how to pronounce Ross Douthat's name, sorry. Ross Douthat and Francis Collins, that have taken the fine-tuning argument for the existence of God seriously, that have gotten some popular exposure on other podcasts or media. What is your best or most accessible argument against the fine-tuning argument?" I'm on record as saying the fine-tuning argument for the existence of God is the best argument that is out there for the existence of God, but it is still a very bad argument. And it's a very bad argument because the people who use it generally aren't being good Bayesians about their probabilities. They say, well, if God existed, this particular feature of the world would look like that. Therefore, this particular feature of the world does look like that.
2:01:32.1 Sean Carroll: Therefore, that is evidence that God exists. But you have to play the game consistently. You have to say, well, what other features of the world would you expect to be true if God existed? As soon as you open yourself up to being honest about what you would expect the world to look like if God existed, if you're really honest, you find out that the world would not look anything like the actual world that we observe. Most obviously, because God wouldn't treat us like this. [laughter] If God actually cared about us, God would give us much more explicit help here on Earth, even if that help is purely in the form of instructions on how to behave better. God has been very, very cagey, almost as if he doesn't exist.
2:02:18.9 Sean Carroll: But there's also a much more scientific objection to the fine-tuning argument. You can get out of it, but only at great cost, and the objection is the following that I've made many times before. The fine-tuning argument says, that the particular arrangement of matter and cosmological structure in the universe, plus the particular choices of the laws of physics, like the fine structure constant or whatever, the cosmological constant, all seem to be delicately tuned in such a way as to allow for the existence of life.
2:02:51.7 Sean Carroll: If you change various constants of nature by a large amount, physics would be dramatically different, and it's much harder to imagine life existing. So, for example, if the neutron were slightly lower mass than the proton rather than slightly higher mass, you wouldn't have any atoms. Everything would just be neutrons, and it would be much, much harder to have chemistry or life. There you go. Is that a fine-tuning argument? If the cosmological constant were anywhere close to it's natural value, there would be no time in the history of the universe to make galaxies, and there would be no life. There you go, another fine-tuning argument, okay.
2:03:26.7 Sean Carroll: But what is actually being allowed by these fine-tunings? These fine-tunings are allowing for the existence of complex chemical reactions. That's what you get by having atoms. That's what you have by giving the universe enough time to form galaxies and things like that. In other words, the thing that these fine-tunings are allowing is precisely the existence of physical structures that have the form of living beings. If you actually believed in the existence of God, you wouldn't need to think that you needed such structures.
2:04:02.5 Sean Carroll: Everyone who believes in the existence of God really sells God short a lot of the times in terms of what God could do. Typically, and again, not always, but most people who believe in God, are also not physicalists about consciousness. They think that there is some spirit or some essence of life or the soul or something like that, that is not purely physical. It would be weird to be a theist but otherwise be purely physicalist about life and consciousness. So consciousness is not something that is merely an emergent outgrowth of the chemical reactions or physical processes in your body if you are a theist.
2:04:45.9 Sean Carroll: So why do you need a fine-tune at all? [laughter] You don't need to set up a universe that allows for complicated chemical reactions if consciousness is not beholden to complex chemical reactions. Now, of course, I already hear the theist objecting to this, because there is a very typical theist move. You wait to see what the universe is like and then you say, yes, that's just how God would have wanted it. [laughter] So, since you know that in the real universe, consciousness is associated with complex chemical reactions, you come up with a reason why that's just how God would have wanted it.
2:05:21.9 Sean Carroll: Just like for the existence of life in the universe or human beings in the universe, you don't need a universe with a trillion galaxies in it. You could easily do it with a universe with just one galaxy, a relatively tiny galaxy, not that many stars in it. And the theist will have to say, yes, but it's easier for God just to do the Big Bang and make the whole universe. It's all very ex post facto and very unconvincing. So, I think that if you're a good Bayesian about these things, even with the fine-tuning argument, the data are pointing you in directions of physical understandings of why things are the way they are, not theological ones.
2:06:00.5 Sean Carroll: Tom Mallory says, "I got a PhD in chemistry, I couldn't find work, and I spiraled mentally. I ended up in a bad way drinking and using drugs. I'm sober now and doing okay, helped in part by finding and following positive inspirational role models like yourself. I have thrown myself into my work and research, taken up running, teaching myself a foreign language, and find time to practice playing the piano every day. I love all that I do, and sober life is infinitely better than using/drinking. However, I still find that despite the love I have for my work and hobbies, occasionally I feel exhausted by doing it all, and worried that I can't maintain it and I don't wanna end up in that dark place again. My question is simply, how does someone so prolific like yourself find time to keep up with everything and never seem to tire of it all? It's quite remarkable. Is it just discipline when you're not feeling it, or do you love what you do so much that you're always feeling it and never tire of it?
2:06:48.8 Sean Carroll: That's a great story, Tom. I'm glad to hear that you've come all this way. Teaching yourself a foreign language, learning to play the piano, taking up running, these are amazing things that a lot of people would be very proud of doing regularly, so good for you. In terms of me being prolific, and do I ever get tired of it? I get tired all the time. I'm tired right now as I'm speaking to you. There's a lot going on, and it's exhausting. Like many other people, nothing special about me here, I have a thing where there's a lot of things I enjoy doing, and I commit to doing them, and then it's too much, and I got to do it all.
2:07:23.7 Sean Carroll: And so, honestly, I get things done just by having deadlines that I have to meet and a feeling of guilt if I don't meet them. I am usually good at meeting them, but not always good. I have a book due. I got to finish that book. I have a podcast that has to come out every week. I have to start teaching quantum mechanics in the fall, so I need to think about how to do that the best. Just a bunch of things that the deadlines come up, the calendar keeps ticking away. And I very often say I should agree to do less things that would burden future me. [laughter] But on the other hand, it's absolutely true that every individual thing is rewarding and fun for me. So I'm in a very, very, very fortunate situation as a human being where most of the things that I have to do, even though they do tire me out, represent things that are individually enjoyable, whether it's the podcast or teaching or writing the book or whatever. So I'm very lucky that way. Not everyone is that lucky, and I absolutely recognize that.
2:08:24.1 Sean Carroll: So, there will always be dark nights of the soul, and you have to deal with them. But human existence is complicated and multifaceted, and there'll be ups and downs. And when you have a down, you have to realize there's gonna be an up. It's not gonna happen automatically, but you can make the ups happen. And it sounds like you're really doing that, and so I hope you can continue. It sounds great. I'm gonna group two questions together.
2:08:52.5 Sean Carroll: Adam Small says, "I always hear that it's basically a given that information can't be destroyed, so Hawking radiation is very important. But how come we are so confident that a black hole can't destroy information, given that our physics breaks down at the singularity or smaller than the Planck length, which I assume they crush everything in a space smaller, hence the singularity?" And then Babyfoot says, "In your view, what is the metaphysical nature of information?" So I'm grouping these two questions together because that Babyfoot's question, the metaphysical nature of information, is relevant to Adam's question about how we're so sure that black holes can't destroy information.
2:09:28.8 Sean Carroll: First, we're not sure that black holes can't destroy information. As we talked about at an earlier question, we have reasons to believe it, but you're never sure about these things. So you remain open-minded. You try to figure out, the black hole information loss puzzle is a puzzle because it is physicists asking themselves, given what we think we know about gravity and quantum mechanics, how can we arrange things so that information is not lost because we think that other principles of physics lead us to believe information shouldn't be lost? But maybe it won't work. We're trying. We had talks here on the podcast with people like Netta Engelhardt, Raphael Bousso. We've discussed these things, but the final answer has not quite yet been written down, so stay tuned on that. But more importantly, I think that the phrasing of Adam's question, gives rise to this idea, or at least stems from this idea, that information is kind of a thing, like a bit of information takes up space. And that's an interesting implicit assumption, because in a sense it does, because information is tied to physical things. This is where we need to answer Babyfoot's question. What is the metaphysical nature of information?
2:10:44.0 Sean Carroll: I know that in physics, there's a motto that information is physical, which is a way of saying that manipulating information requires physical capacities. You generally increase entropy or you require some energy or whatever it is. You can't just push around information without physical goings on. So the way that I like to think about it though, is rather than saying information is physical, I say that information is a way of describing physical stuff. It's the physical stuff that is real. So when you say, I'm crunching things down to the singularity, you're not crunching information, really. You're crunching stuff, whatever it is. Maybe it's just an abstract quantum mechanical wave function that you're crunching. And that's in a regime where we don't know what's going on. That's why it's interesting. But it's not like there's information bits that are the physical things you're pushing around. We have bits, quantum bits, qubits, and we attach information to them in the way that we describe it. This sounds fuzzier than it really is. Energy or momentum are exactly the same thing. There's not energy out there in the universe separate from stuff.
2:11:54.6 Sean Carroll: Energy is a property that stuff has, and so is information. Okay, that's the metaphysical nature of information. It's a property that is attached to physical stuff. As far as Adam's question and the singularity is concerned, so number one, we're not so confident, but number two, the interesting thing about the black hole information puzzle, is that it doesn't require understanding what's happening at the singularity. As long as you think, and maybe it's not true, but as long as you think that information is completely preserved from the time before you made the black hole, to the time after it evaporated, the puzzle about how the information gets out, is there regardless of what happens at the singularity. Because the information is coming from just outside the event horizon, not from the singularity. So that's why people are so interested in the puzzle. It seems to be a puzzle that cannot be resolved just by playing with physics near the Planck length or near the singularity.
2:12:54.9 Sean Carroll: Marie Roscoe says, "I recently rewatched the 2024 Philosophy of Physics workshop on YouTube, and at the beginning you said, that you sometimes promise to give a talk on a subject you haven't yet done the research on and hope it will be done. Sometimes it is, sometimes it isn't. I wonder how this strategy has worked for you." Yeah, this strategy has become more common later in life. Honestly, when I was younger, I would wait until I actually did some research to promise to give a talk on it or whatever. These days, going back to the previous question, one of the ways I keep myself motivated to do some work is to say, all right, I promised to give a talk on it. I guess I better think about it. It doesn't always happen that I get a satisfactory answer, because you can't really predict the pace of progress in theoretical physics or philosophy for that matter. I think, again, I have a lot of things going on, so I need to motivate myself in various ways. In some sense, it's been useful to promise to give talks on things and then be forced to follow up on them. In another sense, I kind of wish I could be more responsible and just do the work first, and then give talks on it later.
2:14:05.8 Sean Carroll: The last four years of my life have been very disruptive, moving across the country, starting a new job, starting up the Natural Philosophy Forum, teaching a bunch of courses that I'd never taught before, trying to finish a bunch of books. A whole bunch of things have been going on, so I'm a little bit less responsible about getting the work done before the talk than I have been, but I hope to return to a level of responsibility that I used to have back in the day.
2:14:29.7 Sean Carroll: Rory Edwards says, "I'm coming to the end of my PhD in experimental neutrino physics and plan to transition to a more hybrid role in medical physics. I still love to learn about fundamental theoretical physics like foundations of quantum mechanics. Of course, I can continue to read textbooks and watch lectures for my interest, but what are your thoughts and recommendations for keeping up with research and developments and continuing to learn in pure/fundamental physics as more of a hobby for a non-layperson?" Well, I think you have a good start. You're listening to this podcast. That's something. I think it very much depends on exactly what you mean by keeping up with research and developments.
2:15:09.9 Sean Carroll: This podcast, for example, makes no pretense to giving you the latest scoop on anything at all. This is a much more slower timescale, big picture, think about things kind of podcast. I wish actually, this is a hint for folks out there who are physics oriented and thinking about starting a podcast. A really good top-notch podcast that really was physics news of the month or an in-depth dive to the most recent interesting physics story, that would be fascinating. I would love that and definitely promote it myself if it were done well. You have to actually put some effort into it. But if you want to go further than that, there's lots of books, there's lots of papers, there's lots of online courses and things like that. I'm just gonna say, it's hard and I can't give you much more specific guidance than that, because there's a very, very large amount of stuff that gets done at the research level. You could just go to archive.org where all the papers appear and you could find your favorite subcategory, HEP-TH for formal theory, HEP-PH for phenomenology, ASTRO-PH for astrophysics or whatever.
2:16:26.9 Sean Carroll: And you could subscribe to daily mailings. You could get abstracts in the mail that are the most recent research in that area and that's something that many people do who are professional physicists. And then you don't need to read every paper, but you would get links to every paper and you can read the titles, you can read the abstracts, you get a feeling for what is going on in the field and then you could actually read the papers that you thought were interesting. It's overwhelming, though. It's a lot. [laughter] A lot of, when you reach a certain age, as I have, you depend on going to seminars or going to conferences to hear what people are talking about for a feeling for what people think is interesting, and also having your students and postdocs tell you, "Oh, don't miss this paper. This is a good one." As well as, just keeping track of the papers online, 'cause sometimes your eyes glaze over, I will confess. There's a lot of university departments that put all their seminars online. So that's one way to, again, get a feeling for what issues are interesting.
2:17:30.0 Sean Carroll: It's not just, here's a million papers, read them all. There are themes within the papers. Oh, there seem to be a bunch of papers on something to do with islands and black hole information. What is that about? You can absolutely catch on to those kind of trends by following it. And it's a little bit of work. It's a little bit of work for all of us. You're basically saying, how can I do a hobby, something that people do also as a full-time job? Of course, if you're not trying to do the research end of it, just reading up on what is happening, then it's a little bit easier, but it's still a lot. So, I absolutely encourage you to do it, but I'm not gonna fool you that it's not a lot of stuff to keep up with.
2:18:14.4 Sean Carroll: Peter Solfest says, "How would you think through determining the amount of federal funding for science at a high level? The recent proposed cuts are worrying, but in a larger sense, I realized I don't know how to think through this. The marginal cost of most science is pretty small compared to the federal budget, but at some point spending would clearly be too much. More money than scientists or in the limit, 100% of the US GDP would clearly be too much. Do you have any thoughts on how to rationally get to an actual top line figure?" I hate to be kind of boring about this, but I think that the best way to do it is pretty close to the way we've been doing it until the last few months. That is to say, you have layers. You have layers of expertise with different amounts of knowledge about the constraints and the promise of different things. At the very micro level, if you just want to hand out grants to specific researchers, and you're faced with grant applications in quantum information from me, and from Scott Aronson and from John Preskill, who do you decide who to give money to?
2:19:16.8 Sean Carroll: That has to be in the hands of experts, and that's exactly what something like the National Science Foundation does. They impanel a bunch of other experts who are not involved in the research proposals, and they rank them and they give out money. And then at the higher level, at the level of the NSF organizational structure, there are debates about, okay, so how much money goes to quantum information versus how much money goes to particle physics, how much money goes to biological physics, whatever.
2:19:43.4 Sean Carroll: And in all those cases, they get input from experts, but then there have to be some higher level people who decide, who is making the best case. What is the promise of the science that we're being offered? What has been the track record of results? How effective is the actual set of proposals that we're getting? How convincing are they? And so forth. And then, at an even higher level, you have to decide how much money to give the National Science Foundation. And that's gonna be Congress. It's not supposed to be the president. I know that we're throwing out all of the laws of the country right now, but the laws say, it's supposed to be Congress deciding this. And Congress, of course, is not a bunch of people who are experts in science policy, but they have staffs, they have committees who are supposed to get information from the experts and process that information and make decisions.
2:20:35.1 Sean Carroll: Obviously, there's no clear algorithm for doing that. You're saying, science would be good, but of course other things would also be good. How do you even compare them to each other? I think that it would be very, very easy to make the case that science is underfunded in the United States, even before the recent cuts, because number one, the return on investment from science is huge. Giving money to universities to do research in many, many ways gives back multiple dollars to the nation, both in terms of new scientific breakthroughs, but also in terms of just the local economy at the university or in the town where the university is located. Spending money gives you return back at an enormous, hilariously large multiplication factor.
2:21:21.9 Sean Carroll: And the other thing is, you could look at the cutoff, like look at who's getting funded, who's not getting funded. Are the people who are not getting funded promising to do pretty good research if they had been funded? I know in my areas, the people who don't get funding are often really good. I've been on the panels for NSF, for DOE, for NASA, giving away grant money, and it was just tough, because we had very famous scientists who had done amazing work throughout their career, and we couldn't fund them because this specific grant proposal wasn't quite as promising as somebody else's. I think it's very easy to make the case that there's a lot of good science that is not getting funded in the current system. So that's the case I would try to make.
2:22:08.8 Sean Carroll: Kevin's Disobedience says, "Having read your books, you've definitely raised my credence for many worlds from effectively zero to something closer to 60%. But my concern is still with the Born Rule. Are you satisfied with rational decision theory as an explanation? You get the right answer, but it seems kind of ad hoc. For me, this is the weakest part of the theory, especially because you already know the answer you want to get." So, Kevin, I don't know if you're Kevin or you're just disobeying Kevin, Kevin's Disobedience, but I don't think you've read my books closely enough, because I don't think that the decision theory argument is the most convincing one. I think that there's an argument based on self-locating uncertainty, that is a more convincing one. But I also agree that they're not 100% convincing. And the reason why I think they're not 100% convincing, even though they seem, I think these arguments seem extremely reasonable to me and I accept them basically completely, they do require leaps of metaphysical imagination. What we're asking you to do in these arguments is a kind of reasoning that is very different than what we have to do in our everyday single universe experience lives.
2:23:14.2 Sean Carroll: And that's always a place where you can get tripped up. Especially because, as you know, you know the answer you want to get. But as I always emphasize, once you look into the math and the logic of the Born rule and many worlds, the correct answer, the Born rule, is really the only sensible answer you could possibly get. There's no chance that you're gonna say the wave function cubed is giving you the probability or anything like that, or even the wave function itself. It has to be the wave function squared. The only argument, is not between the wave function squared versus some other probability distribution, it's between the wave function squared and you can't get probabilities out at all. And I think that the case to be made that you can get probabilities out at all is very, very good. So I'm not really worried about it myself, but I feel for the people who are worried about it, that I don't think that they have quite gotten to the right place in their thinking through the issues here, but I suspect that given enough time they would do so.
2:24:16.7 Sean Carroll: Ben Lloyd says, "Any thoughts on the Sixers getting the third pick in the NBA draft? I'm excited, we couldn't capture the flag, but we still have some great options to add to our team." Yeah, I have mixed feelings about the whole thing. I have to say, it's been tough this year being a Sixers fan. For those of you who are not basketball fans, you don't know or care, but the Sixers came into this year with enormous hopes and pretty good odds, like it wasn't just us talking ourselves up. We had a great team as far as the judgment of the rest of the world was concerned, but then an unprecedented set of injuries just hit the team. All the best players were injured for many, many games, and so we did terribly. The Sixers did, I say we. And as a result, we finished near the bottom of the league rather than near the top, and we got a good draft pick out of it. In this year's NBA draft, there usually is hard to completely rank players, 'cause players are very different, but so people try to put them into tiers. There's the first tier, the second tier, the third tier.
2:25:18.6 Sean Carroll: For what it's worth, the consensus this year is there's a very clear first tier, that only has one player in it, Cooper Flagg from Duke. That's why capture the flag is what people say. And there's a pretty clear second tier with Dylan Harper from Rutgers University, and only one person in the second tier. So then the third tier, has a lot of different people in it. And guess what? The Sixers are picking third. So on the one hand, they get their choice out of that tier. On the other hand, that tier seems to be a couple of steps lower than the other two. So we'll have to see how it goes. I guess my frustration is not that there were injuries. That is frustrating for sure. But the discourse around the team and the sport, et cetera, is really terrible, because people act as if injuries are some kind of moral failing. Or even judgmental failing. Like, oh, the team was constructed badly because they did badly. No, they did badly 'cause they got injured. And of course, there's a probability that you will get injured and you can take that into account.
2:26:19.5 Sean Carroll: But you got to do the best you can. And I was completely on board with the team building strategy that we had before the season started. Everyone has gotten their surgeries in the meantime, and we're hoping everyone comes back at full strength. And I think it's completely possible, that the Philadelphia 76ers win the NBA championship next year. It's also completely possible, they have a losing record again because that is in the nature of injuries, that it can be a boom or bust cycle. We'll have to see.
2:26:50.7 Sean Carroll: Anonymous says, "Do you view what Brutus and the Roman Senate did to Caesar as heroic or villainous? Can actions like theirs ever be morally justified?" Well, I 100% think that their actions can, in principle, actions like that, can be morally justified. There can be times when you have to step out of the legal framework to do something because the legal framework is being distorted by autocratic influences. And in the case of Julius Caesar declaring himself to be the emperor of Rome, that is certainly arguably a case where that is absolutely happening. I certainly don't know enough about the details of Caesar and Brutus and so forth to know whether, in that particular case, the assassination was morally justified.
2:27:34.4 Sean Carroll: But I am a believer that that can, in principle, be morally justified. The sad news about it, is that it's individuals who have to make that decision and individuals are terrible at making those decisions. And that's why I think there should be a huge presumption against doing things like that. You better be rock solid certain that not only is it the morally right thing to do, but the practical implications of what you're doing are the right thing, and that you are willing to take whatever consequences for that action might happen. And no matter how terrible for you they might be. So not recommending ever doing something like that, but I can imagine, philosophically, in principle, it being morally justified.
2:28:15.6 Sean Carroll: Robert Ruxandrescu says, "If the many worlds interpretation and inflation are correct, shouldn't this mean that since since the decay or stop of inflation is a quantum process, there are potentially infinite branches in which inflation stopped after 0.2 seconds, after 1.7 seconds, after a thousand years, and so on, and we live in one of those branches and observe one of potentially infinite ways in which the cosmic microwave background could look like? In other words, a sort of inflationary anthropic principle?"
2:28:42.0 Sean Carroll: Yeah, 100%. That's absolutely possible. Basically, this is just the many worlds way of saying eternal inflation. Eternal inflation and the usual way of thinking about it, is different regions of space have a probability of inflation continuing for a while versus ending. And if space is big enough and expanding fast enough, there will always be some region in which inflation continues. The many worlds version of it is to say, even in a single region, you have branching into different branches where inflation has ended at different points. In fact, if you wanna look up a blog post I wrote a number of years ago called, Are the Multiverse and Many Worlds the Same Idea? I've often said, and people do try to say, that there is a difference between the many worlds interpretation of quantum mechanics and the inflationary multiverse. In many worlds, there are different branches that are next to each other in Hilbert space. In the cosmological multiverse, there are different regions of space where conditions are very different. But there were two papers that came out, one by Raphael Bousso and Leonard Susskind, both former Mindscape guests, and the other one by Yasunori Nomura at Berkeley.
2:29:57.4 Sean Carroll: And they both pointed out that if you sort of combine ideas from inflation and also horizon complementarity, which is a way of saying that you shouldn't think about space-time outside our cosmological horizon, you should only think about what's inside and on the boundary. Then basically the quantum state of the universe for any one observer is a superposition of many different parts of the landscape of inflationary possibilities. So you get this kind of anthropic multiverse even for one observer, or even for one region of space-time, if that even conceptually makes sense. And of course, this is exactly what bothers people like Paul Steinhardt, who was one of the pioneers of inflationary cosmology, but now worries that it doesn't predict anything because it predicts everything.
2:30:46.2 Sean Carroll: I'm not nearly as worried as Paul is, because I think that that's just a question for which you need to have the right probability distribution. Like, yes, it predicts everything, but it doesn't predict everything with equal likelihood. If it predicts that some things are common and other things are uncommon, then that's fine. We just deal with probabilities. And that's probably because I'm used to doing this in the case of many worlds. So I'm just, I don't think that we know the right probability measure in the case of inflation quite as well. But if you dig deep into the paper I wrote with Charles Sebens about probability in many worlds, there's a little section in there on cosmological probabilities that at least moves towards answering that question.
2:31:28.7 Sean Carroll: Daniel Crespo says, "Can you explain more why life is a generator of entropy?" I can try. I'm not sure what the word why means there. Most things are generators of entropy in the world. Not everything. The earth going around the sun, it doesn't really generate any entropy. It's a very sort of clean kind of thing, but the sun itself generates a huge amount of entropy. You don't need life to generate entropy. Any dissipative, any process with friction or air resistance, et cetera, is going to increase the entropy of the universe. Life in particular, needs to generate entropy because it is an example of a system that is not in thermal equilibrium. And it's also not in mechanical stable equilibrium. In other words, there's a difference between the earth and the sun. The earth is solid basically because the atoms and molecules in the earth support each other through their pressure.
2:32:31.7 Sean Carroll: And you don't need any motion. Everything can be completely static. If the earth's core cools off and the temperature goes to zero, it will still be solid. Whereas the sun is a very different thing, because the sun is kept up by burning fuel. Its pressure, the hydrodynamic equilibrium inside the sun is maintained by that energy source at the center in the form of nuclear fusion, which creates entropy. So when the sun stops burning it's nuclear fuel, it's gonna collapse. Well, it's gonna puff up first, and then there's an elaborate story of the end of a stellar lifetime, but eventually it will collapse to a white dwarf.
2:33:08.4 Sean Carroll: And that's because it is currently maintained in kind of a quasi-stable state by generating entropy. Life is like that. Life is obviously much more complicated, but life is an example of a system that needs fuel from the outside world to keep it's processes going. And in the using of that fuel, it degrades the energy from a low entropy state to a high entropy state. It needs that fuel to repair itself, it needs that fuel to generate memories and thoughts. It needs that fuel to do everything that life does, to locomote, to gather other fuel and food, to sense the world, to process the information. All of these processes involve the generation of entropy. So I'm not sure if that's quite answering your question, but life is deep into the regime where generating entropy is fundamentally part of what it is.
2:34:09.5 Sean Carroll: Ryan Cobine says, "I've heard Copenhagen interpretation-like statements by public intellectuals a few times in the last few months. For example, Ross Douthat on Conversations with Tyler from February, "I deny that you can have a measuring stick without a process of consciousness. The measuring stick without a process of consciousness is itself just a ruler. Absent your consciousness, a ruler is a collection of atoms and molecules cut out of a tree with some markings on it. In order for the ruler to be an instrument of measurement, you have to be perceiving it and to be conscious of it." What do you believe to be the best way to respond to these kinds of statements when made in face-to-face conversations? I have a notion of something's wrong with this, but I don't feel my understanding is solid enough to be able to be coherently educational."
2:34:55.2 Sean Carroll: Well, the problem is with this statement, the problem with responding to this statement, is it's as stated, it's just a bunch of assertions. In order for the ruler to be an instrument of measurement, you have to be perceiving it and be conscious of it. Why? If I have a camera, video camera recording something and it records a ruler being put next to something else, is that a measurement or not? And the question becomes, what do you mean by measurement? If your definition of measurement, is it needs to be recorded by a conscious agent, then it's a vacuous statement. Yes, then you need consciousness to do a measurement. If you have a more objective notion of measurement, then clearly you don't need a conscious agent to do that. That statement didn't specifically invoke quantum mechanics, but so you're right, it's Copenhagen-like without actually being quantum mechanical. But so, you have to make this choice. Are you just defining measurement to be only as perceived by conscious creatures or not? If you are, then that's fine, but that doesn't mean that consciousness is anything special.
2:36:01.1 Sean Carroll: Conscious creatures can still be perfectly physical and they can do the measuring. There's nothing mystical or ineffable about that. But I think that it's just better to define measurement in a more objective way that needn't assume consciousness from the start.
2:36:16.5 Sean Carroll: Steve Bonner says, "I recently watched a talk you gave about a year ago, Saturday Morning Physics, where you speculated that distance could be an emergent property of entanglement, and this could happen even in totally empty space. Two points in empty space are said to be close together if they're heavily entangled, and further apart if they're not. But how can two points in empty space be entangled? For things to be entangled, don't there need to be things?" [laughter]
2:36:39.0 Sean Carroll: Yes, there do need to be things in order for things to be entangled, but the idea of a thing becomes rather abstract in quantum mechanics. One way of saying it is, you need a degree of freedom in order for there to be things that are entangled. We don't know the fundamental degrees of freedom out of which nature is made, but certainly even in quantum field theory, where we do have a pretty good idea of what we're claiming to exist, empty space is not empty.
2:37:07.2 Sean Carroll: It's not featureless is what I should say. There are quantum fields in empty space. And one way of thinking about this question at a level where we do understand what's going on, is to talk about the amount of entanglement between the modes of quantum fields in different regions of space-time, even in the vacuum. The vacuum is just the lowest energy state of a quantum field theory, but that doesn't mean the quantum fields disappear, it means that they have a certain state, which is the vacuum state. This is confusing because classically, you can say, okay, maybe I have an electric field, and the vacuum state of the electric field is the electric field equals zero. There's no electric field there in some sense. But that sense is not the right sense. That's not really the really rigorous way of talking about it. When the electric field equals zero, even classically, it's still there. There is an electric field, and what I mean by that is you could poke it. You could perturb the electric field and suddenly it would not be zero anymore. So we don't say that there isn't an electric field there.
2:38:13.8 Sean Carroll: There is always the possibility of electric field, it was just equal to zero. Once you're doing quantum mechanics, the whole story becomes even more subtle. Because it's not even true, that the electric field value is zero quantum mechanically. You have to be careful and talk about the state of the electric field, and that state is gonna be some particular field with a wave function, et cetera, et cetera. Were you to measure it, classically if you measure the electric field, when the electric field is in it's vacuum state, you get zero. Quantum mechanically, you will get some distribution of values around zero, but they will not usually be or always be zero itself, because of the nature of quantum mechanics. So anyway, the point is, even empty space has stuff in it in quantum field theory. There might be a deeper explanation where quantum field theory is not the right way of thinking about things, but there are still quantum mechanical degrees of freedom, which if you like, you can say are in empty space, but it would be even more accurate to say, which make up empty space at the emergent level. It's the quantum mechanical degrees of freedom that are more fundamental.
2:39:19.3 Sean Carroll: Empty space is something that emerges out of that. And in this picture that I was sketching in the lecture, those degrees of freedom have an entanglement structure. Even if there's no particles or non-zero fields, there's still entanglement between the degrees of freedom that make up empty space itself.
2:39:39.3 Sean Carroll: Alex DuBrow says, "As a materialist, I believe there is an objective physical world governed by consistent laws, but neuroscience shows that everything we experience is generated by an internal model in the brain, and even our sense that this model corresponds to reality is itself part of the model. Nothing exists for us outside of the model, can it? Do you think materialism taken seriously requires us to acknowledge that we can never directly access the objective world, and that all evidence for it's nature is filtered through this recursive model? And if so, what implications do you draw from that realization?" Well, I would deny that neuroscience shows that everything we experience is generated by an internal model in the brain. I don't know about you, but I have eyeballs and ears and fingers, and I interact with the outside world. And a much more accurate way of saying it, is there's a constant flow of information between my brain and the outside world.
2:40:32.3 Sean Carroll: It's not like my brain is doing all the work and coming up with a model of reality. My brain absolutely comes up with a model of reality, but that model is influenced by the outside world. Now you say, let's see, where did it go? Requires us to acknowledge that we can never directly access the objective world. I think the words directly and objective are doing an enormous amount of work there. Why do we need to access things directly? Of course, everything is mediated by something else. That's fine. But as scientists or even as rational human beings, we try to make the best model of the world given the data we have. We were able to come up with the idea of quarks inside protons and neutrons, even though we never directly access quarks. Why? Because we say, well, if quarks were in protons and neutrons, what would the implications be? And then we go looking for those implications and we find them and they're all true. So we say, I think my theory is right. I think that there are quarks out there. At a somewhat less rigorous level, this is how we make pictures of the world in our brain.
2:41:35.7 Sean Carroll: We say, well, if things were like this, these things would follow and then we interact with the world. Were those babies pulling on things and seeing what happens? And the objective outside world is absolutely playing a crucial role in my model of what's going on. I don't know about anybody else's.
2:41:51.8 Sean Carroll: Tom Arabia says, "Who was your pick for NBA MVP, Jokic or SGA?" SGA is Shai Gilgeous-Alexander who is a young player for the Oklahoma City Thunder. "Have any favorites to win East-Western Championship?" Cheers. I'm completely unqualified to say who was the MVP this year. I was just not following the NBA closely enough, both 'cause I was busy, and because the Sixers were terrible. So I really didn't see these players play. I'm not gonna say who won. But by now, SGA has already won. So congratulations to him. Nikola Jokic, who is clearly one of the best basketball players alive or ever, for that matter, has won a few MVPs already. So I think that people wanna spread the wealth a little bit. So I don't really know what the objective call would be there.
2:42:38.0 Sean Carroll: Do I have any favorites to win the East-Western Championship? So right now, for those who don't know, it's New York Knicks versus Indiana Pacers in the East, Oklahoma City versus the Minnesota Timberwolves in the West. And Indiana is up 2-1 and Oklahoma City is up 2-1. And I think that I would just vote for the people who are ahead right now. I'm not gonna pretend to have any special insight. In fact, I will anti-pretend to have special insight. On the West, I thought that Oklahoma City would clearly romp through the playoffs and they more or less are. They struggled with Denver a little bit, but they're, I think, a really, really good team, even by historical standards. In the East, I didn't think any of the teams were that good, but I thought that Cleveland was the best and they flamed out very early. I thought that Indiana had no chance and they're doing the best. So what do I know about basketball? I can say that my boy TJ McConnell, former Sixer, now plays for Indiana. So I'm rooting for him. I would love Indiana to win, even though I didn't give them much of a chance at the beginning.
2:43:40.0 Sean Carroll: Conor Kostick says, "In conversation with Annaka Harris, you were quick to dismiss the idea that chairs could be sentient. Fair enough, I agree with you. But I got the impression that she had a different perspective on that to you. This discussion however, moved on another direction. Remembering how Andy Clark made the case for the extended mind, one might be able to easily rule out chairs on their own as being sentient, but perhaps it's not so straightforward to rule out chairs as ever being part of a mind. Do you think that inanimate objects like chairs can ever be part of something we would recognize as a mind?" Sure, in principle, it could happen. I think that you need to make the case that that is the best way to carve up the world. This is the thing that the mind, the idea of minds, in my view, as a materialist, as a physicalist, are not part of the fundamental architecture of reality. They're part of the emergent higher level. So you tell me what your picture of the emergent higher level is, in which the idea of a mind includes a chair. And I will tell you mine in which it doesn't include chairs. And we'll see which maps on better to what actually happens out there in the world.
2:44:45.0 Sean Carroll: I'm open minded about this. I do think that there's a case to be made that you can think of pocket calculators or smartphones as part of your extended cognition, sure, or even pads of paper and pencils. The question is, how useful is that? Is it better to think of them as that, rather than merely just aids to the actual mind, which is in your brain? I don't know. I'm open minded about that, like I said, but you'd have to make the case that that's the best theory that we have.
2:45:16.9 Sean Carroll: Gregory Egan says, "Has the extrapolated size of the universe changed since recent JWST revelations?" Nope, not really. The James Webb Space Telescope has found galaxies earlier on in the history of the universe than we might have guessed, but that's just very normal science-y. Like, okay, we found some galaxies forming early. What that means is not the universe is bigger, it just means galaxies formed faster than we thought they did. Okay, that's interesting, and we'll try to figure it out. The size of the universe doesn't really depend on the earliness of galaxies. It depends on, if anything, the Hubble parameter and the total amount of stuff in the universe.
2:45:54.6 Sean Carroll: And as you know, from our recent talk with Marc Kamionkowski, as well as earlier talks with Adam Riess, et cetera, there are serious worries about things like the Hubble tension, the possible evolution of dark energy. We don't know exactly, but look, all of this is 10% level stuff. These are small tweaks on the size of the observable universe. When I was your age, we had a factor of two discrepancy. Some people thought the Hubble constant was 50, other people thought it was 100. Now we're arguing over is it 68 or 72. So it's a much different world in which we live in. There's not gonna be any radical overthrow of what we think about the size of the universe.
2:46:33.3 Sean Carroll: Casey Mahone says, "Have you watched the show Severance? What do you think of all the philosophical questions that the show poses about identity and consciousness?" Yeah, this is a great question. Actually, David Chalmers on Twitter, I think, asked the question, do we think that the innies and outies are separate people? So for those who have not seen Severance, the conceit of the show is that you can have a job, you go to work, and they basically sever a version of you that works inside, and you call that the innie, from the version of you that lives the outside life.
2:47:05.3 Sean Carroll: And so, you have your old ongoing consciousness as an outie, and you live your life, and you go to work, and the outie doesn't know what happens inside the job, and the innie has no memory of what the outie was like or what they did or who their marriage partner was or anything like that. And there is some magic process in the elevator when you go to work that flips you from innie to outie and back. Of course, this is problematic for all sorts of reasons, but we're just asking the philosophical question, are they two different people? I think they're clearly two different people? Why would you say you're the same person if you have two different ways of talking, sets of memories, all that stuff? It's interesting because of course, as happens in the show, you can imagine trying to unify them again, and then what would that mean? Is one the boss? Should you share the experiences? Are they both real? Can you cheat on your partner? If your innie does something your outie doesn't know about? These are all questions that the show has a lot of fun with.
2:48:05.4 Sean Carroll: But let me just point out that one of the reasons why these questions are hard, is the innie and outie a different person, is because they're not real world. Because we have never in human history actually dealt with a question like this. So if our notion of what a person is, is an emergent higher level thing, that we develop through the experience of the world, that notion is just not equipped to deal with these situations, and that's okay. We shouldn't expect to have a clear-cut answer to some of these questions because we're being faced with situations we've never dealt with before. That's okay. We shouldn't say, here's the right answer. We should expand our space of what is possible and how to think about these things, and I think the show does a good job of encouraging us to do that.
2:48:52.8 Sean Carroll: Rob Adkerson says, "How do you combat the despair of knowing most people don't or can't understand even the basics of your work?" I have to deny the premise of this question, 'cause number one, I think that most people can understand the basics of my work, and I would have no despair if they couldn't. I don't see why that would be a cause of despair. If anything, it's exactly the opposite. I am astonished that as many people are interested and try to understand the basics of my work as they do, for example, all of the lovely Mindscape listeners out there, listening to this right now. Most other human beings here on Earth, that's okay. I don't think that having an effect on most other human beings is the point. I enjoy doing my work. I enjoy talking about it to other people. Some people seem to be interested in it. I think that's all great.
2:49:42.8 Sean Carroll: Nick J says, "Is entanglement a binary concept, I.e. Things need to be entangled or not entangled, or is there an underlying entanglement-ness, I.e. Things can be more or less entangled? It's definitely the latter. Entanglement is not binary at all. You can have two quantum systems, and they can be a lot entangled or a little entangled or anywhere in between. There is a maximum amount of entanglement, that depends on the size of the systems, and there's a minimum amount of entanglement, which is zero.
2:50:11.0 Sean Carroll: But anything in between is completely allowed. And in fact, a lot of people in the world of quantum information theory, put a lot of thought into thinking about, well, how should we think about what happens when you have three systems that are entangled with each other or multiple systems? Multipartite entanglement entanglement is something you need to worry about.
2:50:31.4 Sean Carroll: Igor Kopylov says, "I'm a bit confused by the relationship between geometry and topology in general relativity. My understanding is that the geometry of spacetime is determined by solving Einstein's equation, but is the topology also, or is it something you're given separately? Like, space could be flat and infinite, or flat and wrap around. Is that determined by the evolution over time, or do you assume you're doing GR in one topology or the other to begin with and go from there?" Largely, it's the last, the latter. You assume that you have a topology, and you have an initial geometry, and then you can solve the equations to see where you go from there, as you said. There are however, some constraints, and this is the fascinating aspect of differential geometry or even differential topology, if you wanna call it that, that if you have local conditions.
2:51:19.9 Sean Carroll: So, if you have not just curvature on a manifold, although that counts, but more generally some knowledge of what a function is doing at a particular location in space, and the derivatives of that function, and you have some reason to believe that there are conditions that are true at every individual point, those conditions can add up to a constraint on what is allowed topologically globally. So if you think about a two-dimensional sphere, at every point on a two-dimensional sphere, by which we mean the surface of the sphere, the inside of the sphere is called the ball, the outside, the boundary is called the sphere. So in every direction, there's curvature. On the sphere, it's sort of curving away from you uniformly in every direction. And if that's true at every point on the sphere, then the sphere, or any surface that has that property, must be a sphere, is the right way to say it. It can't go on forever. If you have positive curvature greater than or equal to a certain amount, your manifold must be topologically compact. Whereas, if the curvature on average is zero, then it may or may not be compact.
2:52:28.1 Sean Carroll: So it does not determine the global topology or even any constraints on the global topology. You can have an infinitely big flat surface, or you could have a torus or something like that that is also flat. So there are some relationships between topology and geometry, but it's not an absolute connection.
2:52:46.5 Sean Carroll: David Whitaker says, "Is the concept of simultaneity meaningless, or could one imagine an event taking place in a distant galaxy light years away at exactly this moment or the next moment? Does that have any meaning?" It has meaning, but it's subjective. It doesn't have absolute meaning. That's the lesson of relativity versus Newtonian absolute space-time. In Newtonian absolute space-time, the idea of an event taking place in a distant galaxy at exactly this moment, is something that is perfectly well-defined and agreed upon by everybody. There is a fact of the matter about whether or not something is happening at this moment very, very far away. In relativity, there is no fact of the matter about whether something is happening at this moment very, very far away, but an individual can make choices about how to divide up space-time into space and time, and subject to those choices, you will have an answer to the question, what is happening at exactly this moment at a galaxy very far away?
2:53:47.5 Sean Carroll: A different person might make different choices, and then they have a different coordinate system, a different way of slicing space-time into space and time, and to them they will get a different answer to that question. So whether or not something is happening at the same time in a distant galaxy is relative to how you're deciding to slice up space and time, and that is not objectively chosen.
2:54:10.9 Sean Carroll: Rue Phillips says, "What are some of your favorite non-fiction books or subjects to read about? I'm curious more about your guilty pleasures rather than highly technical books for your work." Well, for non-fiction books, I guess I don't have a lot of guilty pleasures in the non-fiction category. Maybe the occasional biography or history book, but honestly, non-fiction books, most of my interest in non-fiction you can pretty much work out by who appears on the Mindscape podcast. This is the stuff that I'm interested in. Or there's stuff that I'm reading, again, for work, for the future projects like Physics of Democracy or learning about complexity or things like that, that have some impact on Mindscape, but it's not exactly the ratio you might think. Most of my guilty pleasures are fiction books, both genre fiction and literary fiction, science fiction, mysteries, contemporary fiction. It's just whatever seems to grab me at any one moment in time.
2:55:07.6 Sean Carroll: Tim Converse says, "I enjoyed the Matt Strassler episode enough so that I followed up and read Waves in an Impossible Sea. Strassler really delivered on his project of giving new intuitions on well-known physics using words only. I was startled by one explanation though, because I hadn't seen it elsewhere. He explains the rest mass of the proton with reference to the high velocity of it's internal components, and then says that this velocity is a legacy of components' velocities from when the protons were formed. 'Now trapped forever inside these droplets, protons and neutrons, quarks, gluons, and antiquarks still dash around at speeds at or near the speed of light, colliding again and again. The bedlam of the Big Bang is caught within, never to escape or fade away.' What do you think of this explanation? It is poetically evocative, but I'm curious whether you recognize it as a mainstream view."
2:55:55.0 Sean Carroll: I think, no, I don't like that explanation at all for two reasons. One, it does, as Tim says, give the impression that there's some relic remnant of the conditions near the Big Bang, which is just wrong. A proton that you made ab initio today would have exactly the same things going on in it as a proton that was left over from the Big Bang. But secondly, I just don't like to talk about quarks dashing around near the speed of light. I understand what Matt's trying to do. He's trying to give some visceral or visual understanding of what is fundamentally a quantum field theory problem. But my way of talking about it is just to admit that it's a quantum field theory problem and point out that the quantum state of the quark fields inside the proton is 100% static. It is not changing from moment to moment in time. There is nothing dashing around in any sense. I think that that is a leftover from our view of the world as fundamentally classical, and we try to force quantum phenomena back into these classical boxes. It leaves us with the wrong impression, in my view.
2:57:05.4 Sean Carroll: Eric Stromquist says, "Have you ever read or studied the philosophy of Nietzsche, certainly a naturalist, although perhaps not a poetic one, or if not him, Arthur Schopenhauer, or if not him, any of the existentialists? If you have, what is your view of it?" I certainly wouldn't claim any expertise on these guys. I've read little bits and pieces of many of them. I've read more Nietzsche than Schopenhauer. I even supervised an independent study course that two undergrads did here at Johns Hopkins on Nietzsche. They had a specific idea about using the will to power to think about dynamical changes in living systems that involve science as well as philosophy, so I was an appropriate person to supervise them. And certainly other existentialists, indeed, remember we had Skye Cleary on the podcast talking about existentialism, so that is something I think is interesting. I have mixed feelings about it. I think Camus is my favorite. He was the best writer I think, among them. Simone de Beauvoir, obviously hugely important for feminist philosophy, and Sartre and also Kierkegaard, as well as proto-existentialists like Dostoevsky and so forth, all have something interesting to say. I do wonder about the emphasis on authenticity that the later existentialists had, Sartre and Beauvoir especially.
2:58:26.5 Sean Carroll: I'm not so sure about authenticity as one's lodestar in these circumstances. I think that maybe we need to accept the more theatrical and dramatic aspects of human life. But I think it's important, to think through these issues. I kind of am sad that these days Western philosophy has less emphasis on these day-to-day living questions than it used to have back in the heyday of existentialism.
2:58:56.3 Sean Carroll: Jesse Rimler says, "Roughly speaking, the progressive left in the United States splits into two major camps. First, accountable capitalists who emphasize antitrust and consumer protections inspired by thinkers like Brandeis, and second, social democrats who focus on expanding universal programs via a welfare state and the nationalization of industries and socialization of wealth, inspired by what exists around the world, especially the Nordic models. Interestingly, Nordic countries like Norway and Finland have highly prosperous, more egalitarian societies and more sector concentration due to state ownership. What do you make of these approaches?"
2:59:32.3 Sean Carroll: I think that there's a lot of different possible... I think you shouldn't be wedded ahead of time to a model. I think you should try things out and see how they work in the world. You should be an empiricist about these things. I am actually a fan of market forces in the right circumstances. We've talked about markets a lot on the podcast. If you go through it, Sam Bowles, Brian Arthur, Doyne Farmer, Daron Acemoglu, Henry Farrell, we had lots of talk about markets in different ways. And I think everyone who's a professional economist thinks that markets are good for some purposes.
3:00:12.2 Sean Carroll: The question is, how much you constrain the markets and how much you put protections on there because markets absolutely can be bad in some cases. So I'm not really in favor of nationalizing industries unless those industries are specifically conceptualized as social goods, communal goods, as we were talking about before, the job of government. I think that, I don't want to privatize the railroads like they did in England. I don't wanna privatize national defense or the roads or education or anything like that. I think that these are things that serve a much broader communal purpose than an individual purpose, and so it's perfectly sensible for government to do them. I would like more nationalization of health care and elderly care. I think that those are two things that here in the United States we do very, very badly at.
3:01:05.3 Sean Carroll: But I don't want nationalization of like the steel industry or the communications industry. I think that here are cases where a little competition is a very, very good thing. I don't wanna privatize the post office, but I do wanna keep most media outlets perfectly private. It's okay to have like a public broadcast system. That's a good and important thing. But mostly, I wanna let things be private, and let the market decide what is good or bad, as long as there are protections for people who are not being well served by the market forces.
3:01:40.8 Sean Carroll: Jeff H. Says, "If you grant our universe is simulatable in principle, would we be none the wiser if it were it, like a record, rewound and replayed, paused for a trillion years, or played in reverse? Need it play at all? What of that other album, self-defining, internally consistent universe volume 2, gathering dust on the shelf, never to be placed atop the turntable? Should we hold a certain prejudice against the inhabitants of that vinyl?" I don't know. [laughter] On the one hand, yeah, a simulated universe is just as real as anything else. But again, you're asking a question, and I'm not poo-poo-ing the question, I'm just saying I don't know. You're asking a question that is super far outside of our experience. So it's very, very hard to really have confidence that your answer to it is on solid ground.
3:02:29.2 Sean Carroll: If our universe is being simulated outside, but there is no observable impact of that on our universe, then I don't see what point it is to be thinking about it. As long as we are being run on the simulation and things are still going on as they do, then we get to exist and treat it as we ever would. If there is some higher level intelligence that gets to pause the simulation and restart it later, like we would do with a video game, and we don't notice, then I don't think it matters to us here, as long as we never notice.
3:03:03.5 Sean Carroll: Mark says, "Are there lessons from the controversy surrounding MIT's de-publishing of the paper Artificial Intelligence, Scientific Discovery, and Product Innovation beyond a cautionary tale? Scientists, whether top or lower third, are not immune to the foibles of human nature and are susceptible to of-design AI sycophancy, perhaps particularly vulnerable given their intense focus on their subjects and their hard-earned pride, characteristics that are not necessarily compatible with self-awareness. I worry that when a top scientist is pleased with their interaction with AI, they might miss that they are looking at a disguised mirror of what they themselves have created."
3:03:38.9 Sean Carroll: So for those who don't know, a graduate student at MIT in economics, submitted a paper that was, it seemed to be an amazing result. It showed that the use of AI in various materials science laboratories greatly enhanced the productivity and creativity of those laboratories. Turns out he fabricated the entire thing. He fabricated the entire thing, ironically enough, using AI. So in fact, he lowered the overall productivity of the world, by making people worry about this. I haven't been following it closely enough maybe, but I don't think it was de-published. I don't think that's quite the right way to say it. The paper was submitted to a journal. It had not gotten to the point where the journal had judged it yet, before they realized that it was falsified and they withdrew it. Some economists had read the paper, 'cause you read papers before they're refereed. That happens all the time. That's fine. And they were very excited about it, which you should be excited about it if you didn't know it was fake.
3:04:41.5 Sean Carroll: It was a very exciting result. Apparently, there were some claims that you should have known it was fake because the results were too good to be true, but sometimes that's what the community is there for. The community judges these things and someone is relevantly knowledgeable enough to say, oh, this must be fake, and then you look into it more carefully and realize that it is. I found a paper that I mentioned on Blue Sky, that was about cosmology. I found it on the archive, and it was very interesting. It was literally about Boltzmann brains and the anthropic principle and other things I know something about, and I'd never heard of the author, and there was no institution listed, and eventually I realized the whole paper was AI-generated.
3:05:20.5 Sean Carroll: No harm, no foul, because who cares about these papers other than people like me, but it's gonna be a problem going forward. The paper I found was not attributed to an existing researcher. So it's not like someone was trying to build up their CV or anything like that. I have no idea why it was done honestly, and I have no idea how many other papers like that there are.
3:05:41.3 Sean Carroll: It does speak to why it's so difficult for people on the street to 'Do their own research.' Because, at first glance, the paper looked completely sensible to an expert like myself, but then, at second or third glance, it looked suspicious. But you have to really be trained to know that it looks suspicious. So I certainly worry that it's not a problem with scientists per se or their self awareness or anything like that. I think that fabrications by AI are gonna become more and more convincing as time goes on.
3:06:13.7 Sean Carroll: The paper that I saw, it would not have been that hard to tweak it so that it was completely unrecognizable as a fake. It wasn't a great paper. It wouldn't have made a big splash or anything like that. The paper from the MIT student got a lot of attention because had it been correct, it would have been very, very important. This one that I found was completely ignorable. [laughter] But we're gonna be faced with all that, and I don't think it's a scientist problem. I think it's a world problem. Separating out the wheat from the chaff, the true from the fake, is gonna become a harder and harder thing to do.
3:06:46.8 Sean Carroll: Pseudonym says, "Do you have a personal line that if a US Administration were to cross it, you would either leave the country or make fighting to reverse it your primary focus, even to the detriment of your career?" No, I really don't. I don't believe in such lines because I think that one's judgment can change over time. Again, as I've said before, I am very unlikely to ever leave the United States for those reasons, but it's always a possibility. I'm more a stay-and-fight kind of guy, even in the meager ways that I can be counted as fighting to some extent. So we'll have to see. Things could get very bad. If there's literally imminent physical danger to my family, then I would absolutely consider it, or if I become homeless or whatever, or careerless. But up until then, I would rather be here and at least raise a ruckus, if not actually literally fighting.
3:07:40.9 Sean Carroll: Johnny says, "I'm trying to understand the Boltzmann brain thought experiment. Is there any insightful conclusion from it that we can put our finger on?" Yeah, I think the conclusion is very, very straightforward. You're imagining being a cosmological theorist. You're comparing the likelihood of different cosmological models. If your model predicts the overwhelming abundance of Boltzmann brains, then your model is no good, and you throw away that model and try to develop a better model. That's it. That's the end of it. How you get to that conclusion is tricky, and people disagree, but I think that's the right conclusion to draw.
3:08:18.5 Sean Carroll: Nanu says, "If supersymmetry does not explain why our universe didn't decay to a lower energy-level universe, what would, from your perspective?" I'm not sure exactly why you're invoking supersymmetry here. There is a relationship between supersymmetry and vacuum energy, namely that in straightforward implementations of supersymmetry, the vacuum energy has to be either zero or negative. So if the vacuum energy is positive, as it appears maybe to be in our real universe, that certainly means that supersymmetry has to be broken. Maybe it means that supersymmetry isn't allowed at all. This is something that is actually still being debated among the people who think about these things.
3:09:02.1 Sean Carroll: But without supersymmetry, it's still perfectly possible that our universe is in it's lowest energy state, or at least the vacuum of our universe is in it's lowest energy state. You just have a lot more freedom in that case. Supersymmetry, because it's a symmetry that mixes internal symmetries of particles with spacetime symmetries, has something to say about what the vacuum energy is allowed to be, whereas most symmetries just don't. So if you don't have any supersymmetry in your theory, then generally, the vacuum state can have whatever energy you want, and you can make it perfectly compatible with the universe as we know it.
3:09:39.2 Sean Carroll: Jonathan Goodson says, "I heard you opine that within 50 years, many worlds will emerge as the dominant interpretation of quantum mechanics. Yet over the past 50 years, the number of different interpretations has increased significantly and no major contender has yet been ruled out. As N. David Mermin remarked, new interpretations appear every year, none ever disappear. What would need to occur for many worlds to buck that trend and emerge as the clear winner? And why did you predict the resolution will occur in this century?" I might be wrong in my prediction, but that is my prediction, and I think that these processes are gradual. I don't think that there's any threshold you cross and suddenly everyone says, oh yes, it must be many worlds.
3:10:17.4 Sean Carroll: I think that what will happen is, there is now, here in 2025, a lot more attention being paid to quantum foundations than there was 50 years ago, so that will lead to more progress. And the way that progress will happen is, people will think deeply about the different models and their implications, and they will both look for experimental tests of them, and they will ask how those models fit in with other things we think are true in physics, whether it's quantum gravity or particle physics or atomic physics or whatever. And certain things will fit better and better, certain things will fit worse and worse, and I do believe that the progress will be made. Maybe it's wrong that it'll be 50 years, maybe it'll be 200 years, I don't know for sure, but at the rate at which I see progress being made, I'm optimistic that it'll be 50 years.
3:11:06.7 Sean Carroll: Peter Lloyd asks a priority question, which is, "Why do you act so dumb whenever you discuss consciousness?" So, again, I will deny the premise of the question. I do not think I act dumb when I discuss consciousness, and I think that if you think that I am not actually dumb usually, but you perceive me acting dumb when I'm discussing consciousness, I think your first guess is that there is at least some reason for me to be having the opinions I do that maybe you have not quite understood. That's not to say you have to agree with me, but I think that I try very hard when smart people are saying things I don't agree with, to try to understand why they would say those things. That doesn't mean I'm gonna start agreeing with them, but understanding why they say those things is very important. I try to be perfectly clear about why I have the opinions I have about consciousness. You're welcome to disagree with them. So I don't think I'm dumb. Maybe I'm not clear. That's always certainly possible.
3:12:02.1 Sean Carroll: Ed Sedstuff says, "In your interview with Eric Topol, you expressed skepticism about AGI, which surprised me given your physicalist worldview. From a physicalist perspective, human intelligence emerges from physical processes, so in principle, shouldn't similar processes be replicable in machines?" Yes, 100% they should. I'm sorry I don't exactly remember what I said in that particular interview, but I try very hard to say the same thing over and over again. Somehow it doesn't always come across.
3:12:32.9 Sean Carroll: I am skeptical about large language models achieving AGI. I am also skeptical about using the concept of AGI as a useful benchmark for progress in artificial intelligence more broadly, but I will say over and over again that there are zero barriers in principle to some kind of computational architecture doing everything that human beings can do. Similar processes should absolutely be replicable in machines.
3:13:04.4 Sean Carroll: David Khodaverdian says, "Could you please explain, in the example of a hydrogen atom emitting photons, why we observe photons with specific energies? Is it because the electron's wave function is in states with definite energies, or do we as observers collapse the wave function into an energy eigenstate? What is so special about the energy operator that we so often speak of its eigenstates and observe them in experiments?" Well, we speak of lots of things. Energy eigenstates are one of the things to talk about. It's the last part of your question that really puts the finger on it. Why do we observe them in experiments?
3:13:39.5 Sean Carroll: The point about quantum mechanics is that, there are things that we call observables. We can observe position, momentum, energy, whatever. They are incompatible in general, so that if you observe position, then you have not observed momentum, as we were talking about before. And we need to make a choice about which of these observables we're going to observe. And what that choice comes down to, is a choice of the physical apparatus doing the observing. Okay, so what is the physical process by which we made that measurement? For a oscillation in the electromagnetic field, we are often observing it by having that oscillation be absorbed by some process with a discrete energy spectrum. Sorry, I said some process, by some system, by some physical object with a discrete energy spectrum itself. And so, the actual interaction, you have to go into the details, it's not a simple thing. I'm sort of talking words around it, but there's math behind this.
3:14:40.1 Sean Carroll: The actual interaction between the electromagnetic quantum wave and the physical system, entangles states of definite energy in the photon with states of energy, definite energy in the system, and we say that we have observed the particular energy. It is also true that in the case of emitting atoms, it is very often the case that due to interactions with the outside world, the atom that we're talking about starts out in a state that is close to a state of definite energy. Probably not exactly there, but it's close to there, and therefore when it undergoes a transition, it undergoes a transition of definite energy.
3:15:20.0 Sean Carroll: It will undergo a transition gradually, so usually that transition is a superposition. If you catch it in between the start and the end, it is typically in a superposition of the different energies, but then by making the observation, you collapse it, or if you want, you branch the wave function of the universe into one where the transition either has happened or hasn't happened. But anyway, long-winded answer to say that, it's because that is the measurement that we choose to do by choosing our physical measurement apparatus to be sensitive to certain properties.
3:15:55.2 Sean Carroll: Julian Voigt says, "If the quantum wave function is not a function of space, how is it that the atom or anything really is not empty but rather filled with or by the wave function?" Excellent question, because this points to the kind of thing we do all the time in quantum mechanics or in explaining physics more broadly is that we cheat. We simplify a little bit. The quantum wave function in general, is not a function of space. The reason why it's not a function of space is because of entanglement. If I have two particles, then the quantum wave function is a function of the position of both of those particles. It's a separate function of x1 and x2. If x1 is the position of particle 1, and x2 is the position of particle 2. So that's not a function of space. That's a function of space two times, which is a different thing than space.
3:16:42.7 Sean Carroll: However, for many practical purposes in the case of an atom or something like that, there is not a lot of entanglement or at least the entanglement is there, but it's so subtle you don't need to worry about it because of identical particles and things like that. But for all intents and purposes, a particular electron in a particular atom can be said to have an unentangled wave function all by itself. Then, when the wave function of a single particle is unentangled with anything else, then the wave function is just a function of space. So we use that simplification to allow ourselves to visualize the shape of the wave function of a certain electron. If you push hard on that concept, you'll have to admit it's not completely accurate.
3:17:28.7 Sean Carroll: Jeff B. Says, "Last month I asked something about measuring field configurations and you said something like, in practice we don't measure the field configuration everywhere at once, nor do we measure precisely at one point. But I'm still wrapping my head around what this means in a real world physical experiment, and how I should imagine fields interacting with a cloud chamber to give the impression of particles. In ordinary quantum mechanics, this makes sense to me, since the particle has a wave function of position, but it no longer makes sense to me if the particle really is smeared out in a field." Well, I would advocate doing, I should have grouped this with the earlier question, I think that the right way to get conceptual clarity on this, is to recognize that the wave function of a field, when you are specifically interested in relatively low energy quantum states, looks exactly like a wave function of a set of particles. That is how particles appear out of quantum field theory, whether those particles are quarks or electrons or photons or gluons or whatever. So it is 100% okay, to think about the wave function as depending on the positions or momenta, whatever, of a set of particles.
3:18:43.5 Sean Carroll: That's okay. So when you do the experiment, you can ask, what is the wave function of the set of particles? Maybe it's a superposition of different numbers of particles, that is something that the field handles perfectly well, but this is the miracle of what is called Fock space, and I talk about it in Quanta and Fields. That the set of quantum states of fields, as long as they're not too energetic, looks like a superposition of zero particle states, plus one particle states, plus two particle states, et cetera. So if you can train your brain to think that way, you'll have a much easier time understanding this kind of question.
3:19:21.1 Sean Carroll: Finally, I'm gonna group two questions together for the final question this month. P. Walder says, "In principle, do you think that advances in neuroscience accompanied by technical advances such as neural link, will eventually lead to thoughts being brought into the public domain and therefore moving thinking from subjective to the objective domain? And Lukasz Honnold says, "Given recent advances in brain-computer interfaces, do you think that in 10, 20, or 50 years we'll be sharing our thoughts directly with our friends, the government, or an AI system, and what's your take on such scenarios? I've always considered this to be a dystopian idea, but given human curiosity, I doubt they could resist the temptation." I think these are super good questions. I do recommend that you go back and listen to the podcast episode I did with Nita Farahany.
3:20:07.0 Sean Carroll: She is a lawyer/philosopher, but she has been looking into digital privacy in the age of brain-computer interfaces. Should our innermost thoughts be private? The good news is, that exactly the thoughts that are the most innermost and the ones that maybe you don't want other people to know about, are the ones that are hardest to access via brain-computer interfaces. That's a little bit of silver lining on the cloud. But I think that both of these questions are leaning in the right direction of being worried. Because on the one hand, there's enormous possibility of useful, fun, provocative new technologies out of brain-computer interfaces.
3:20:54.5 Sean Carroll: On the other hand, there's enormous prospects for abuse. And human beings, just like with AI, we very often vote for convenience and usability over safety and privacy. So it might be a thing, like if you go back to the podcast with Daron Acemoglu, where he was thinking about literally the economic impacts of new technologies. Now we're talking about the privacy impacts of new technologies. But I think that the analysis is a little bit parallel, a little bit similar, namely that the initial effects can be really bad, really disastrous, like workhouses and terrible factories at the Industrial Revolution.
3:21:38.7 Sean Carroll: But then, the public fights back a little bit. They organize, and they make unions and whatever, and they put in protection for workers, and things get better. So yeah, I think that that's the kind of thing that will happen. That's my guess. My guess is, we will have some terrible, terrible things happening with brain-computer interfaces, as well as some wonderful, wonderful things. And it will take time to sort out how to put protections on things to keep our innermost thoughts where we want them. So that's actually an optimistic take. The optimistic take is we will eventually sort it out, and we'll keep our thoughts private, and I think that's good. There is a dystopian version, which I don't like to think about, but hopefully we're at least having the conversation enough, maybe we are, maybe we're not, to be aware of what the challenges are and the worries about them, and hopefully we take care of them.
3:22:28.7 Sean Carroll: All right, that's it for this month. Thanks for sticking with me. With this month's AMA, thanks for supporting the Mindscape Podcast. Wish me luck, at the upcoming Natural Philosophy Symposium. Very excited about that. I'll report back on how it went later on. Talk to you next time. Bye-bye.
[music]
The Piers Morgan/Eric Weinstein conversation was sickening. He is just a petty, unhinged selfish person. The less said about it the better.
Pingback: El tiempo más largo del universo – Mundos Múltiples