55 минут, которые ПЕРЕВЕРНУТ Ваше представление о РЕАЛЬНОСТИ - Алексей Семихатов — Transcript
Full transcript
- 0:00Does a three-dimensional world even
- 0:02exist if it can all be placed on a
- 0:04surface? Perhaps I live on a surface,
- 0:06and here I only seem to exist, seem to.
- 0:09Can we then say that we are living in a
- 0:10simulation? We don't pay attention to
- 0:13this, but in fact, people didn't
- 0:15understand it right away.
- 0:17Well, it's counterintuitive.
- 0:18It is, to put it mildly,
- 0:20counterintuitive because of, it seems,
- 0:22this possible trickery. An amazing
- 0:25thing happens. thing. Information as an
- 0:27abstract phenomenon suddenly turns out
- 0:29to be something quite physical. Then
- 0:31the question arises: where is the
- 0:33universe located that is running a
- 0:34computer of such power? The first
- 0:36reality, let's put it that way.
- 0:41Many people in Silicon Valley have
- 0:43become obsessed with the simulation
- 0:45hypothesis, the claim that what we
- 0:47perceive as reality is actually created
- 0:49by a computer. The New Yorker claims
- 0:53that two tech billionaires have gone so
- 0:55far as to secretly fund scientists to
- 0:57work on ways to break us out of the
- 0:59simulation. And before we draw any
- 1:04conclusions on this matter, I would
- 1:06like to discuss the holographic
- 1:08principle, what it means, and what
- 1:09conclusions we can draw from it.
- 1:12So, we're setting the simulation aside.
- 1:14Setting it aside, yes.
- 1:16Well, it's just that in the simulation,
- 1:17your main word was "secret." As soon as
- 1:19words about secrets appear in any
- 1:21information, well, that's it,
- 1:23conspiracy theory; plus there's the
- 1:25word "billionaires" and the word "
- 1:26secret." No, well, of course, of course
- 1:28, it's discussed in society that
- 1:30billionaires are doing something
- 1:32secretly behind our backs. Well,
- 1:33well, yeah, what else would you expect
- 1:35from them?
- 1:36It excites the mind.
- 1:37Yes, yes, yes, of course, of course,
- 1:39it's discussed.
- 1:40And, accordingly, the holographic
- 1:42principle, what is it?
- 1:44A silly story. Very impressive. A very
- 1:48impressive, silly story. What is
- 1:50holography? What seems to you to be a
- 1:543D figure is created because you are
- 1:56illuminating a certain surface with
- 1:58coherent light. The light reflected
- 2:01from it behaves in such a way that a 3D
- 2:03figure seems to appear before you in
- 2:05space. At the same time, if you remove
- 2:09some piece of this surface on which the
- 2:11hologram is recorded, the figure won't
- 2:14disappear, it will just become lower
- 2:16quality. This means that the properties
- 2:20of this 3D object, which you recorded
- 2:22as a hologram on a special plate, are
- 2:24distributed, they are very widely
- 2:26smeared across the entire plate. You
- 2:29can't say that if this is a holographic
- 2:32Obi-Wan Kenobi, then Obi-Wan's head is
- 2:34here, his hands are here, and his
- 2:36lightsaber is recorded over there. No,
- 2:38no, no, it's all, it's all smeared in a
- 2:41very complex way. And in a way that is
- 2:43very difficult to explain.
- 2:44It is impossible to say how the
- 2:46information was redistributed across
- 2:48this holographic plate. That is
- 2:50ordinary holography. Now, what is, uh,
- 2:52cosmological or whatever it's called,
- 2:55this kind of holography? The
- 2:56holographic principle of a black hole.
- 2:58In the context of BHs, you haven't
- 3:00mentioned a black hole yet; you stole
- 3:01that from me. Yes, yes. A black hole,
- 3:04of course, is relevant, yes. This story
- 3:06begins with the following observation.
- 3:08A black hole is the result of gravity
- 3:10winning over matter, when a body
- 3:12disappears and empty space remains. At
- 3:14the center, there is something. Uh, we
- 3:16don't know what—we call it by the
- 3:17word "singularity" and don't understand
- 3:19what's there. But everywhere except for
- 3:21the center itself, it is empty.
- 3:22Therefore, a black hole has no surface.
- 3:24Uh, we can, of course, draw it—well,
- 3:26this is the case where visual things
- 3:28and what seems intuitively clear to us
- 3:30—we, of course, how to draw a black
- 3:32hole, well, I take and draw it in black
- 3:34. Aha.
- 3:35There. Well, uh, they always show it as
- 3:37a black sphere. No, it is not a sphere.
- 3:40It is a region of space that has
- 3:42nothing—nothing resembling a solid
- 3:44surface. So what does it have?
- 3:46Uh-huh.
- 3:47It has what is called the horizon
- 3:48surface. This is a surface
- 3:50mathematically drawn in space that
- 3:52possesses the following property. If
- 3:54you are beneath it, then even if you
- 3:56are light, you can no longer return. It
- 3:59is the point—the boundary of no
- 4:01return, inside of which gravity becomes
- 4:03too strong. And your fate, whoever you
- 4:06are, even light, the fastest thing
- 4:08there is, is to fall into the center.
- 4:11Something will happen to you there; we
- 4:12don't know what. Look at what happens.
- 4:14You created a black hole by throwing a
- 4:16huge amount of things into it. So, you
- 4:18got tired of sitting in the studio and
- 4:20threw everything here into a black hole
- 4:21. Then you got angry at something, came
- 4:24home,
- 4:24threw all your children's toys into the
- 4:26black hole, and just threw everything
- 4:28in there. Uh, the things you throw in
- 4:30carry with them, in a sense, a large
- 4:32amount of information about themselves.
- 4:34They are made of something.
- 4:36And this information, in principle, is
- 4:38preserved. It seems to us that when I
- 4:40burn something, the information is
- 4:42destroyed. But, actually speaking, in
- 4:45principle, I have the possibility to
- 4:47track pieces of ash, bits of smoke, or
- 4:50something else. The information is
- 4:54terribly scattered. Terribly scattered,
- 4:57but, in principle, it is preserved. A
- 4:59black hole is structured in a peculiar
- 5:01way. It, uh, shows nothing on the
- 5:05outside—the information about what
- 5:07you threw in there, it hides it. In a
- 5:12way unknown to us. But the amount of
- 5:15information you've thrown in there is
- 5:17proportional to the surface area of the
- 5:19horizon. Once again, if you’ve thrown
- 5:23something into a black hole, it becomes
- 5:25a little bit—how does this happen?
- 5:28How so? Look, there is no volume. How
- 5:30does it happen from the perspective of
- 5:32gravity? You’ve thrown something into
- 5:35it, it has become more massive, its
- 5:37gravity has become stronger, and the
- 5:39greater the mass of the black hole, the
- 5:41larger its horizon radius. I repeat,
- 5:43this is an imaginary surface that
- 5:46simply denotes the point of no return.
- 5:48But, uh, when you’ve fed it, it has
- 5:51eaten, your black hole, its horizon has
- 5:53become larger. You might think that the
- 5:56information about it having eaten is
- 5:57proportional to the volume inside the
- 5:59horizon. Well, after all, for us, a
- 6:01black hole is what’s inside—in
- 6:03general, of course, it’s everything
- 6:05together, but, well, the hole, the hole
- 6:07is exactly what’s inside the horizon,
- 6:09right? It’s what is bounded by the
- 6:10horizon. Well, nothing of the sort.
- 6:12It’s not like that with my closet.
- 6:15When I put things into my closet, I
- 6:17stuff something in there. And if my
- 6:21closet suddenly became twice as long,
- 6:23twice as wide, and twice as high,
- 6:27then the volume would increase
- 6:28eightfold. I could stuff eight times
- 6:30more junk in there, right?
- 6:32But with a black hole, the situation is
- 6:34different. It means, uh, it's as if
- 6:37every time I put sneakers into the
- 6:39closet, I took a standard label, wrote
- 6:42"sneakers" on it in a small font, and
- 6:44stuck it onto the walls of my closet.
- 6:48Well, let’s take a bold step. Let’s
- 6:52look around, go out on a starry night,
- 6:55maybe to some observatory, look at the
- 6:57surrounding universe, and all the
- 6:59information contained within it is
- 7:01placed on an imaginary boundary. Its
- 7:04imaginability refers to the fact that
- 7:05it is the boundary of the observable
- 7:07universe. It is the distance from which
- 7:09light has managed to reach us. In a
- 7:10second, it will reach us from somewhere
- 7:12slightly further away, one light-second
- 7:14further. But whether there is something
- 7:16there, or there was something there.
- 7:17Well, I have just fixed it completely
- 7:19for the New Year. So, I somehow drew
- 7:21the boundary of the observable universe
- 7:24and was talking about these volumes.
- 7:26Everything that is located in this
- 7:27volume,
- 7:28it is a fairly large volume. Basically,
- 7:30everything we see and are capable of
- 7:32seeing with all our Hubbles, James
- 7:34Webbs,
- 7:34gravitational waves in the future, and
- 7:37so on. All this information about
- 7:40everything that happens here can, in
- 7:42principle, be placed on the surface of
- 7:44this volume. This is very strange.
- 7:47Because it seems that volume is even
- 7:49more capacious. all the information,
- 7:50everything that happens, absolutely
- 7:52everything, about everything, about the
- 7:54entire visible universe can be placed
- 7:55on the boundary of this visible, well,
- 7:57observable universe, according to
- 7:59three-dimensional onto two-dimensional,
- 8:00yes, and that is very strange. If you
- 8:03are an impressionable person, you start
- 8:05to think: "Where do I actually live?
- 8:07And what was that stuff the guys,
- 8:09especially the billionaires, said about
- 8:11the simulation? So, does a
- 8:13three-dimensional world even exist if
- 8:15it can all be recorded, placed on a
- 8:17surface? Maybe I live on the surface,
- 8:20and here I only seem to myself to exist
- 8:23. I am exaggerating this greatly,
- 8:26because why does the volume turn out to
- 8:28be filled only as much as the surface
- 8:30allows due to gravity? because the
- 8:34limitation of this process—stuffing
- 8:37something into a volume—was set by
- 8:39gravity: if you overdo it—bang, a
- 8:41black hole. and a black hole sets a
- 8:44limit on how much a volume can be
- 8:46filled. So, in that ridiculous
- 8:49description, there is no gravity. It
- 8:51exists in a space with one dimension
- 8:54less. And what's worse, there is
- 8:58nothing at all in that theory, which
- 9:00rewrites information on the boundary,
- 9:02information about our volumetric world,
- 9:04that resembles it. It has nothing
- 9:10similar to our theory, to our, to our
- 9:12familiar world. This is rewriting like
- 9:15holography. That is, it is very
- 9:18non-local. Some pieces are smeared over
- 9:21significant parts, over significant
- 9:23parts of this very, this very surface.
- 9:26Information about something located
- 9:27here is smeared across it, and about
- 9:29something located here, it is also
- 9:31smeared, superimposed on that, and so
- 9:33on. This is a big story, it is not
- 9:35fully understood, it is interesting
- 9:37from a fundamental point of view. And
- 9:39all this comes to an end. I must now,
- 9:42with all my characteristic directness,
- 9:44conclude with a pessimistic statement.
- 9:47Is this all applicable to our Universe?
- 9:51Yes, it is. It is. It is applicable to
- 9:54an anti-de Sitter universe with a
- 9:55negative cosmological constant. In our
- 9:58Universe, the cosmological constant is
- 10:00positive.
- 10:01This is the one who argued with
- 10:02Einstein.
- 10:03That's right. This originated from what
- 10:05came from Einstein originally. This is
- 10:07what Einstein introduced into his
- 10:08equations, considering it his biggest
- 10:10mistake, because he introduced it
- 10:12guided by incorrect ideas and continued
- 10:14to believe until the end of his life
- 10:16that he shouldn't have done it. And
- 10:18everyone thought so, but then it turned
- 10:20out that we do live in a Universe where
- 10:22it exists. And this thing, by the way,
- 10:24is responsible for the accelerated
- 10:26expansion of the universe. Not for the
- 10:28expansion of the Universe as such, but
- 10:30for the fact that over time its rate,
- 10:32its rate increases. Here is how people
- 10:35relate to this. People relate to this
- 10:37as follows. So, this is a model story
- 10:39that teaches us a great deal. And these
- 10:43models are not meaningless. Well,
- 10:45perhaps when we relax something local,
- 10:47perhaps due to the physical processes
- 10:49occurring there, they might not care
- 10:51whether they are in a de Sitter or
- 10:52anti-de Sitter universe. The mechanism
- 10:54of how we can map something onto a
- 10:56boundary within the framework of a
- 10:57completely different theory. And that
- 10:59new, completely different theory, which
- 11:01describes what is happening on the
- 11:02boundary, may allow us to make progress
- 11:04in calculations, in understanding where
- 11:06here, in our three-dimensional world,
- 11:08we are hitting a wall. That is,
- 11:09probably, of course, very useful, very
- 11:12great. And maybe this idea is a very
- 11:13general one, and therefore studying it
- 11:15is not meaningless. Maybe someone will
- 11:17eventually come up with a way to do it.
- 11:19And then, of course, the experience we
- 11:21gained in the anti-de Sitter space will
- 11:23prove useful. And perhaps it will turn
- 11:25out that we can simply transfer the
- 11:26results from there to here. I don't
- 11:28know. I want to reiterate: is the
- 11:30statement true that the information
- 11:32that fell into a black hole is encoded
- 11:34in bits on its surface?
- 11:37In bits on its surface, yes. So, one
- 11:39bit is one per every Planck area,
- 11:42that’s 10 to the power of, don’t be
- 11:44alarmed, minus sixty-six square
- 11:46centimeters, roughly that much.
- 11:48Well, is that not a hypothesis, is it
- 11:50confirmed?
- 11:51No, it’s not a hypothesis. It’s not
- 11:54a hypothesis. It is our, wait, what it
- 11:56is, is our understanding of gravity
- 11:59within the framework of general
- 12:01relativity and, uh, such minor ventures
- 12:07. Well, let’s stick to general
- 12:09relativity, yes, it is our best theory
- 12:11of gravity. Uh, perhaps it is not
- 12:13complete, because it is not quantum, it
- 12:15needs to be made quantum, and it is so
- 12:17reluctant to become quantum that some
- 12:19people already doubt whether that is
- 12:21even necessary. That I don't know, that
- 12:23is a separate conversation.
- 12:25Nevertheless, it is our best
- 12:26understanding of gravity. It describes
- 12:27black holes. It describes the expansion
- 12:29of the universe. It describes the
- 12:31deflection of light, which is
- 12:32measurable, describes the precession of
- 12:34the perihelion of Mercury, which is
- 12:35experimentally verified, and so on. But
- 12:37it is not complete, that is, it is not
- 12:39quantum and it refutes itself in
- 12:41singularities. So this means that, in
- 12:43principle, it might someday, most
- 12:45likely not, we just won't get that far.
- 12:47Well, perhaps it will someday be
- 12:49replaced by a more advanced theory, and
- 12:50perhaps that more advanced theory will
- 12:52correct this statement. But, um, when
- 12:54you move from a, generally speaking,
- 12:57working theory to one that corrects,
- 12:59improves, or generalizes it, you don't
- 13:01cross out everything that came before.
- 13:04In those areas where the old one worked
- 13:06well, the new one must also reproduce
- 13:08the same results. Exactly like that,
- 13:10perhaps even slightly more accurately,
- 13:12yes, sometimes that precision isn't
- 13:14needed, but sometimes it is. Well, this
- 13:16is exactly the relationship with
- 13:17Newtonian theory of gravity. We launch
- 13:20satellites, probes, track asteroids,
- 13:23and all that, following Newtonian
- 13:25gravity. Einstein's theory refines it.
- 13:29For launching a satellite to the Moon
- 13:31or even landing a person on the Moon,
- 13:33we don't need the General Theory of
- 13:34Relativity. But for an accurate
- 13:37description of the planet Mercury's
- 13:38motion, it is necessary. That is,
- 13:40Newton describes it, but not quite
- 13:42accurately. But this doesn't mean you
- 13:43can't use Newtonian theory. Therefore,
- 13:45with the General Theory of Relativity,
- 13:47eventually, perhaps the same thing will
- 13:49happen. Some other theory will emerge,
- 13:52but this one will remain in the areas
- 13:53where it currently works well. And
- 13:56within its framework, the picture is as
- 13:58follows: you can, to have some visual
- 14:00reference, imagine the horizon of a
- 14:02black hole. Again, this is not a
- 14:06surface, it is imaginary, there is
- 14:08nothing there, but it is as if zeros
- 14:10and ones were mentally inscribed upon
- 14:12it. This encodes everything that fell
- 14:16in. Only the zeros and ones, of course,
- 14:18cannot be drawn, because it is
- 14:20something else. It is a cell in which
- 14:22something is effectively recorded. It
- 14:26has a Planck area. I wasn't trying, I
- 14:31certainly didn't say that the black
- 14:33hole horizon is divided into lines or
- 14:35cells, that it looks like an apartment
- 14:37building with very small apartments. No
- 14:41, I didn't say that. This is a
- 14:44conceptual mathematical construct,
- 14:46but these are golden words and this is
- 14:48information. And when talking about
- 14:51information, we can visualize it this
- 14:53way. It is simply convenient for us.
- 14:56Not just a black hole. A black hole, on
- 14:58the surface of which, in principle, one
- 15:00can place information about everything
- 15:02inside it.
- 15:02How, in that case, should we understand
- 15:04the thesis of physicist Susskind that
- 15:06everything around us is a hologram, a
- 15:08two-dimensional surface on the
- 15:09cosmological horizon?
- 15:10No, Susskind understands perfectly well
- 15:12that, strictly speaking, this is not
- 15:14applicable to our Universe. And, well,
- 15:16this holographic duality is a model
- 15:18scheme, not exactly for our universe.
- 15:21And 't Hooft, he said that the
- 15:23three-dimensional world is a hologram
- 15:25encoded on the boundary of the Universe
- 15:27. What is meant by encoding? A real
- 15:30object or a mathematical model? What
- 15:34can we say about this, considering we
- 15:36live in a de Sitter universe?
- 15:39Listen, I don't know. I mean, uh, he
- 15:42should certainly be asked, yes, but the
- 15:45fact remains. We can think that things
- 15:49are arranged in our universe just as
- 15:51they are in the anti-de Sitter model
- 15:53case. But please, please, let's not
- 15:56forget. Perhaps we’ve simply
- 15:58discovered a principle here that likely
- 16:00works always. It’s just that how it
- 16:04specifically, technically works in the
- 16:06case of our de Sitter universe with a
- 16:08positive cosmological constant, we
- 16:10technically do not know at all. Anti-de
- 16:13Sitter must be negative, otherwise,
- 16:15otherwise this boundary ends up in the
- 16:18wrong place, there’s nowhere to draw
- 16:20it, roughly speaking. Well, so that’s
- 16:23an advance, that’s an advance. Uh,
- 16:25maybe that’s just how it is? Maybe
- 16:27the people working on this more closely
- 16:29feel it so well that they have a sense
- 16:31it doesn't depend on some triviality
- 16:33like positive or negative cosmological
- 16:35constants? I don’t know about that, I
- 16:37can’t say, but Susskind himself...uh,
- 16:39from time to time, when he’s really
- 16:41backed into a corner, or when he feels
- 16:43ashamed, perhaps, or wants to finally
- 16:44share something valuable, he says:" No,
- 16:46guys, well, of course, let's be honest,
- 16:48let's just admit it, we live in a de
- 16:50Sitter universe, and in it, in
- 16:51principle, everything we did for the
- 16:53anti-de Sitter one needs to be started
- 16:55over from scratch. "
- 16:56Well, that’s an important addition.
- 16:58Also,
- 16:58that is an important addition. It’s
- 17:00just that the people who work on this
- 17:02truly think that, perhaps, we are just,
- 17:04you know, our computational and
- 17:06observational capabilities are limited,
- 17:08that maybe we just discovered—using
- 17:10this model example, we discovered some
- 17:12general situation, and it isn’t
- 17:14limited to this model example. We just
- 17:16don’t know how to technically apply
- 17:18it to our world. Maybe,
- 17:19it’s quite a pop-culture thing.
- 17:21We already had the pop-culture stuff
- 17:22about simulation.
- 17:23Well yes, yes. And
- 17:24and about billionaire conspiracies. But
- 17:26this is the most
- 17:27which, yes, which by the way is
- 17:28impossible to disprove, yes,
- 17:30like many other conspiracies. Not any
- 17:32absolutely correct. And this statement
- 17:34is in principle impossible to disprove,
- 17:36because, of course, one can assume that
- 17:38yes, but then the question arises:
- 17:40where is that universe located in which
- 17:42a computer of such power is running?
- 17:44The first reality, let’s say, yes?
- 17:46Well yes, where is the previous one?
- 17:48And then, and then the question:" And
- 17:49then what is that one? "" And then you
- 17:51need to somehow explain the status of
- 17:52its existence. But why such [
- 17:54complications] then? If you can explain
- 17:56the status of its existence, why all
- 17:57these complications? Let's just explain
- 17:59instead that we ourselves are not any
- 18:01kind of simulation. I mean, this is
- 18:03such a complication of the world
- 18:05picture that adds nothing to you,
- 18:07absolutely nothing. Because all the
- 18:09questions you might have had, you feel
- 18:11as if they've been resolved. They
- 18:12haven't been resolved at all; they've
- 18:14just been shifted over there, toward
- 18:15those guys who—and I don't even know
- 18:17who they are—are messing around with
- 18:19this simulation, but for unfathomable
- 18:20reasons were able to launch a computer
- 18:22on the scale of an entire functioning
- 18:24universe.
- 18:24But if we, hypothetically, in our
- 18:25reality were able to create a
- 18:27simulation of some world on some
- 18:28computer, would that convince you that
- 18:30...
- 18:30No, no, and we, what? No, we just
- 18:33wouldn't be able to, because, well, how
- 18:35many bits do you need? Uh, how many
- 18:37qubits do you need in that quantum
- 18:39computer? You need a qubit for
- 18:41approximately every Planck area element
- 18:43. Well, that means you need another
- 18:45universe.
- 18:45Well, what if we make it a bit simpler,
- 18:47hypothetically?
- 18:48Then that would be, well, look, in a
- 18:50quantum computer you can make it
- 18:52simpler. For example, yes, we expect
- 18:54that good, powerful ones—not
- 18:55prototypes like we have now, but when
- 18:57quantum computers actually work—we
- 18:59will be able to simulate, for example,
- 19:01the interaction of some drug, or
- 19:03conversely, a poison, depending on your
- 19:05goals, with some protein. It lands on
- 19:06the protein, and we will have exactly
- 19:08that. Yes, that we will most likely be
- 19:10able to simulate. Excuse me, one
- 19:12complex molecule with another very
- 19:15complex molecule. That is, well, quite
- 19:18far from a universe. Well, quite far.
- 19:21Even if the number of these molecules
- 19:24increases a thousandfold, you know,
- 19:26even if a millionfold, even if a
- 19:27trillionfold, which is already hard to
- 19:30believe,
- 19:32that is, uh, you won't be able to
- 19:34describe a cube of matter in all of its
- 19:36details. So to launch, to emulate a
- 19:40universe, you need a number of qubits
- 19:42approximately equal to the amount of
- 19:45information in the universe. And any
- 19:47complex system contains these—you say
- 19:49, I'll take something simpler. Well
- 19:52okay, but even in that, if you
- 19:54understand it in terms of qubits and
- 19:56want to emulate it, that is, run it on
- 19:58a computer, you still need
- 20:01an unimaginable quantum computer with
- 20:03an unimaginable number of qubits and,
- 20:05by the way, with error correction.
- 20:08That’s encouraging, because that
- 20:10error correction means—do you know
- 20:11what error correction means? Error
- 20:14correction means that you need at least
- 20:16several times, if not an order of
- 20:18magnitude more physical qubits, so that
- 20:20your, so that some logical qubits work
- 20:22as if error-free, to fix the errors.
- 20:26Do we have any equations that resemble
- 20:28error correction in the laws of physics
- 20:30?
- 20:32Ah, maybe there are. Maybe there are;
- 20:38in holography, it is possible to
- 20:40discern something—I cannot comment on
- 20:43it in detail—but in the holographic
- 20:45idea, one can discern something similar
- 20:47to error correction. And error
- 20:51correction in a quantum computer is not
- 20:54just comparing two texts with a pen; it
- 20:58is a certain use of quantum
- 21:00entanglement and the redistribution of
- 21:02information across various qubits. No,
- 21:05in this world, you cannot literally
- 21:06duplicate information. There is the
- 21:08no-cloning theorem. The no-cloning
- 21:10theorem, which states that you cannot
- 21:12create a copy of an unknown
- 21:13wavefunction without destroying the
- 21:15original. But you can, in a sense,
- 21:18expand the information by entangling
- 21:21your one qubit, which, like, is just
- 21:23one qubit. Well, it doesn't have too
- 21:26much information, but you can entangle
- 21:28several others with it. And this
- 21:32entanglement is not entirely foreign to
- 21:34how information is redistributed in the
- 21:36holographic principle. That is, it is
- 21:42about how you, where it lands, where it
- 21:44goes, and through what channels the
- 21:46information is distributed. Ah, you see
- 21:51, we are all about information,
- 21:53information about what is happening,
- 21:55about your physical system. So, in this
- 21:59sense, this entanglement, which error
- 22:01correction is based on, can find
- 22:03parallels between it and holography.
- 22:06We are coming back, it turns out, to.
- 22:08Holography again, even.
- 22:09Yes. Well no, holography is a thing. No
- 22:11, holography is not a useless thing. No
- 22:13, not useless. Well, perhaps, perhaps,
- 22:15perhaps the next progress,
- 22:17yes, I’m saying,
- 22:18will come directly from holography.
- 22:20Maybe, maybe. We just don’t know, you
- 22:22understand? Uh, there is no such thing
- 22:23right now.
- 22:24I thought everything in physics had
- 22:25already been discovered, it turns out.
- 22:27On the other hand, you know, some
- 22:28people object, they tell me: "Look, you
- 22:30have moved so far away from the real
- 22:32world in your theoretical objections;
- 22:33you are reasoning about some surface of
- 22:35yours, and not even quite in our
- 22:36universe." Seriously, guys, what are
- 22:38you talking about? What does this have
- 22:40to do with our processes? Maybe at some
- 22:42point someone will break through, find
- 22:44some secret door or break down some
- 22:46wall with a particularly extraordinary
- 22:48device, and then it will turn out...we
- 22:50will just, well, yes, we will enter
- 22:52some new theoretical world with a
- 22:54brand-new understanding of how the
- 22:56world is built, perhaps—that is what
- 22:58people are hoping for. Well, it's
- 23:00impossible to predict when this will
- 23:02happen or if it will happen at all.
- 23:04Don't forget that this situation is
- 23:06different from, for example, the time
- 23:08when quantum mechanics was created.
- 23:10Back then, people's hair stood on end
- 23:12because of what we didn't understand.
- 23:14We knew, roughly, that here is an atom,
- 23:16and something is happening inside it.
- 23:18Our previous theory couldn't describe
- 23:20how the atom was structured. We needed
- 23:22to invent something. They always had
- 23:24the atom right in front of their eyes,
- 23:25you see? This kind of atom, that kind
- 23:27of atom, another kind of atom, and in a
- 23:29magnetic field, yet another. And
- 23:30everyone kept checking against that.
- 23:34But here, this is a theoretical
- 23:36construction we are waiting for, which
- 23:38is based not on studying atomic spectra
- 23:40, but on a previous theoretical
- 23:42construction. So, it is a very indirect
- 23:47thing. Maybe there will be super
- 23:48progress that will return to us here in
- 23:50the physical world. Maybe not. In that
- 23:52case, I'd like to discuss the
- 23:54double-slit experiment. A photon
- 23:56behaves like a particle if we observe
- 23:58it. And we see an interference pattern
- 24:00if we do not observe it. How should we
- 24:02understand this in mathematics and
- 24:04physics? Why is the observer important
- 24:05and why does it change the results of
- 24:07the experiment?
- 24:08Quantum mechanics understands this in
- 24:10only one way. There is a wave function
- 24:14that says if you have two slits, even
- 24:16if you send single photons, they land
- 24:19on the screen behind those two slits as
- 24:21if each of them passed through both
- 24:23slits simultaneously. Which cannot be,
- 24:27because a photon does not divide into
- 24:31parts. Well, that's it; that's what
- 24:34quantum mechanics says. You test the
- 24:36experiment and, in the experiment, you
- 24:38see that this is exactly what happens.
- 24:40It's a bit easier to talk about doing
- 24:42an interference experiment with
- 24:43electrons. It makes absolutely no
- 24:45difference. You send an electron
- 24:47through the slits. It's easy to send an
- 24:49electron. Do you remember 20 th-century
- 24:51televisions? No,
- 24:52well, that thing, that kind of tube.
- 24:54Basically, electrons ran to the screen
- 24:56in there. You just need to release them
- 24:58very weakly. A very weak signal.
- 25:00Release them one by one, one by one, so
- 25:02that the electron goes through. And the
- 25:04two slits are something like
- 25:08a crystal on which you scatter them, or
- 25:10some atoms that limit its movement, and
- 25:16you catch them one by one. But the
- 25:20pattern is as if it passed, each one
- 25:22going in turn passed through two slits
- 25:24simultaneously. And you say: "Guys,
- 25:30I'll figure out now whether the
- 25:31electron comes through the left one,"
- 25:33and you start illuminating it with weak
- 25:36light. But as soon as, and as you start
- 25:40to receive more information, more
- 25:42reliable information about whether it
- 25:44passed through the left or the right,
- 25:46this interference pattern disappears.
- 25:50And it turns out that you found out it
- 25:51passed through the right, but it also
- 25:53arrived where the path through the
- 25:55right slit leads, as if the left one
- 25:57didn't exist. Found out it passed
- 25:59through the left, it arrived where the
- 26:01road, where the path passing through
- 26:03the left slit leads. Now it seems that
- 26:05you are needed here, conditionally you.
- 26:07Yes, an observer.
- 26:09Always in such cases, ask yourself: "Is
- 26:11a PhD really necessary, or maybe a
- 26:13graduate student is enough?" What if
- 26:15it's a kitty? What if it's a kitty?
- 26:18Now this experiment has been done in a
- 26:20different way quite recently,
- 26:22and an earlier experiment is like this:
- 26:25instead of electrons, they send
- 26:27fullerenes. This is a large C60
- 26:29molecule. It's a ball made up of a
- 26:31large number of carbon atoms. There are
- 26:33quite a lot of electrons there,
- 26:35accordingly, these are molecules after
- 26:36all. Unlike an electron, this kind of
- 26:38molecule has, yes, they behave exactly
- 26:40the same way. Well, you just have to
- 26:42define the slit correctly. Uh-huh.
- 26:44You won't be able to cut the slits with
- 26:46a knife to perform this experiment, but
- 26:48it can be done.
- 26:49And unlike an electron, this molecule
- 26:51has an internal life. Here is what you
- 26:53can do with it. You can illuminate it
- 26:55with light, so the electrons there
- 26:57start to flutter, and along the way,
- 26:59while this molecule moves to the two
- 27:01slits and then to the screen, it will
- 27:03emit light, because you first gave it
- 27:05extra energy to these electrons that
- 27:07are there, and they will start to give
- 27:09it off. The radiation they will emit
- 27:11will settle on the walls of your
- 27:13laboratory. No one collects it in a box
- 27:16, no one gathers it, observes it, or
- 27:18sums it up. But that is enough for the
- 27:21interference pattern to disappear. That
- 27:24is, the walls of your laboratory
- 27:26receive information, in principle
- 27:28receive information, and you are not
- 27:30needed, about whether it passed through
- 27:32the right slit or the left. In
- 27:34principle, if you were to collect some
- 27:36finest radiation, removing everything
- 27:39else, it is impossible to do this. You
- 27:42would be able to reconstruct which path
- 27:44it took, the left or the right. Just
- 27:46from the fact that, in principle, you
- 27:47could have reconstructed it, the
- 27:49interference pattern, of course,
- 27:51disappears. It passes either on the
- 27:53left or the right, as it happens, but
- 27:55it disappears. Meaning, you aren't
- 27:57needed; the environment is. The
- 28:01environment, in the form of your room's
- 28:03walls, constantly monitors the system
- 28:05because it receives the radiation that
- 28:07signals where it is located. Quantum
- 28:11mechanics. Feynman once said that, in
- 28:13general, this is a place that cannot be
- 28:15understood. Meaning if you think you
- 28:17understand it, you don't; you don't
- 28:18understand anything. What does that
- 28:20mean? He understood it better than most
- 28:22others, believe me. But the point was,
- 28:24the meaning of that statement is that
- 28:25it's something that doesn't translate
- 28:27into our usual abstractions or our
- 28:29familiar reality; it just doesn't
- 28:31translate, that's all.
- 28:32So we can't grasp it with our own logic
- 28:34?
- 28:34No, we can grasp it with logic. When we
- 28:37watch our language, forgive me, watch
- 28:38everything we say, and act within the
- 28:40framework of quantum mechanics—that
- 28:42is, we look at the wave function,
- 28:44assume nothing, solve the Schrödinger
- 28:46equation, or make an experimental
- 28:48observation that confirms the solution
- 28:50to that Schrödinger equation, and so
- 28:52on. Yes, in that sense, if we don't do
- 28:54anything stupid, there are no problems
- 28:57with logic. But translating that into
- 28:59our familiar abstractions and the
- 29:00behavior of our familiar world—Do you
- 29:02understand what is familiar to us? Some
- 29:05grain of sand, some gear, a drop of
- 29:06water falling, the wind blowing,
- 29:08something shifting a little, pushing
- 29:10another thing a little. Yes, fine, even
- 29:12a magnetic field having a slight
- 29:13influence. And so it moved somewhere.
- 29:15You know, like a Rube Goldberg machine,
- 29:17that kind of thing. Someone pushed
- 29:18someone else, something fell onto a cup
- 29:20, something reconfigured, and so on. We
- 29:22see everything roughly like that, as
- 29:24some sort of little things. But there
- 29:26are no such little things there, that's
- 29:27all. They simply don't exist. They
- 29:29simply don't exist. That's it. They
- 29:30don't exist, period.
- 29:31Uh-huh. And in that case, how does
- 29:33Wheeler's delayed-choice experiment
- 29:35work? What is the participatory
- 29:36anthropic principle, and does it even
- 29:38exist at all?
- 29:39No. So, Wheeler's delayed-choice
- 29:41experiment—great. It doesn't
- 29:43contradict quantum mechanics in any way
- 29:45. And it doesn't suggest, nor does the
- 29:47quantum eraser suggest, that we are
- 29:48changing the past. That is a false
- 29:50interpretation of it. What happens
- 29:52there, I can say very briefly, is what
- 29:54is called post-selection. In the
- 29:59quantum eraser, the quantum eraser, how
- 30:02to put it, the quantum eraser...the
- 30:04idea is supposed to be that we can make
- 30:06a choice about whether we observe the
- 30:08left path or the right path after the
- 30:10electron has already hit the screen.
- 30:16It's not like we're predetermining its
- 30:18past,
- 30:19well, it's not necessarily through
- 30:21consciousness, but we are changing
- 30:22something in the experimental setup. No
- 30:26, there is no paradox here, and there
- 30:29is no traveling back in time. The hype
- 30:32surrounding the quantum eraser is due
- 30:35to the fact that people—those who
- 30:37interpreted it in a hyped-up way—
- 30:38simply forgot that post-selection is
- 30:40involved. You see, based on what has
- 30:44already happened, you need to select,
- 30:46roughly speaking, half of the
- 30:48experiments, and say that the other
- 30:50half doesn't fit the criteria. And that
- 30:53is not called directly influencing the
- 30:55past. You know, if I'm picking out when
- 30:58my team won or lost in a football match
- 31:00, and I select only the instances where
- 31:02they won, it doesn't mean that my
- 31:04choice influenced them to win. No, no,
- 31:08no. It's pretty much the same situation
- 31:11here; it's called post-selection.
- 31:13Please, you shouldn't forget about that
- 31:15in the case of the quantum eraser.
- 31:17Landauer's work. Alexey Mikhailovich,
- 31:20what is that about? It’s a pretty
- 31:21cool thing. To jump ahead, it’s about
- 31:25the thermodynamic cost of information.
- 31:29Since information is everything these
- 31:30days, it’s quite an interesting topic
- 31:32. But it didn't start with Landauer; it
- 31:35started with Maxwell's demon.
- 31:37Uh-huh.
- 31:37We'll have to spend some time on
- 31:39Maxwell's demon. The demon isn't
- 31:41anything bad. It's a guy who deals with
- 31:43information. It's an imaginary being.
- 31:46It's what science calls a thought
- 31:48experiment. When you're not trying to
- 31:51build an actual device, but you're
- 31:53interested in the fundamental
- 31:55possibility, and that's how you test
- 31:56physical laws. Now, the law of
- 32:00increasing entropy determines the arrow
- 32:03of time. And it says that if you have a
- 32:07hot body and a cold body—for example,
- 32:09hot water here—after some time, the
- 32:11temperatures will equalize. The
- 32:14explanation for this is roughly as
- 32:16follows:
- 32:18There are more ways—many more ways—
- 32:20to realize a state where temperatures
- 32:22are equalized than to realize a state
- 32:24where it is hot here and cold outside.
- 32:29And time flows toward the direction
- 32:31where there are more states, where
- 32:33there are more ways to realize the same
- 32:35macroscopic picture. Maxwell, having
- 32:38realized a significant part of what I
- 32:40am talking about now, came up with the
- 32:42following thought experiment. "Let's,"
- 32:45he said, "start the opposite process."
- 32:49One part of the vessel will get colder,
- 32:50and the other will get hotter. How to
- 32:53achieve this? Hotter and colder is
- 32:56because molecules move faster or slower
- 32:58. Well, look, when people go to a
- 33:01restaurant, they think about which
- 33:03restaurant they will go to,
- 33:05uh, how much money they are ready to
- 33:07spend there, and so on. But imagine
- 33:10that all people go to restaurants
- 33:11completely randomly, completely by
- 33:13chance. So you were let out onto the
- 33:15street next to us here; there are quite
- 33:17a few establishments there. And you
- 33:19randomly go into any door, regardless
- 33:21of anything. Now, if you are a
- 33:25restaurateur who wants people to come
- 33:27to you and spend more money, how do you
- 33:29do it? You make a quickly opening and
- 33:33closing door and put a specially
- 33:34trained person at the entrance who
- 33:36determines if this one has a lot of
- 33:37money or a little. And to those who are
- 33:41poorer, he says: "No, it's closed,
- 33:43that's it, go anywhere else. I'm not
- 33:46interested in where you go." But for
- 33:48those he assesses as having more money,
- 33:50he opens the door. He doesn't even
- 33:52invite them; he just opens the door.
- 33:54What will be the result? In this
- 33:55restaurant, people will start spending
- 33:57much more money. But this is under the
- 33:58assumption, of course, that people go
- 34:00there randomly. Well, molecules do
- 34:01exactly the same thing. They rush
- 34:03around randomly, and they are
- 34:04constantly colliding. And some have
- 34:06higher speed, others have lower speed.
- 34:07Speed is a measure of kinetic energy.
- 34:09Roughly speaking, it's MV squared over
- 34:11two. Uh, that is, the energy they carry
- 34:13depends on the speed. That is, those
- 34:15who are faster, when it is hot, it
- 34:17means that the molecules are simply
- 34:19moving faster, yes. Well, strictly
- 34:21speaking, it's not about speed, but the
- 34:23average kinetic energy of molecular
- 34:25motion, but we can talk about speed for
- 34:27simplicity. Now imagine that you have a
- 34:29vessel which is divided, uh, into two
- 34:31parts by a partition with a little door
- 34:33. Well, just like in my restaurant, the
- 34:35door is practically weightless. And
- 34:37sitting there is a small creature, like
- 34:40a bouncer, but tiny, tiny, called
- 34:42Maxwell's demon, which does what? If a
- 34:45fast molecule approaches him from this
- 34:47side on my right, he quickly opens the
- 34:49door. And the molecule goes into the
- 34:51left part of the vessel. And if it's
- 34:52slow, he closes the door. And if a slow
- 34:55molecule approaches from the other side
- 34:56, conversely, that's even cooler than
- 34:58in my restaurant,
- 34:59If someone wants to leave, someone poor
- 35:01has entered the restaurant and wants to
- 35:03leave, please, go ahead. But a rich
- 35:04person, on the contrary, is not let out
- 35:06. Yes, if someone slow approaches from
- 35:08the other side, he says: "Please,
- 35:10please, the door is open." If someone
- 35:12fast approaches, he closes the door.
- 35:14What happens? Fast molecules will
- 35:17gather in one part of the vessel, and
- 35:19slow ones in the other.
- 35:20Uh-huh.
- 35:21This means that in one part the
- 35:22temperature will be higher, and in the
- 35:24other lower. And this means that you
- 35:26can, uh, perform work. You can, like
- 35:30our piston, how a steam engine works,
- 35:33why a locomotive moves, you make
- 35:34something hot, and the hot steam in
- 35:36this case expands and pushes the piston
- 35:39forward toward the side that is colder.
- 35:43So, if part of my vessel has heated up,
- 35:45I can attach a piston to this, and this
- 35:48piston will push toward the cold side.
- 35:50What kind of load can I lift? So, this
- 35:52is a thought experiment. It is
- 35:54completely unimportant here how it is
- 35:55implemented in practice. This is
- 35:57discussing the foundations, well, of
- 35:58our understanding of how the world is
- 36:00structured. Therefore, these are all
- 36:02just thought experiments. It turns out
- 36:04that I performed, extracted useful work
- 36:07, lifted some load, the piston moved, a
- 36:09winch turned, some weight was raised. I
- 36:12performed useful work. But what is the
- 36:14source of energy for me? Actually,
- 36:17nothing at all. I didn't spend anything
- 36:19. You might say that this demon spends
- 36:21something on opening and closing the
- 36:23door. But then it turns out that you
- 36:26can—again, this is a matter of a
- 36:28thought experiment, a theoretical
- 36:30question—but it turns out that you
- 36:32can make the door practically
- 36:33weightless, and he can spend a
- 36:35negligibly small amount of energy
- 36:37opening and closing the door. So, there
- 36:39will be an energy gain. He spent very
- 36:43little, and the hotter gas he created
- 36:45pushed the piston, and the piston
- 36:47performed some work. Now, uh, you might
- 36:50say: "Well, he needs to determine which
- 36:53molecule is fast and which is slow."
- 36:56Don't worry, you can shine infrared
- 36:58light on them with very low energy. And
- 37:02this information can also be read, in
- 37:05general, almost for free. And energy
- 37:07can be obtained, resulting in a
- 37:09paradoxical conclusion. Uh, that is
- 37:12what was called the Maxwell's Demon
- 37:14paradox. Maxwell came up with this idea
- 37:18to show, look, how we are used to the
- 37:20fact that you cannot extract energy,
- 37:23perform useful work, in such a way that
- 37:25it becomes cold here and hot there. It
- 37:28always happens the other way around for
- 37:30us. If it is hot somewhere and cold
- 37:32somewhere else, the temperatures become
- 37:34roughly equally warm. And so that you
- 37:36push here, it gets hot here, and then
- 37:38this heat starts pushing some piston.
- 37:39We don't have anything like that. That
- 37:41would be a source of energy, more or
- 37:43less. This is called, in short, a
- 37:45perpetual motion machine of the second
- 37:47kind. Well, simply put, a perpetual
- 37:48motion machine of the first kind is
- 37:50just
- 37:51a direct violation of the law of
- 37:52conservation of energy. Well, you just
- 37:54say that energy isn't conserved, I'm
- 37:56making it out of thin air or some other
- 37:58nonsense you made up on the spot. Which
- 38:00doesn't exist. And, as is well known,
- 38:02since some long-forgotten year, even
- 38:03the Paris Academy of Sciences no longer
- 38:05considers perpetual motion machines of
- 38:07the first kind. I don't know exactly
- 38:09since when for the second kind. A
- 38:11perpetual motion machine of the second
- 38:12kind is this type of heat engine.
- 38:14Conversely, you just take molecules and
- 38:16sort them for free into hot and cold
- 38:18ones. Well, that's a good question. Is
- 38:20Maxwell's demon possible or not? Can we
- 38:24really reverse the increase of entropy?
- 38:28And can we violate this second law of
- 38:30thermodynamics, which is close in its
- 38:32formulation to the impossibility of a
- 38:34perpetual motion machine of the second
- 38:36kind, that you can't get work just by
- 38:38cooling. As the only result of a
- 38:41process where the only outcome was
- 38:42cooling one body and heating another.
- 38:47What will stop the demon, this
- 38:48wonderful creature, from doing this? By
- 38:53the way, over time it became clear that
- 38:55you don't have to perceive him as a
- 38:57creature. A computer will handle this
- 39:00job perfectly. It's a somewhat
- 39:02computerized task.
- 39:03And in general, you'll quickly figure
- 39:05out that it's all about information.
- 39:06Because what happens? That the demon,
- 39:09no matter how it's constructed,
- 39:11Uh-huh. Needs to record information
- 39:16about an approaching molecule, if only
- 39:18to then pass it to the device that
- 39:20controls the door. And you suddenly
- 39:27realize the following thing: that his
- 39:29memory will start to bloat. He will
- 39:31handle one molecule, then the second,
- 39:34third, fourth, fifth, tenth. If you
- 39:36have a person standing at a restaurant
- 39:38door, he doesn't notice that he forgets
- 39:40what happened with the previous ones.
- 39:42But here, the question of forgetting
- 39:44becomes quite relevant. This demon is
- 39:47made of something. There is some device
- 39:49there, and something has changed inside
- 39:52it. Some molecule moved somewhere, some
- 39:54little thing got magnetized differently
- 39:56. Imagine it however you like. You can
- 39:58imagine any hypothetical scheme. It
- 40:02doesn't have to be realistic, but the
- 40:04state must change. This means his
- 40:06memory is finite. After a while, his
- 40:10memory will overflow, and he won't be
- 40:12able to handle the next incoming
- 40:14molecule. That's it, the process will
- 40:17end.
- 40:17But why can't he just forget after
- 40:18every molecule?
- 40:19Ah, an excellent question. What does it
- 40:21mean to forget?
- 40:22Well, erasing information.
- 40:24Okay, good. What does "erase" mean? To
- 40:27erase means, look, exactly what it
- 40:29means is wonderful. This is the root,
- 40:32the root of this whole story. Erasing
- 40:35means the following. Regardless of
- 40:37whether
- 40:37Uh-huh.
- 40:38it was a zero or a one, it must become
- 40:40one specific thing. That is, you must
- 40:43come up with a procedure for what it
- 40:45means to erase, to write everything as
- 40:47a zero,
- 40:48but these are some specific physical
- 40:50systems. In it, one state is called a
- 40:54zero, and another state is called a one
- 40:56. Szilard came up with the Szilard
- 40:59engine. It illustrates very well
- 41:01everything you are talking about. And
- 41:03we will now discuss exactly what it
- 41:04means to erase. Everything will become
- 41:06as clear as day. This is a wonderful
- 41:08thing. It is almost what I started with
- 41:10. It is a vessel divided into two
- 41:13halves. Uh-huh.
- 41:14And you have only one molecule. It is
- 41:16either in one half or the other.
- 41:18If it is in the left half, it is a zero
- 41:20. If it is in the right half, it is a
- 41:22one. This is one way of recording
- 41:24information. Yes, it is ridiculous, but
- 41:26it is very good. But we do not care if
- 41:28it is ridiculous or not. What is
- 41:29important to us is the principle. In
- 41:31principle, it is an excellent way of
- 41:33recording information. Now look, what
- 41:35does it mean that you need to erase
- 41:37this information? You need to ensure
- 41:40that these two, it does not matter, you
- 41:41do not know in advance where the
- 41:43molecule is: in the left half or the
- 41:45right. You need a universal procedure
- 41:47that will erase it. What do you do? You
- 41:51push the piston to the same side all
- 41:52the way, and then you pull it out, and
- 41:54then the entire vessel will definitely
- 41:56be empty. And you have erased the
- 41:58information, you are freeing up the
- 41:59volume. But in doing so, you experience
- 42:01resistance from this molecule.
- 42:03We are doing work.
- 42:04You are doing work, you are spending
- 42:06energy. It turns out that erasing
- 42:09information is something that
- 42:11inevitably dissipates a certain amount
- 42:13of energy. Landauer, whom we started
- 42:16with, calculated this amount of energy.
- 42:19This is called the Landauer principle.
- 42:21You can treat it in different ways. One
- 42:24can consider it self-evident, or that
- 42:26it follows from the second law of
- 42:27thermodynamics. Or it follows from
- 42:31others under some reasonable
- 42:33assumptions, and then it helps you
- 42:35justify the second law of
- 42:36thermodynamics. Here, in such
- 42:40foundational things, regarding the
- 42:42basic laws, quite a lot depends on what
- 42:44you accept as axioms and what you
- 42:46derive from those axioms. Well, in
- 42:51general, the principle is: erasing one
- 42:54bit of information costs you the
- 42:56dissipation of kT log 2 energy.
- 43:00Boltzmann constant multiplied by
- 43:01absolute temperature multiplied by the
- 43:03natural logarithm of two. This is a
- 43:05certain terribly small, but quite
- 43:07specific number. And it turns out that
- 43:12maintaining your demon, its operations,
- 43:14and its memory will require energy and
- 43:17will require the dissipation of energy.
- 43:21The following surprising picture
- 43:23emerges. It seems as if, at first
- 43:25glance, the demon has disrupted the
- 43:27flow of entropy. It led to a decrease
- 43:29in entropy because it collected hot
- 43:32molecules on one side, fast molecules
- 43:34on one side, and slow molecules on the
- 43:36other. And there are fewer ways for
- 43:38this. This state has fewer ways to be
- 43:40realized than when they are all
- 43:42uniformly mixed. But in reality, there
- 43:46is a compensating increase in entropy
- 43:48in its memory. And dealing with memory
- 43:52clutter requires the dissipation of
- 43:53energy. As a result, an amazing thing
- 43:56happens, which people did not
- 43:58understand immediately, and not even in
- 44:01Maxwell's time, but significantly later
- 44:03.
- 44:03Maxwell was puzzled. Maxwell probably
- 44:05remained puzzled. Well, Maxwell is an
- 44:07absolutely great person, he invented
- 44:09many things. Maxwell's equations on
- 44:11electricity are great. So, it was
- 44:13discovered that the compensating growth
- 44:15of entropy happens in the memory, and
- 44:18the dissipation of energy occurs when
- 44:20erasing information. Immediately, new
- 44:23supporters of the demon appeared.
- 44:26Uh-huh.
- 44:27They said: "Let's come up with this."
- 44:29We know how to archive information. You
- 44:32take a file and zip it. It is smaller
- 44:34in size, takes up less volume. So let
- 44:39the demon buy some kind of archiving or
- 44:41Zip device. There. And connect it to
- 44:46itself and it will archive the contents
- 44:48of its memory. It will be smaller in
- 44:51volume, and then it won't erase every
- 44:53time, but only when some archive
- 44:55accumulates. That is, a large amount of
- 44:57memory accumulates, it compresses it,
- 44:59and it fits into a smaller number of
- 45:01cells. It will have to spend less
- 45:03energy, and it will still deceive us.
- 45:05Yes, it will spend some amount of
- 45:07energy on this erasure, but it won't be
- 45:09for every cell because it will all fit
- 45:11into a smaller number of cells. The
- 45:13file size, roughly speaking, and the
- 45:15point of archiving is that the file
- 45:17size decreases. The demon fooled us,
- 45:19and here the idea of Kolmogorov or some
- 45:22other complexity appears, which says: "
- 45:24You know, some sequences zip very well.
- 45:28If I tell you a sequence of 1s 10 and
- 45:34keep going for an hour,
- 45:37then you can describe it in just a few
- 45:39simple symbols. 1 repeats many, many
- 45:42times. That's it. That's it. As a
- 45:46result, it turns out that some
- 45:48sequences zip well, while sequences
- 45:50that have no order or pattern zip very
- 45:53poorly. Well, by the way, we know that
- 45:56there are things like JPEG files that
- 45:58don't compress well because everything
- 46:00that can be archived in them has
- 46:02actually already been archived. So, if
- 46:04you take a random sequence, it zips
- 46:06very poorly; it is very difficult to
- 46:09archive. Due to random reasons, it
- 46:11might archive well, but on average, it
- 46:13will archive very poorly because there
- 46:16is no order in it whatsoever. And the
- 46:20sequence of zeros and ones that gets
- 46:22recorded in the demon's memory is
- 46:24deeply and fundamentally random,
- 46:26because there is nothing to control
- 46:28which molecule will approach it first,
- 46:30which second, which third, fourth, or
- 46:32fifth. Thus, it turns out once again
- 46:36that the demon will not be able to
- 46:38deceive us. But I want to emphasize how
- 46:40, uh, how several sciences have become
- 46:43intertwined here. Uh, it started with
- 46:46thermodynamics, the science of heat.
- 46:49And is it possible to reverse the
- 46:51increase of entropy through such a,
- 46:53well, almost possible trickery? It,
- 46:55this trickery, this Maxwell’s demon,
- 46:57it seems, in general, it seems possible
- 47:00. But what is the objection? And then
- 47:01the objection moves into a different
- 47:03field. One needs to understand that for
- 47:05it to operate well, for it to work well
- 47:07, information is needed,
- 47:08right? And then one needs to realize
- 47:10that information is inseparable,
- 47:12starting from a certain level,
- 47:14inseparable from the physical nature of
- 47:16things. Information does not just exist
- 47:19; it is recorded in something physical.
- 47:21And then you manipulate this physical
- 47:23medium and find that when you
- 47:24manipulate the demon's memory, one way
- 47:26or another, you dissipate information.
- 47:28As a result, Maxwell's demon provides
- 47:31no gain. The story has come full circle
- 47:33, taking several decades. Szilard is
- 47:36the thirties, by the way. Of the
- 47:38twentieth century. So. But Maxwell is
- 47:40still the mid-nineteenth century. He
- 47:42started a cool process. And these
- 47:45questions, well, it's not that they are
- 47:47still actively discussed today, but
- 47:49they do not completely leave our focus,
- 47:51because this is a very, very deep thing
- 47:53. So, the Landauer principle is that
- 47:56the erasure of information has a
- 47:57thermodynamic cost. You dissipate a
- 47:59certain amount of energy and increase
- 48:01entropy. And this is, on the whole, an
- 48:04instructive story, telling us that, uh,
- 48:06well, it communicates a few things. It
- 48:08says that information, on one hand,
- 48:10seems like an abstract thing, but on
- 48:12the other hand, it is not incorporeal;
- 48:14it has a physical content. And when you
- 48:18start digging at a microscopic level,
- 48:20uh, information as an abstract
- 48:22phenomenon suddenly turns out to be
- 48:24something quite physical. And
- 48:27information theory and thermodynamics
- 48:29get tangled and become one thing. This
- 48:31is a very beautiful story that began
- 48:33with Maxwell.
- 48:34And can we then say that information is
- 48:36a fundamental physical quantity on par
- 48:38with energy and mass?
- 48:40I don't know. I simply don't know about
- 48:42that. I don't I don't have a point of
- 48:45view on this. But let's, for the
- 48:48duration of today's conversation,
- 48:50assume that this is one of those
- 48:52wonderful undefinable concepts. Either
- 48:56you define it through something else
- 48:58that is undefinable, or you consider it
- 49:00fundamental. Right? And you see,
- 49:03information has this very dangerous
- 49:06aspect: information for whom?
- 49:09And it immediately seems that this
- 49:11leads to philosophical musings that are
- 49:14, more or less, endless. Information is
- 49:17information only when someone can use
- 49:19it, when it is information for someone.
- 49:21Well, even if you didn't know the
- 49:23language I'm speaking, you would still
- 49:25draw some conclusions about me. But if
- 49:28you were completely detached, like a
- 49:30Martian or something, you wouldn't
- 49:32understand my intentions, my state, or
- 49:34anything at all. So, information is a
- 49:38complex thing, but indeed quite
- 49:40fundamental.
- 49:41Well, I think we've figured it out.
- 49:42Well, let's hope so.
- 49:44Yes. Alexey Mikhailovich, thank you so
- 49:46much for this podcast. It was very
- 49:48interesting for me to listen to you.
- 49:50Many things that people used as
- 49:52arguments for living in a simulation,
- 49:54we have, one way or another, debunked.
- 49:57Do you have any final words? Perhaps,
- 50:02I wish all your viewers and listeners
- 50:03to be interested in interesting things.
- 50:09And to remember that science is
- 50:10something in constant development. That
- 50:13is what makes it interesting. We are
- 50:15always turning, scientists are turning
- 50:17the unknown into the known. They are
- 50:19surrounded by the unknown. You have to
- 50:20be able to endure that. A scientist's
- 50:22professional state is saying:" I don't
- 50:24know this, I don't know that, and I
- 50:26don't know the other. "That is their
- 50:28job. To look at what they don't know,
- 50:31and at how things become known to us.
- 50:33Moreover, things that are hard to even
- 50:36imagine, hard to conceive, and
- 50:37counterintuitive. It is a hymn to our
- 50:40brain, to our human nature. We turn out
- 50:45to be capable of stepping so far beyond
- 50:48the familiar, beyond what is
- 50:49intuitively clear to us. We can
- 50:54logically describe the world in much
- 50:56greater complexity than is apparent
- 50:58from our immediate experience.
- 51:00And do you think we will make progress
- 51:02soon in understanding what information
- 51:04actually is? In terms of fundamental
- 51:07things, perhaps not, but in terms of
- 51:09the interconnection of various
- 51:11fundamental concepts, yes, science is
- 51:13heading there. Redefining fundamental
- 51:15concepts is very difficult, but finding
- 51:17deep connections between them is the
- 51:19task of science.
- 51:20Everyone should definitely subscribe to
- 51:22your social media. Which ones should be
- 51:24mentioned?
- 51:25I don't have social media, so you
- 51:26should definitely subscribe to Gleb's
- 51:28channel. Ah, I have become
- 51:32I have become the author of two books
- 51:34some time ago. One is called Everything
- 51:36That Moves: A Walk Through the Restless
- 51:38Universe, and the other was released
- 51:40for the quantum mechanics centenary.
- 51:42And right now we are in the year of the
- 51:45quantum mechanics centenary, it's
- 51:46called 100 Years of Ambiguity. Quantum
- 51:49Mechanics for Everyone: 25 Essays. 100
- 51:51Years of Ambiguity. 100 years. Is it
- 51:54clear why" ambiguity "? Because quantum
- 51:56mechanics is such a thing-in-itself.
- 51:57That's what I talk about. I saw the
- 51:59book has already
- 52:00been a bestseller for a long time,
- 52:02honestly speaking. Yes.
- 52:03Alexey Mikhailovich, thank you very
- 52:05much once again for this podcast, dear
- 52:08friends. Well, and be sure to subscribe
- 52:10to this YouTube channel, subscribe to
- 52:12my Telegram channel, leave your
- 52:14comments, join the discussion, and
- 52:16definitely read our guest's books. See
- 52:19you soon. Bye.
- 52:20Take care. Thank you for your attention
- 52:22.
About this transcript
This page contains the full transcript of 55 минут, которые ПЕРЕВЕРНУТ Ваше представление о РЕАЛЬНОСТИ - Алексей Семихатов by Глеб Соломин, generated from the public captions YouTube serves with the video. The transcript has 9,041 words across 1,449 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.