Making Sense - Sam Harris - #86 — From Cells to Cities

00:00:00.000 Welcome to the Making Sense Podcast.

00:00:08.820 This is Sam Harris.

00:00:10.880 Just a note to say that if you're hearing this, you are not currently on our subscriber

00:00:14.680 feed and will only be hearing the first part of this conversation.

00:00:18.440 In order to access full episodes of the Making Sense Podcast, you'll need to subscribe at

00:00:22.720 samharris.org.

00:00:24.140 There you'll find our private RSS feed to add to your favorite podcatcher, along with

00:00:28.360 other subscriber-only content.

00:00:30.520 We don't run ads on the podcast, and therefore it's made possible entirely through the support

00:00:34.640 of our subscribers.

00:00:35.880 So if you enjoy what we're doing here, please consider becoming one.

00:00:46.380 Today I am speaking with Jeffrey West.

00:00:49.720 Jeffrey is a theoretical physicist whose primary interests have been in fundamental questions

00:00:55.080 of physics and biology.

00:00:56.840 He's a senior fellow at Los Alamos National Laboratory and a distinguished professor at

00:01:03.360 the Santa Fe Institute, where he served as president from 2005 to 2009.

00:01:09.560 He's been named to Time Magazine's list of 100 most influential people in the world, and

00:01:15.280 he is the author of the very fine book, Scale, The Universal Laws of Growth, Innovation, Sustainability,

00:01:22.440 and the Pace of Life in Organisms, Cities, Economies, and Companies.

00:01:28.100 And we talk about his book at length here.

00:01:31.400 As you'll hear, Jeffrey is an extremely interesting guy.

00:01:35.120 We ran into a few audio problems at the end, so apologies for that.

00:01:39.760 All I can say is that our robot overlords don't yet have this internet thing fully worked out.

00:01:44.380 But I should say that this conversation is pretty dense.

00:01:48.460 I didn't really appreciate how dense it was until I re-listened to it.

00:01:52.840 There's a lot of information here.

00:01:54.940 Those of you who are students of physics and mathematics will absolutely love it.

00:01:59.940 But some of you will find that you really need to concentrate to follow Jeffrey where he goes.

00:02:06.160 And you might need to rewind from time to time or just listen to the whole thing twice.

00:02:10.720 But this will repay your attention because Jeffrey is doing some very deep and interesting work.

00:02:16.540 And his book is really wonderful.

00:02:19.380 And now, without any further delay, I bring you Jeffrey West.

00:02:30.200 I am here with Jeffrey West.

00:02:32.100 Jeffrey, thanks for coming on the podcast.

00:02:33.940 Yeah, pleasure to be here, Sam.

00:02:35.220 Thank you for inviting me.

00:02:37.320 You have written this fascinating book called Scale, which links the underlying properties of complex systems to both biological and cultural phenomenon, really everything from cells to cities.

00:02:52.560 Yes.

00:02:52.980 And it's a fascinating route into basically everything we care about.

00:02:58.100 And the book is filled with disarmingly simple-sounding questions, which turn out not to be simple at all.

00:03:04.980 But they're questions like, why do we live 100 years rather than 1,000?

00:03:09.460 Why do we stop growing?

00:03:10.980 We keep eating all the time, but at some point we stop growing.

00:03:13.640 It's not obvious why that should be the case.

00:03:15.620 Yeah.

00:03:15.820 Why do people die and companies die, but cities don't seem to die?

00:03:20.780 And before we get into answering these questions, first tell our listeners how you got into this, because you're a theoretical physicist by training.

00:03:30.020 And now you're focusing on biological and even socioeconomic questions.

00:03:34.360 And it seems to have been inspired both by the death of the supercollider project in the U.S.

00:03:40.540 and your growing sense of your own mortality.

00:03:42.880 So give us the context of your investigations.

00:03:46.380 Yes.

00:03:46.740 No, thank you.

00:03:47.440 Yes, indeed.

00:03:48.120 You have pinpointed, so to speak, the genesis of this in that I was at some stage happily doing research into quarks and gluons and string theory and fundamental questions of physics, dark matter and so forth.

00:04:08.860 And associated with that, of course, was this marvelous project of the superconductive supercollider to be built in Texas.

00:04:20.040 And, of course, the vision was to open up new vistas at extremely high energies and, therefore, in very short distances and confirm some of our ideas about fundamental forces and the fundamental constituents of nature.

00:04:37.840 But also, you know, just the usual search for, you know, new science, new physics.

00:04:45.180 And sadly, that was canned in the early 90s.

00:04:48.380 And I had been somewhat involved in it.

00:04:52.480 And at the same time, I was into my 50s.

00:04:59.100 And it so happens that I come from a line of short-lived males.

00:05:06.700 Very few live beyond about 60.

00:05:09.260 And many have died in their 50s.

00:05:12.520 I've got a similar problem.

00:05:13.780 I'm just edging into my 50s.

00:05:16.000 And I'm a year shy of the age my father made it to.

00:05:20.300 So, yeah, I follow your mind.

00:05:22.460 Well, very similar.

00:05:23.420 Well, my father did make it to almost 61.

00:05:26.860 But his father, you know, died at 57.

00:05:31.540 And my father's brother died at 54 and so forth.

00:05:34.680 It was that.

00:05:35.080 So it's a similar kind of thing.

00:05:36.220 And so I'd grown up with this idea that, you know, I'd probably die somewhere in my early 60s.

00:05:45.180 That was sort of the lifespan of what was to be expected.

00:05:48.420 And in my 50s, I began to realize that, my gosh, you know, I may only have five to 10 years at most to live.

00:05:57.040 And it was the confluence of that and the death of the super collider and some of the things that were being said in terms of, so to speak, justifying why we shouldn't continue with this huge project.

00:06:16.420 That got me to start thinking about some of these big questions in biology originally.

00:06:21.620 And the one thing that really stimulated me and sort of got me emotionally was a statement that many people are familiar with that was being banded around, especially in the early 90s, was, you know, physics was the science of the 19th and 20th centuries, whereas biology is clearly going to be the science of the 21st century.

00:06:42.780 And I must say it's sort of hard to argue with that, but there was a corollary to it that was sometimes actually made explicit, oftentimes just implicit.

00:06:54.100 And that was that, you know, we know all the physics we need to know, and, you know, there's no point in, you know, going any further, kind of philistine view of the intellectual enterprise.

00:07:08.040 And that really got me, you know, because even though, as I said, I agreed with the first part that no doubt biology was going to be significantly important during this century.

00:07:19.740 Nevertheless, I arrogantly and out of ignorance, frankly, sort of came back with the statement that, well, yes, that may be the case, but it won't be a real science unless it starts to...

00:07:35.740 Until it gets a proper case of physics envy.

00:07:38.000 Exactly. Yeah, something like that. Yes, I was hesitating. But to really, you know, start to incorporate the paradigm of physics in terms of it being quantitative, more analytic, based on principles, and therefore more predictive.

00:07:59.920 That kind of paradigm, that kind of paradigm, and also some of the techniques of physics and the question.

00:08:04.720 So the big question is, you know, to what extent can biology be mathematized and put on a kind of principled basis beyond just, in quotes, the principle of natural selection, but to put that into a more solid foundation.

00:08:20.440 That was where I was coming from before. And I must say, I knew almost no biology at the time.

00:08:27.080 But it was that sort of emotional reaction that got me to start thinking at some stage, well, you know, maybe I should think about that seriously.

00:08:38.280 Maybe I should actually start thinking about how you would, in fact, take this fantastic set of ways of thinking and tools that we've developed in physics.

00:08:49.620 How could you take that to biology? And that's where it coupled up with the question of aging and mortality, that I started thinking a little bit about that.

00:08:59.560 And the way I framed it in my head was not just what is the mechanism of aging and why do we die, but to make it slightly more quantitative and say, you know, where in the hell does 100 years come from for the lifespan of a human being?

00:09:16.420 You know, what is that related to? We ought to have a theory, you know.

00:09:19.680 So to begin, put it in this kind of arrogant physics way, if biology were a serious science, then, you know, you should be able to pick up a biology textbook and there would be a chapter about aging and mortality in which there'd be a little calculation that ends up with saying lifespan of a human being should be approximately 100 years.

00:09:42.100 And by the way, the lifespan of a mouse should be of the order of two or three years, et cetera.

00:09:48.640 And what I discovered as I started to take this more seriously and read not just biology textbooks, but read the literature on aging and mortality, gerontology in general, was that this was not a very well-developed area at that time.

00:10:06.320 And in particular, as far as I could tell, no one seems to have asked the question in that form.

00:10:12.560 And so I kind of, as a little exercise, so to speak, to, you know, spend my spare time in the evenings or the weekends, I thought, well, maybe I should start thinking about that.

00:10:25.920 How would you go about trying to show that 100 years is the expected lifespan of an animal our size?

00:10:35.400 And what that led me to, first of all, was, you know, if you're going to start thinking about aging mortality, you have to start thinking about what is it that's going wrong in terms of what's keeping you alive?

00:10:51.060 I mean, apply that to any machine, for example, what is it that, so to speak, wears out or starts to become dysfunctional in terms of its, during its lifespan?

00:11:04.300 And of course, you know, what's keeping you alive is metabolism.

00:11:09.220 That is, you eat and metabolize food to form energy, ATP molecules, currency of energy.

00:11:16.160 And so then I started reading about metabolism, and in so doing, discovered, I didn't discover, but I learned about these amazing scaling laws.

00:11:31.500 And in particular, this remarkable scaling law for how metabolic rate, that is the amount of energy any organism needs, you know, per second or per hour to stay alive.

00:11:43.580 How does that, how that scales with the size of an animal?

00:11:48.820 And to my amazement, I learned that it was extremely simple and regular.

00:11:56.820 Jeffrey, before we jump into biological scaling, let's just answer a couple of higher level questions here, because I don't even think people understand what is implied by the word scaling.

00:12:07.840 Yes, I was going to come to this, absolutely.

00:12:09.520 Yeah, so let's, let's, the big picture here is that you point out that the phenomenon that we, that really concern us, that form the space in which we live, span a range of more than 30 orders of magnitude, from molecules to cities.

00:12:25.180 So can you put that in context first?

00:12:27.420 Yes, sure.

00:12:27.880 So, so first of all, just take organisms.

00:12:33.300 We go from the smallest organism, which is mycoplasma, it's a, you know, tiny sub-bacterial kind of organism, all the way up to the blue whale, that's about 20 orders of magnitude, you know, 20 powers of 10.

00:12:49.640 So it's, it's enormous.

00:12:51.660 If you include, if you go down to molecules, you add several more orders of magnitude, and of course, if you go up to ecosystems and cities, many more.

00:13:01.420 So, you know, I mean, you could even stretch this to 40 orders of magnitude in terms of the structure of life.

00:13:09.800 You know, all of these things are, to some extent, living, I mean, even at the molecular level, you could talk about living, things that are doing things that we would call life, sort of primitive viruses.

00:13:23.200 But all the way up to, as I say, a large ecosystem, and in particular, a city, which you can sort of think of for these purposes as a kind of pseudo-organism.

00:13:33.800 So that's kind of amazing, because, you know, as I think I point out in the book, this is much greater scale than the relationship of us to the entire Milky Way, for example, or an electron to a cat.

00:13:51.820 You know, these, we as life span much more than that, and it's kind of amazing.

00:13:57.560 And so that's the range over which the phenomena I discuss in the book are discussed.

00:14:04.280 But the phenomenon of scaling, that is usually called scaling, is how do the characteristics of, you know, let's say, let's just be a little more modest and stick to, say, just all mammals, for example.

00:14:21.580 How do their characteristics scale as you change the size of a mammal?

00:14:26.360 So mammals go from the smallest, which is a shrew, which sits easily on the palm of the hand, all the way up to the blue whale, which is as big as the building I'm sitting in.

00:14:38.620 And that covers approximately eight orders of magnitude in its mass.

00:14:45.220 And scaling asks the question, well, let's look at characteristics of these mammals, everything from the one I mentioned earlier, metabolic rate, to something a little more mundane, like the length of their aortas.

00:15:01.520 The aorta is the first tube that comes out of the heart, or even, you know, the size of their hearts, or the length of a limb.

00:15:10.400 But all these various things that you could measure, how long they live, how many offspring they have, and so on.

00:15:16.680 So that's just the concept of scaling.

00:15:20.360 And the remarkable thing is that when you look at any of these quantities, and, you know, one can list maybe 50 to 75 of such characteristics,

00:15:32.200 and ask how do they change with the size of the mammal, they all scale, in that sense, in a very regular fashion.

00:15:41.720 And not only in a very regular fashion, but all in a similar way, mathematically.

00:15:49.300 And that's extremely surprising, naively, at a naive level, because, you know, we believe in natural selection.

00:15:58.280 We believe that all of these organisms have evolved by natural selection, with highly contingent histories.

00:16:05.440 Each subsystem of them, each organ, each cell type, each genome, has its own unique history.

00:16:15.620 So you might have expected that if you plotted any characteristics, such as its metabolic rate versus size,

00:16:25.400 you would get points scattered all over the graph.

00:16:30.600 And quite the contrary, you find that there's a tremendous regularity.

00:16:35.440 It gets revealed suggesting that underlying this extraordinary complexity, because, after all, something like metabolism is maybe the most complex process in the universe, for all we know,

00:16:47.620 because it's sort of, you know, at its most primitive level, it takes, you know, matter, stuff, and creates life.

00:16:54.880 That's what we're doing, you know, as we eat, and so on.

00:16:57.760 You know, here's this unbelievably complex process, and yet, if you ask how it scales across this huge range of organisms, it scales in this very simple way.

00:17:10.540 And the amazing thing is this even extends to cities that have different cultural histories and different geographies.

00:17:16.840 Absolutely. So the same thing, after we did this work and explained where these scaling laws come from, it was very natural to ask the question, you know,

00:17:27.980 are there other forms of life, such as, you know, more synthetic ones, so to speak, like cities, or even companies,

00:17:35.660 that express similar kinds of regular systematic scaling.

00:17:41.480 And, as I say, later following understanding the biological scaling, when we looked at the data on the scaling of cities,

00:17:53.360 we found a similar kind of scaling, similar in the sense that there was a regular systematic behavior, and the mathematics was the same.

00:18:02.780 The details of it are different, and the details are different in a very important and powerful way.

00:18:11.480 But it was quite similar, and similarly, even with companies, you know, going from a small company of a couple of hundred employees to Walmart or General Motors,

00:18:25.520 there were similar kinds of scaling.

00:18:27.000 So, you know, there's this ubiquitous behavior that is quite surprising when you first meet it, that says that despite, you know, the daunting complexity and diversity that we see out there,

00:18:42.220 underlying it seems to be a kind of simplicity, which can be expressed both graphically and mathematically in very powerful and simple terms.

00:18:56.560 So maybe I should say a little bit about what the nature of that scaling is.

00:19:00.460 Yeah, yeah.

00:19:00.860 Would that be helpful?

00:19:02.220 Yeah, that'd be great.

00:19:02.980 So let's start with the case of biological scaling, and I just want you to go through the significance of the fact that this scaling tends to be nonlinear.

00:19:14.000 It's either sublinear or superlinear on your account.

00:19:16.380 So what's the significance of that?

00:19:17.960 Yeah, that's very important because, you know, if you ask yourself, well, look, if I double the size of an organism or if I look at an organism that's twice the size of another one,

00:19:30.560 or in particular, let's take mammals, as I said, a mammal that's twice the size of another, then it contains, roughly speaking, twice as many cells.

00:19:39.440 And that's linear scaling.

00:19:41.900 You know, and if it's three times as big, it contains three times as many cells.

00:19:47.580 And that's, roughly speaking, correct.

00:19:49.200 It's a simple linear relationship.

00:19:51.900 However, the scaling of all other characteristics of an organism are nonlinear in the following sense.

00:19:59.860 Take metabolic rate.

00:20:01.420 If you double the size of an organism, instead of needing twice as much energy, twice as much food, if you like, to stay alive,

00:20:09.860 what you discover is you don't need twice as much.

00:20:13.020 You only need 75% as much, even though there are twice as many cells.

00:20:18.240 And this happens systematically.

00:20:20.740 So if you double the size from 4 grams to 8 grams or 4 kilograms to 8 kilograms, it doesn't matter where you start.

00:20:32.460 As long as if you double, you only need, roughly speaking, 75%, three quarters, roughly, the amount of energy.

00:20:39.520 There's a 25% savings on the average every time you double.

00:20:44.240 And that's called an economy of scale.

00:20:46.380 That's a classic economy of scale and means, of course, that the individual cells, since they do scale linearly,

00:20:56.480 it means that the energy needed to support an individual cell is systematically smaller by this 25% rule the bigger you are every time you double.

00:21:11.160 And so, you know, your cells work less hard in a predictable way than your dogs or cats.

00:21:18.800 But, you know, your horse or your elephant are working even less hard.

00:21:24.900 So this is a pervasive phenomenon throughout biology, this economy of scale, and has far-reaching consequences.

00:21:35.960 So that similar kind of scaling gets repeated across any measurable quantity, whether it's physiological, like the one I mentioned,

00:21:47.460 just something sort of mundane like the length of an aorta, or something quite sophisticated,

00:21:54.020 like the rates at which oxygen diffuses across membranes or how long an organism lives and so on.

00:22:01.140 And these also are governed by an analog to this 25% rule.

00:22:08.640 So time scales increase according to this 25% rule the bigger you are.

00:22:15.040 And generically, the pace of life slows down.

00:22:18.960 So that, in fact, you know, if you took an elephant and you followed these scaling laws for all its physiology and all its rates of life history

00:22:32.680 and scaled it according to that and just kept scaling down, you would end up with a mouse.

00:22:39.020 You know, a mouse is a scaled, you know, at this...

00:22:41.400 A tiny elephant.

00:22:41.920 ...18, 90% level is a scaled-down elephant.

00:22:44.640 And by the way, that brings up something that's very important about the nature of these rules, these laws,

00:22:51.800 and that is that they're not like the laws of physics, which we think of as being precise,

00:23:00.160 like Newton's laws or Maxwell's equations for electricity and magnetism or quantum mechanics,

00:23:04.920 where we have this, you know, roughly speaking, this paradigm that you can...

00:23:11.380 With these principles and laws of physics, you can calculate any physical phenomenon in principle to any degree of accuracy

00:23:19.440 so that, you know, we know the positions of all the planets at any time.

00:23:25.560 We know the positions of satellites at any time.

00:23:29.520 That's why we can, you know, get our...

00:23:32.060 Exchange messages on cell phones and so on.

00:23:35.980 Our cell phones work precisely and so forth.

00:23:38.300 So all this works because the laws of physics are extremely precise

00:23:43.440 and we can calculate things and predict things in a highly precise fashion.

00:23:48.960 That is not true of the kinds of laws that I'm talking about,

00:23:54.520 the kind of scaling laws I'm talking about in biology.

00:23:58.000 These are laws that we technically call coarse-grained,

00:24:02.720 meaning that they're only true to, say, 80%, 90% accuracy

00:24:08.940 so that we can predict or you can predict from these laws the following.

00:24:15.200 So just to give you another example,

00:24:18.320 if you give me the size of a mammal,

00:24:21.380 I can tell you pretty much anything about it,

00:24:24.980 everything from, as I said, its metabolic rate,

00:24:28.160 the complete structure of its circulatory system or its respiratory system.

00:24:33.100 I can tell you about how long it will live,

00:24:38.480 how many offspring it will have, and so on and so forth.

00:24:41.940 You know, all these various measurable characteristics.

00:24:45.020 But I can only do it to 80%, 90% accuracy.

00:24:49.060 And if I ask to make a prediction about a very specific elephant

00:24:52.660 or a very specific mouse,

00:24:55.140 I couldn't do it with anything more than that accuracy.

00:24:59.480 So it's, so to speak, it's for the average animal of that size.

00:25:06.560 But of course, that's extremely powerful,

00:25:10.700 not only because it connects, you know,

00:25:13.160 all these different organisms that seemingly look alike

00:25:17.180 and live in very different environments.

00:25:20.480 It connects them all under sort of one umbrella

00:25:24.820 and shows the kind of unity of life.

00:25:27.860 But also, it provides a baseline for asking about specific cases.

00:25:33.500 You know, you can then look at specific animals

00:25:36.900 or specific individuals of that species

00:25:40.120 and start asking questions using the scaling laws as a baseline.

00:25:45.900 Well, so to talk about one variable here, lifespan.

00:25:50.040 So as you get bigger as an animal,

00:25:53.040 perhaps we should confine this to mammals.

00:25:55.160 As you get bigger, you tend to live longer.

00:25:59.080 And this follows the scaling law that you,

00:26:01.720 this is a consequence of metabolism slowing down

00:26:05.880 and economies of scale?

00:26:07.780 Yeah.

00:26:08.060 So the, so this is, so let me back off now

00:26:12.260 and talk more generally about the origin of these scaling laws.

00:26:16.620 What is, where in the hell do they come from?

00:26:18.740 For example, we just talked mostly here about mammals,

00:26:22.060 but the same scaling laws apply to trees and plants

00:26:25.960 in the following sense.

00:26:27.540 Their metabolic rate scales in the same way as ours does.

00:26:31.540 That is, every time you double the size of a tree

00:26:34.500 in terms of its weight,

00:26:36.480 it uses only 75% more energy, just like we do.

00:26:41.220 But for example, the way it's trunk scales,

00:26:47.200 the trunk of a tree scales,

00:26:49.000 is essentially identical to the way our aorta scales.

00:26:54.040 The tree is its own aorta.

00:26:55.520 It's all aorta.

00:26:56.800 It's all circulatory system.

00:26:57.820 Well, the trunk is the aorta.

00:26:58.960 Yeah, exactly.

00:26:59.680 No, so, so let me, let me take that.

00:27:01.980 Let me, let me go from there.

00:27:03.820 The analog to the tree inside us

00:27:08.780 is our circulatory system.

00:27:11.080 The aorta, the analog to the aorta,

00:27:13.860 which is that first tube, as they say,

00:27:15.700 coming out of the heart,

00:27:17.080 the analog to that is the trunk of the tree,

00:27:20.060 the part that goes up before it branches

00:27:22.320 into two or three other big branches.

00:27:26.160 And indeed, the origin of these scaling laws,

00:27:29.920 because you ask yourself, you know, what is it?

00:27:31.720 What is it that's common among plants, trees, mammals,

00:27:36.260 birds, fish, et cetera,

00:27:38.080 that they all seem to obey these same scaling laws,

00:27:41.980 even though their evolved engineered design

00:27:46.080 is quite different.

00:27:47.640 Obviously, you know, we have beating hearts.

00:27:49.680 Trees certainly don't have beating hearts,

00:27:51.640 just to take a dramatic example.

00:27:54.380 So you ask, what is it that's common among all of them?

00:27:57.060 And what you realize is what's common among all of them

00:27:59.420 is that they have all evolved

00:28:02.260 to be hierarchical branching network systems.

00:28:05.880 And you sort of understand that

00:28:07.740 because, you know, just think of yourself,

00:28:09.540 you're made of 10 to the 14th cells,

00:28:12.280 roughly 100 trillion cells.

00:28:14.580 And each one of those has to be serviced

00:28:17.620 in some, roughly speaking,

00:28:19.660 democratic and efficient fashion.

00:28:21.160 And the way that problem has been solved

00:28:24.120 is by evolving these networks

00:28:27.180 that deliver oxygen and nutrients and so on

00:28:32.280 and information from, if you like,

00:28:35.780 a central reservoir down to the cellular level.

00:28:40.200 And as I say, one, we're very familiar

00:28:42.080 with our circuitry system,

00:28:43.460 our respiratory system,

00:28:45.080 our neural system,

00:28:47.000 our renal system,

00:28:48.000 and so on.

00:28:49.740 And all of these have those characteristics.

00:28:52.460 And the idea is

00:28:53.620 that it is the mathematics and physics,

00:28:59.060 the sort of universal generic mathematics and physics

00:29:02.800 of these network systems

00:29:05.240 at all scales

00:29:06.640 that are being reflected in the scaling laws.

00:29:10.380 And that was the work that I got involved in.

00:29:13.280 And we, you know,

00:29:14.920 it's quite complicated mathematics

00:29:17.880 to work it all out.

00:29:19.660 But out of that pops

00:29:20.960 these remarkable scaling laws.

00:29:24.300 And I want to say a couple of things

00:29:26.060 about the networks

00:29:27.060 because it's not just the networks,

00:29:29.740 but they have special properties

00:29:31.760 which are, roughly speaking, universal.

00:29:35.300 And one of them

00:29:36.660 is that they are

00:29:38.640 what we technically call space filling.

00:29:40.520 It's a very simple concept.

00:29:43.020 And it's simply that

00:29:44.940 whatever the structure of the network,

00:29:48.060 its terminal units,

00:29:49.520 in our case,

00:29:50.360 for example,

00:29:50.900 the circuitry system,

00:29:52.020 the terminal units

00:29:52.880 are capillaries

00:29:53.980 that feed cells. 0.97

00:29:55.700 Those capillaries,

00:29:57.180 so to speak,

00:29:57.700 have to go everywhere

00:29:58.640 because every cell in the body

00:30:00.740 has to be fed

00:30:02.160 by oxygen

00:30:04.060 diffusing from blood,

00:30:06.060 from the capillaries

00:30:06.900 to cells.

00:30:07.600 So the endpoints

00:30:10.160 of the network

00:30:11.140 have to end up

00:30:12.380 close by cells.

00:30:15.420 And so the network

00:30:17.080 in that sense

00:30:17.820 has to be space filling

00:30:19.020 and go everywhere.

00:30:20.180 As, for example,

00:30:22.260 in a city,

00:30:23.360 the road networks

00:30:24.920 essentially have to service

00:30:27.460 all buildings

00:30:28.600 and ultimately all people.

00:30:30.520 the street system

00:30:36.340 doesn't leave

00:30:38.280 vast areas of houses

00:30:40.480 without any access to them.

00:30:43.040 So it is with our bodies.

00:30:45.180 So that concept

00:30:45.980 is called space filling

00:30:47.160 and that has to be put

00:30:48.020 into some mathematical terms

00:30:50.300 and that's one of the inputs

00:30:51.940 to the,

00:30:53.100 or one of the constraints,

00:30:54.360 I should say,

00:30:55.560 on the network.

00:30:56.420 I think we could introduce

00:30:57.860 a mathematical concept here

00:30:59.260 that will be familiar

00:31:00.560 to people

00:31:01.220 but it seems relevant

00:31:02.520 that the concept

00:31:03.320 of a fractal,

00:31:05.060 which, you know,

00:31:05.860 I think,

00:31:07.140 it seemed like in the 80s

00:31:08.680 literally everyone

00:31:10.080 knew what fractals were.

00:31:11.520 I mean,

00:31:11.800 like the barber

00:31:12.520 was telling you

00:31:13.260 about the Mandelbrot set.

00:31:15.060 Exactly.

00:31:15.640 So we had reached

00:31:17.700 peak fractal back then

00:31:18.940 but I'm not sure

00:31:20.020 the knowledge has stuck

00:31:21.080 so perhaps you could

00:31:22.160 remind people about

00:31:23.160 what fractals are

00:31:24.200 and their significance.

00:31:24.880 Yes, let me do one last thing

00:31:28.160 before doing that

00:31:29.020 because it relates

00:31:29.760 directly to it

00:31:30.860 and that is another constraint

00:31:33.340 on the network

00:31:34.320 and that is that

00:31:36.640 in some sense

00:31:38.220 the network optimizes the system.

00:31:41.500 I say that loosely

00:31:42.820 but let me give you an example

00:31:44.120 because it leads to fractals

00:31:45.620 and that is that

00:31:47.480 the circulatory system

00:31:50.720 that we have

00:31:52.660 and that has evolved

00:31:54.500 by the process

00:31:55.240 of natural selection

00:31:56.220 and by we have,

00:31:57.840 I mean the we

00:31:58.440 I'm referring to

00:31:59.200 is all mammals.

00:32:00.440 That is all mammals

00:32:01.680 that now exist

00:32:02.600 and all mammals

00:32:03.380 that have ever existed.

00:32:05.700 The one that we have

00:32:07.780 minimizes

00:32:09.340 the amount of energy

00:32:11.220 our hearts have to do

00:32:12.780 to pump blood

00:32:14.060 through our circulatory system

00:32:17.160 to feed cells

00:32:18.380 so that we can maximize

00:32:21.660 the amount of energy

00:32:22.960 we can devote

00:32:24.060 to what is called

00:32:26.980 Darwinian fitness

00:32:27.920 meaning that we can devote

00:32:29.680 to having sex 0.93

00:32:31.020 and rearing children. 0.98

00:32:33.600 And so

00:32:34.220 that's very important.

00:32:36.940 So that means that

00:32:38.380 whatever the structure

00:32:39.160 of the network is

00:32:40.080 not only does it have

00:32:40.900 to be space filling

00:32:41.720 but

00:32:42.540 its structure

00:32:44.000 has to be

00:32:44.860 that if we changed it

00:32:46.940 in any significant way

00:32:49.180 you know

00:32:49.880 by just say

00:32:50.520 doubling

00:32:51.020 the length

00:32:52.700 of the third branch

00:32:53.960 of your arterial system

00:32:56.000 that would increase

00:32:58.180 the amount of energy

00:32:59.280 your heart has to do

00:33:00.580 and similarly

00:33:01.940 if you halved

00:33:02.900 this eighth branch

00:33:04.300 of your arterial system

00:33:05.640 it would increase

00:33:06.840 the energy.

00:33:07.380 So we sit

00:33:08.200 in a kind of basin

00:33:09.220 of optimization

00:33:11.540 so to speak

00:33:12.340 of minimizing

00:33:13.340 the energy

00:33:13.900 our hearts have to do.

00:33:15.260 It seems to me

00:33:15.680 that that need not be so

00:33:17.800 in evolutionary terms

00:33:19.300 I mean there's a lot

00:33:20.320 obviously

00:33:20.960 there's a lot about us

00:33:22.320 that an engineer

00:33:23.600 would not have put in place

00:33:25.240 and I put

00:33:25.920 I put the prostate gland

00:33:27.360 high on the list

00:33:28.220 of things

00:33:29.040 you would not have engineered.

00:33:30.420 Absolutely.

00:33:31.300 That's just mathematically

00:33:32.160 so at this point

00:33:33.200 we can say

00:33:33.700 that it is optimal.

00:33:34.680 So here was the idea

00:33:35.900 the idea was that

00:33:37.260 you know

00:33:37.840 in order to start

00:33:39.500 to take this idea

00:33:41.060 that networks

00:33:41.800 underlie the scaling laws

00:33:43.060 you have to start

00:33:44.320 putting together

00:33:45.120 the mathematics

00:33:45.720 of the networks

00:33:46.580 and as in all physics

00:33:48.680 you need

00:33:49.420 generic principles

00:33:50.640 that transcend

00:33:51.640 you know

00:33:52.580 the individual system

00:33:53.600 you're looking at

00:33:54.340 and you need

00:33:55.040 you know

00:33:55.680 certain assumptions

00:33:56.500 and one of the

00:33:58.720 simplest assumptions

00:33:59.540 was to assume

00:34:01.000 that there was

00:34:02.100 this kind of optimization

00:34:03.120 that by the

00:34:04.340 continuous process

00:34:05.740 continuous feedback

00:34:07.600 process

00:34:08.400 inherent in natural

00:34:09.680 selection

00:34:10.240 the you know

00:34:11.940 mammals that have 0.60

00:34:13.560 survived

00:34:14.500 that we are

00:34:16.100 tend towards

00:34:17.540 minimization

00:34:18.320 of this

00:34:19.580 you know

00:34:20.400 the amount of energy

00:34:21.200 that we use

00:34:21.960 to keep ourselves alive

00:34:23.100 we minimize

00:34:23.820 the amount of energy

00:34:26.180 that is

00:34:27.260 the mundane process

00:34:28.400 of remaining alive

00:34:30.020 so that we can

00:34:32.200 maximize

00:34:33.100 the amount of energy

00:34:34.680 that we put into

00:34:36.220 our genes

00:34:37.500 going forward

00:34:38.360 and

00:34:39.420 you know

00:34:40.860 that of course

00:34:42.020 need not be

00:34:42.900 on the other hand

00:34:44.200 you know

00:34:44.800 if you believe

00:34:45.240 in Darwinian fitness

00:34:46.280 and you had

00:34:48.020 long enough time

00:34:48.860 which we've had

00:34:49.660 you would expect

00:34:50.540 something like that

00:34:51.360 to happen

00:34:51.780 but anyway

00:34:52.120 that was

00:34:52.560 a hypothesis

00:34:53.860 and it was

00:34:55.160 very natural

00:34:55.720 to hook it up

00:34:56.500 to traditional ideas

00:34:58.320 of Darwinian fitness

00:34:59.560 but you're absolutely

00:35:00.960 right

00:35:01.500 there are many

00:35:02.360 aspects of our

00:35:03.440 physiology

00:35:04.100 especially at my age

00:35:06.280 that you begin to realize

00:35:07.600 weren't exactly

00:35:09.240 designed in the way

00:35:10.580 that maybe

00:35:11.360 they were optimal

00:35:12.240 but you know

00:35:13.380 you have to remember

00:35:14.780 that having said that

00:35:17.380 that

00:35:18.320 that's always the case

00:35:20.180 when you look

00:35:20.860 at one individual

00:35:22.120 component

00:35:23.060 you know

00:35:24.560 like you mentioned

00:35:25.340 you know

00:35:25.600 if you look at one

00:35:26.300 specific thing

00:35:27.160 but you have to remember

00:35:28.100 that that is

00:35:29.340 interconnected

00:35:30.280 with everything else

00:35:31.620 it's a systemic problem

00:35:33.240 and the optimization

00:35:34.600 and that's what

00:35:36.040 part of this idea was

00:35:37.700 is not so much

00:35:38.880 that it's taken place

00:35:39.940 at the highly local level

00:35:41.680 and this is extremely important

00:35:43.280 but it's taken place

00:35:44.480 at the systemic level

00:35:45.680 I'm talking about

00:35:46.700 the systemic level

00:35:48.300 of each one

00:35:49.360 of these

00:35:49.800 network systems

00:35:50.980 so it's the entire system

00:35:52.460 going from

00:35:53.620 the heart

00:35:54.680 and the entire

00:35:56.540 structure

00:35:58.000 structure of the

00:35:59.080 circulatory system

00:36:00.340 from your aorta

00:36:01.060 downwards

00:36:01.620 feeding through tissue

00:36:03.480 through capillaries

00:36:05.660 to cells

00:36:06.460 and how that

00:36:07.180 diffusion takes place

00:36:08.360 for example

00:36:08.780 all of this

00:36:09.560 is one huge system

00:36:10.780 and it has to be

00:36:12.040 and the idea

00:36:12.740 is that

00:36:13.240 it is the systemic

00:36:14.380 optimization

00:36:15.220 rather than

00:36:16.760 the local

00:36:17.620 optimization

00:36:19.680 I'll take your point

00:36:20.880 Jeffrey

00:36:21.140 but if you're going

00:36:21.740 to argue

00:36:22.080 that the prostate

00:36:22.940 gland is a masterpiece

00:36:24.040 of nature

00:36:25.060 and it's God

00:36:25.780 you're going to have

00:36:26.860 a tough time

00:36:27.440 on this podcast

00:36:27.860 it is not

00:36:28.160 and I certainly

00:36:32.200 agree with you

00:36:32.880 with that

00:36:33.320 or the way

00:36:34.860 backs are designed

00:36:35.920 for example

00:36:36.580 or the way

00:36:38.320 I mean

00:36:38.680 I still

00:36:39.960 am amazed

00:36:41.680 at the whole

00:36:42.160 process

00:36:42.820 of both

00:36:44.260 reproduction

00:36:45.760 and child

00:36:47.620 delivery of

00:36:49.360 fetuses

00:36:49.940 into the world

00:36:51.620 why does it have

00:36:52.500 to be a medical

00:36:53.080 emergency every time

00:36:54.240 yeah exactly

00:36:55.060 I mean

00:36:55.720 so you know

00:36:56.780 obviously

00:36:57.600 all of this

00:36:58.200 but you know

00:36:58.840 those do not

00:37:00.100 happen

00:37:00.460 my point is

00:37:01.400 they do not

00:37:03.040 happen

00:37:03.520 in a vacuum

00:37:04.680 I mean

00:37:05.200 they're all

00:37:05.720 interconnected

00:37:06.420 you know

00:37:07.180 and no doubt

00:37:08.220 something

00:37:08.780 about that

00:37:10.020 birth delivery

00:37:11.520 has all kinds

00:37:12.780 of other

00:37:13.400 implications

00:37:14.220 not just

00:37:14.980 physiological

00:37:15.600 but social

00:37:16.640 implications

00:37:17.280 of course

00:37:18.080 and so on

00:37:19.500 so I don't

00:37:19.960 want to argue

00:37:20.480 this

00:37:20.840 this is not

00:37:21.500 you know

00:37:22.680 this is a

00:37:23.180 secondary thing

00:37:23.860 really

00:37:24.640 to the main

00:37:25.420 point

00:37:25.840 that

00:37:27.140 you know

00:37:27.720 when you

00:37:28.120 look at

00:37:28.580 these networks

00:37:29.360 and you

00:37:29.880 apply these

00:37:30.880 underlying

00:37:31.840 generic

00:37:32.760 systemic

00:37:33.380 principles

00:37:34.000 to them

00:37:34.740 one of the

00:37:35.920 things that

00:37:36.540 you learn

00:37:37.300 is that

00:37:38.580 the optimal

00:37:39.760 system

00:37:40.320 is fractal

00:37:41.980 like

00:37:42.380 I'll use

00:37:42.820 that word

00:37:43.400 and fractal

00:37:44.240 means

00:37:45.320 another

00:37:45.960 word for

00:37:46.500 it is

00:37:46.760 self-similar

00:37:47.500 and we're

00:37:48.960 all very

00:37:49.480 familiar with

00:37:50.160 it

00:37:50.380 I'm looking

00:37:51.480 out at the

00:37:52.040 moment

00:37:52.380 at a tree

00:37:53.100 and you

00:37:54.980 know it

00:37:55.200 has this

00:37:55.780 hierarchical

00:37:56.540 branching

00:37:57.060 network

00:37:57.540 and the

00:37:58.480 fractality

00:37:59.080 is expressed

00:37:59.740 by the fact

00:38:00.380 that if you

00:38:00.840 cut some

00:38:01.660 branch

00:38:02.240 and remove

00:38:03.260 it

00:38:03.520 it looks

00:38:04.100 like a

00:38:04.420 little

00:38:04.640 tree

00:38:05.060 right

00:38:05.520 and then

00:38:07.420 you can

00:38:07.800 take that

00:38:08.340 little

00:38:08.700 tree

00:38:09.100 and cut

00:38:10.160 a branch

00:38:10.700 of that

00:38:11.200 and take

00:38:12.540 it away

00:38:13.000 and it

00:38:13.340 looks like

00:38:13.800 an even

00:38:14.080 smaller tree

00:38:14.920 and so on

00:38:15.600 and that's

00:38:16.060 the idea

00:38:16.620 that you

00:38:17.340 have this

00:38:17.860 repetitive

00:38:18.660 self-similarity

00:38:19.980 and the

00:38:22.240 theory

00:38:23.020 is one of

00:38:24.760 the things

00:38:25.260 that comes

00:38:26.620 out of it

00:38:27.060 is that

00:38:27.560 there is in

00:38:28.580 fact that

00:38:29.160 the systems

00:38:29.900 should be

00:38:30.620 fractal

00:38:31.220 in order to

00:38:32.280 optimize in

00:38:33.020 the way I

00:38:33.480 said and

00:38:34.060 also

00:38:34.640 critical

00:38:35.760 fill all

00:38:37.600 of space

00:38:38.120 that is

00:38:38.720 that it

00:38:39.260 needs to

00:38:40.260 every part of

00:38:42.180 the system

00:38:42.660 needs to be

00:38:43.220 serviced

00:38:43.660 by the way

00:38:44.720 I use the

00:38:45.360 word fractal

00:38:46.000 like

00:38:46.520 because

00:38:47.240 actually

00:38:48.240 the rules

00:38:49.400 that evolve

00:38:50.060 that come

00:38:50.520 out of the

00:38:51.000 theory

00:38:51.460 actually

00:38:52.760 are

00:38:53.580 variants

00:38:54.460 of a

00:38:54.800 fractal

00:38:55.240 there you

00:38:55.880 know to

00:38:56.120 be a bit

00:38:56.460 more technical

00:38:57.020 about it

00:38:57.620 there it's

00:38:59.340 not a

00:38:59.680 precise

00:39:00.200 self-similar

00:39:01.240 in other

00:39:01.580 words and

00:39:02.480 it's in fact

00:39:03.020 true the

00:39:03.800 data shows

00:39:04.460 this that

00:39:05.020 if you do

00:39:05.560 take a

00:39:05.980 tree and

00:39:06.960 you cut a

00:39:07.540 piece out

00:39:08.020 of it it

00:39:08.740 does look

00:39:09.400 like the

00:39:09.940 tree but

00:39:10.340 if actually

00:39:10.800 if you do

00:39:11.740 measurements

00:39:12.260 the theory

00:39:12.800 predicts

00:39:13.240 this

00:39:13.420 it deviates

00:39:14.540 in a

00:39:15.020 predictable

00:39:15.460 way from

00:39:16.340 the original

00:39:16.840 tree but

00:39:18.140 it's very

00:39:18.500 close to

00:39:19.020 this idea

00:39:19.740 of repetitive

00:39:21.420 self-similarity

00:39:22.500 and one of

00:39:23.860 the wonderful

00:39:24.460 things that

00:39:25.740 you know you

00:39:26.820 discover in

00:39:27.580 all of this

00:39:28.200 and that is

00:39:28.840 related to

00:39:29.420 the scaling

00:39:29.940 laws is

00:39:31.220 that all

00:39:31.980 of these

00:39:32.340 systems have

00:39:33.640 somewhere in

00:39:34.640 them some

00:39:35.980 manifestation of

00:39:38.660 this regularity

00:39:40.620 this fractal

00:39:41.460 regularity

00:39:42.320 that seems

00:39:43.540 that seems

00:39:43.560 to permeate

00:39:44.300 nature

00:39:44.740 and some

00:39:46.540 of that

00:39:46.920 is no

00:39:48.740 doubt

00:39:49.140 related to

00:39:53.540 let's put it

00:39:54.020 that way

00:39:54.420 some of it

00:39:54.980 is related

00:39:55.540 to this

00:39:56.700 idea that

00:39:57.720 something is

00:39:59.160 being optimized

00:40:00.100 yeah so

00:40:00.760 to connect

00:40:01.940 the self-similarity

00:40:03.100 and the

00:40:03.400 seemingly

00:40:03.840 endless

00:40:04.340 divisibility

00:40:05.200 of these

00:40:05.720 branching

00:40:06.200 networks

00:40:06.680 to the

00:40:07.920 space filling

00:40:08.420 problem

00:40:08.860 just in a

00:40:09.420 vivid way

00:40:09.900 this is a

00:40:10.680 fact you

00:40:11.320 describe in

00:40:12.020 the book

00:40:12.360 so if you

00:40:13.060 ask what the

00:40:14.060 size of our

00:40:14.780 lungs are

00:40:16.020 they are about

00:40:17.820 the size of

00:40:18.400 a football

00:40:20.060 but the

00:40:21.260 surface area

00:40:22.500 of the

00:40:23.460 respiratory

00:40:23.880 membranes

00:40:25.120 in there

00:40:25.900 is about the

00:40:27.060 size of a

00:40:27.560 tennis court

00:40:28.060 because of

00:40:28.720 just how

00:40:29.420 how endlessly

00:40:30.220 branching it

00:40:31.020 is down

00:40:31.500 at the

00:40:31.820 smallest

00:40:32.400 scale

00:40:33.220 yeah so

00:40:34.000 it's kind

00:40:34.320 of wonderful

00:40:34.760 feeling that

00:40:35.520 inside you

00:40:36.200 is a

00:40:36.740 tennis court

00:40:37.400 you know

00:40:37.860 I mean

00:40:38.200 actually

00:40:38.620 and indeed

00:40:40.040 if you took

00:40:40.740 your circulatory

00:40:41.440 system

00:40:41.880 and you

00:40:42.200 laid all

00:40:42.680 those

00:40:43.140 vessels

00:40:45.580 end to

00:40:46.220 end

00:40:46.680 I forget

00:40:47.680 the precise

00:40:48.720 answer

00:40:49.020 I think it

00:40:49.380 was a hundred

00:40:50.120 thousand

00:40:50.500 kilometers

00:40:51.820 I believe

00:40:52.280 yes

00:40:52.560 yeah you

00:40:53.780 go around

00:40:54.240 the earth

00:40:54.700 certainly

00:40:55.420 more than

00:40:55.840 once

00:40:56.220 and that's

00:40:57.880 kind of

00:40:58.140 amazing

00:40:58.640 you know

00:40:59.020 it's an

00:40:59.300 amazing

00:40:59.840 image

00:41:01.580 it's almost

00:41:02.480 spiritual

00:41:03.120 that feeling

00:41:03.940 that inside

00:41:04.540 you is this

00:41:05.220 unbelievable

00:41:06.260 length

00:41:07.560 of tubing

00:41:09.700 but that

00:41:11.140 and that

00:41:11.640 it's

00:41:12.060 very systematic

00:41:13.640 you know

00:41:14.120 it's

00:41:14.340 it's

00:41:14.700 it's

00:41:15.300 structure

00:41:15.720 is obeying

00:41:17.100 very simple

00:41:18.280 mathematical rules

00:41:19.700 that are

00:41:20.620 like these

00:41:21.480 kinds of

00:41:21.900 rules

00:41:22.160 that I

00:41:22.380 mentioned

00:41:22.720 earlier

00:41:23.140 these

00:41:23.660 so-called

00:41:24.280 power laws

00:41:25.020 there's

00:41:25.640 something

00:41:25.880 spooky

00:41:26.360 about the

00:41:26.920 power laws

00:41:28.180 themselves

00:41:28.880 I mean

00:41:29.140 there's

00:41:29.500 this one

00:41:29.940 number

00:41:30.380 to which

00:41:30.800 you've

00:41:31.000 alluded

00:41:31.260 that

00:41:31.540 runs

00:41:32.040 through

00:41:32.320 this

00:41:32.600 that

00:41:32.780 almost

00:41:33.160 could

00:41:33.420 put

00:41:33.720 someone

00:41:34.100 in the

00:41:34.380 mind

00:41:34.740 of the

00:41:35.220 pseudoscience

00:41:36.600 of numerology

00:41:37.440 the fourth

00:41:38.040 power

00:41:38.460 the fact

00:41:39.460 that basically

00:41:40.140 all these

00:41:40.540 living systems

00:41:41.300 scale

00:41:41.900 to the

00:41:42.460 one-fourth

00:41:42.940 power

00:41:43.280 well the

00:41:44.280 one-quarter

00:41:44.780 yes

00:41:45.160 the number

00:41:45.560 four

00:41:46.220 is

00:41:47.500 the number

00:41:48.260 that permeates

00:41:48.920 all of

00:41:49.300 these scaling

00:41:49.760 laws

00:41:50.060 I said

00:41:50.420 the three

00:41:50.760 quarters

00:41:51.140 for metabolic

00:41:51.660 rate

00:41:52.160 it's one

00:41:53.180 quarter

00:41:53.620 for time

00:41:55.440 scales

00:41:56.100 for example

00:41:57.520 and so on

00:41:59.620 and for

00:42:00.000 lengths

00:42:00.380 it's very

00:42:00.820 similar

00:42:01.180 it's always

00:42:01.740 one quarter

00:42:02.420 comes in

00:42:03.040 and that's

00:42:03.840 that 25% 0.75

00:42:04.860 savings

00:42:05.540 that's not

00:42:06.220 an accident

00:42:06.720 and that's

00:42:07.640 the thing

00:42:08.020 that comes

00:42:08.440 out of

00:42:08.800 the theory

00:42:09.420 based on

00:42:10.540 these

00:42:11.000 mathematical

00:42:11.600 principles

00:42:12.460 of network

00:42:14.640 design

00:42:15.240 and that

00:42:16.520 pops out

00:42:17.240 and that

00:42:18.220 four

00:42:18.700 by the

00:42:19.140 way

00:42:19.520 if you

00:42:20.740 look at

00:42:21.120 the

00:42:21.240 mathematics

00:42:21.680 and ask

00:42:22.140 where it

00:42:22.440 comes

00:42:22.680 from

00:42:22.980 it's

00:42:24.200 the

00:42:24.580 following

00:42:24.840 it turns

00:42:25.180 out the

00:42:25.480 four

00:42:25.920 is actually

00:42:27.200 so to speak

00:42:28.300 not four

00:42:28.960 it's actually

00:42:29.380 three plus

00:42:30.000 one

00:42:30.360 which sounds

00:42:31.020 you know

00:42:31.480 like a

00:42:32.640 paradox

00:42:33.140 it's three

00:42:34.280 plus one

00:42:34.800 meaning

00:42:35.240 the three

00:42:36.640 part of the

00:42:38.160 three plus one

00:42:39.000 comes from

00:42:40.120 the fact

00:42:41.340 that we live

00:42:41.840 in three

00:42:42.160 dimensions

00:42:42.680 the up

00:42:43.800 down

00:42:44.160 and sideways

00:42:44.780 and the

00:42:47.700 one

00:42:48.180 is a

00:42:49.300 reflection

00:42:49.800 of the

00:42:50.600 fractality

00:42:51.340 of these

00:42:52.820 systems

00:42:53.420 that

00:42:55.520 it's well

00:42:57.380 known

00:42:57.800 among those

00:42:58.560 that learn

00:43:00.120 about fractals

00:43:00.940 that they

00:43:02.080 have peculiar

00:43:03.160 sense of

00:43:03.860 dimensions

00:43:04.360 and a

00:43:06.420 fully fractal

00:43:07.180 system

00:43:07.740 is one

00:43:08.720 that effectively

00:43:09.720 adds an

00:43:10.580 extra dimension

00:43:11.380 so there's

00:43:12.040 this kind

00:43:12.400 of weird

00:43:12.980 extra dimension

00:43:13.920 and that's

00:43:16.000 this one

00:43:16.580 so the

00:43:17.420 four is

00:43:17.880 actually

00:43:18.120 three plus

00:43:18.600 one

00:43:18.800 so if we

00:43:19.240 lived in

00:43:19.760 eight

00:43:20.580 dimensions

00:43:21.280 we would

00:43:23.660 be dominated

00:43:24.120 we had

00:43:24.600 life

00:43:25.020 it would

00:43:26.180 be

00:43:26.560 then

00:43:27.320 everything

00:43:27.960 would be

00:43:28.320 dominated

00:43:28.800 by the

00:43:29.300 one-ninth

00:43:29.880 power

00:43:30.320 and we

00:43:32.520 would be

00:43:32.900 instead of

00:43:33.320 saving

00:43:33.940 25%

00:43:35.960 we'd be

00:43:36.980 saving

00:43:37.640 you know

00:43:38.240 one-ninth

00:43:39.020 about 11%

00:43:40.020 I feel the

00:43:41.120 temptation

00:43:41.500 to remind

00:43:42.080 people

00:43:42.380 of just

00:43:42.740 what it

00:43:43.380 means

00:43:43.900 to be

00:43:44.300 adding

00:43:44.660 a dimension

00:43:45.220 here

00:43:45.620 in fractal

00:43:46.420 terms

00:43:46.740 so

00:43:47.480 it's a

00:43:48.480 little hard

00:43:48.760 to do

00:43:49.180 on a

00:43:49.660 podcast

00:43:49.980 but people

00:43:50.540 can look

00:43:51.020 up something

00:43:51.720 a figure

00:43:52.540 I think

00:43:52.940 it's called

00:43:53.200 the

00:43:53.460 is it

00:43:54.140 the Koch

00:43:54.620 curve

00:43:55.220 the K-O-C-H 1.00

00:43:56.740 oh yes

00:43:57.660 look this up

00:43:59.500 if you're

00:43:59.880 if you want

00:44:00.400 to follow us

00:44:00.900 down this

00:44:01.220 rabbit hole

00:44:01.620 but there's

00:44:02.420 there's

00:44:02.620 this image

00:44:03.560 or this

00:44:04.860 this curve

00:44:05.500 which is

00:44:05.940 essentially

00:44:06.380 formed by

00:44:07.440 an equilateral

00:44:08.120 triangle

00:44:08.740 being divided

00:44:09.920 on each

00:44:10.480 of its

00:44:10.760 sides

00:44:11.180 by a

00:44:11.620 smaller

00:44:12.040 one-third

00:44:12.920 size

00:44:13.440 equilateral

00:44:13.900 triangle

00:44:14.380 and you

00:44:14.680 keep doing

00:44:15.220 that

00:44:15.400 just adding

00:44:15.940 triangles

00:44:16.480 upon triangles

00:44:17.260 and you

00:44:17.940 develop a

00:44:18.640 kind of

00:44:18.900 snowflake

00:44:19.680 looking image

00:44:20.560 and then

00:44:21.480 when you

00:44:21.720 ask what's

00:44:22.520 the size

00:44:23.620 of that

00:44:24.160 curve

00:44:24.680 of that

00:44:25.080 figure

00:44:25.560 you know

00:44:26.320 given that

00:44:26.920 in the pure

00:44:27.520 mathematical

00:44:28.040 space

00:44:28.440 you keep

00:44:28.900 doing this

00:44:29.300 infinitely

00:44:29.680 well it's

00:44:30.220 a fully

00:44:30.620 self-contained

00:44:31.700 object

00:44:32.400 which actually

00:44:33.360 has an

00:44:34.060 infinite

00:44:34.460 length

00:44:35.720 of its

00:44:36.400 circumference

00:44:37.140 and this

00:44:38.340 has

00:44:38.760 now this

00:44:39.420 doesn't map

00:44:40.260 onto the

00:44:40.660 real world

00:44:41.420 totally

00:44:42.280 because we're

00:44:43.020 not talking

00:44:43.600 about infinite

00:44:44.180 lengths

00:44:44.600 in terms of

00:44:46.100 the world

00:44:46.860 in which we

00:44:47.200 live

00:44:47.380 but it

00:44:47.820 does to a

00:44:48.860 surprising

00:44:49.380 degree

00:44:49.820 and this

00:44:50.520 Jeffrey

00:44:51.080 you could

00:44:51.560 perhaps

00:44:52.020 remind us

00:44:52.580 of how

00:44:53.100 this was

00:44:53.460 first

00:44:53.720 discovered

00:44:54.220 where

00:44:54.720 you try

00:44:55.240 to measure

00:44:55.680 the boundary

00:44:56.400 between

00:44:56.840 two countries

00:44:57.600 and that

00:44:58.240 becomes

00:44:59.080 remarkably

00:45:00.080 dependent

00:45:00.500 on basically

00:45:01.740 how big

00:45:02.340 a measuring

00:45:02.760 stick

00:45:03.040 you use

00:45:03.480 yes

00:45:04.040 indeed

00:45:04.380 yes

00:45:04.820 that's

00:45:05.180 one of

00:45:05.580 those

00:45:05.760 marvelous

00:45:06.180 discoveries

00:45:06.800 that

00:45:07.300 sort of

00:45:08.520 came out

00:45:09.560 of the

00:45:09.760 blue

00:45:10.040 and something

00:45:10.800 that should

00:45:11.780 have been

00:45:11.980 known since

00:45:12.900 the Greeks

00:45:13.400 but wasn't

00:45:14.440 and it was

00:45:15.160 discovered by

00:45:15.720 a man named

00:45:16.380 Richardson

00:45:16.920 who was a

00:45:18.640 kind of a

00:45:19.760 polymath

00:45:20.200 but he was a

00:45:21.080 kind of a

00:45:22.140 geographer

00:45:22.600 and one of the

00:45:24.100 things that he

00:45:24.540 was interested

00:45:25.080 in

00:45:25.480 was

00:45:27.400 the length

00:45:29.860 of boundaries

00:45:30.780 between

00:45:31.800 countries

00:45:32.540 and he

00:45:33.800 was interested

00:45:34.260 in this

00:45:34.660 by the way

00:45:35.300 because he

00:45:35.840 had a

00:45:36.120 theory of

00:45:36.640 war

00:45:37.000 that

00:45:37.760 somehow

00:45:38.360 the

00:45:38.980 incidence

00:45:40.320 of

00:45:40.840 conflicts

00:45:41.440 between

00:45:42.000 nations

00:45:42.600 was

00:45:43.360 proportional

00:45:43.860 to the

00:45:44.520 length

00:45:45.180 of their

00:45:45.460 boundaries

00:45:45.860 and by

00:45:46.180 the way

00:45:46.400 we're

00:45:46.760 talking

00:45:46.980 about

00:45:47.280 he

00:45:49.200 developed

00:45:49.720 that

00:45:50.180 around

00:45:50.760 the time

00:45:51.100 of the

00:45:51.280 first

00:45:51.460 world war

00:45:52.000 but this

00:45:52.800 work on

00:45:53.340 measurement

00:45:53.760 came much

00:45:54.340 later

00:45:54.760 when he

00:45:55.140 was trying

00:45:55.480 to really

00:45:56.540 get a

00:45:56.900 quantitative

00:45:57.320 handle

00:45:57.880 on this

00:45:59.300 and so

00:46:01.280 he got

00:46:02.020 hold of

00:46:02.540 all these

00:46:02.900 maps

00:46:03.420 of

00:46:04.580 various

00:46:05.040 places

00:46:05.620 and he

00:46:06.920 started

00:46:07.160 measuring

00:46:07.600 their

00:46:07.820 boundaries

00:46:08.200 and one

00:46:08.740 of the

00:46:09.000 curious

00:46:09.300 things

00:46:09.600 that he

00:46:09.860 first

00:46:10.100 discovered

00:46:10.580 was that

00:46:12.000 I think

00:46:13.820 the first

00:46:14.240 one was

00:46:14.600 between

00:46:14.880 Spain and

00:46:15.520 Portugal

00:46:15.940 where he

00:46:16.860 found

00:46:17.280 looking at

00:46:18.080 different

00:46:18.480 maps

00:46:18.940 very detailed

00:46:19.660 maps

00:46:20.080 that he

00:46:21.820 got

00:46:22.000 completely

00:46:22.780 different

00:46:23.140 answers

00:46:23.560 I mean

00:46:24.280 I don't

00:46:24.720 remember the

00:46:25.100 numbers

00:46:25.420 I wrote

00:46:25.940 them in

00:46:26.200 the book

00:46:26.540 but

00:46:27.380 you know

00:46:27.880 instead

00:46:28.140 of

00:46:28.320 you know

00:46:28.600 one

00:46:28.820 map

00:46:29.060 might

00:46:29.340 give

00:46:29.740 1100

00:46:31.380 kilometers

00:46:31.960 and then

00:46:33.500 he'd look

00:46:34.500 at another

00:46:34.880 map

00:46:35.240 and you'd

00:46:35.960 get

00:46:36.200 650

00:46:37.040 kilometers

00:46:37.720 and he

00:46:39.000 would look

00:46:39.340 these up

00:46:39.760 in various

00:46:40.180 places

00:46:40.580 and indeed

00:46:41.100 he'd find

00:46:41.520 different

00:46:42.400 books

00:46:44.340 recording

00:46:44.980 these things

00:46:45.600 giving

00:46:45.880 completely

00:46:46.500 different

00:46:46.820 numbers

00:46:47.260 and this

00:46:49.140 was very

00:46:50.040 mysterious

00:46:50.580 and he

00:46:51.860 started

00:46:52.160 looking around

00:46:52.840 and he

00:46:53.160 looked across

00:46:53.820 many

00:46:54.280 countries

00:46:54.720 and he

00:46:55.160 discovered

00:46:55.540 the same

00:46:56.100 phenomenon

00:46:56.600 and he

00:46:58.540 you know

00:46:59.040 he was

00:46:59.300 very puzzled

00:46:59.860 by this

00:47:00.440 and he

00:47:02.580 did realize

00:47:03.280 what was

00:47:03.700 going on

00:47:04.500 but he

00:47:05.640 didn't

00:47:06.180 formalize

00:47:07.680 it

00:47:07.920 it was

00:47:09.140 formalized

00:47:09.660 later by

00:47:10.180 a man

00:47:10.460 named

00:47:10.640 Benoit

00:47:10.960 Mandelbrot

00:47:11.620 who termed

00:47:12.260 the phrase

00:47:12.640 fractal

00:47:13.200 and it's

00:47:14.280 the following

00:47:14.900 it was

00:47:16.060 that

00:47:16.580 when people

00:47:19.140 made these

00:47:19.720 measurements

00:47:20.140 when you

00:47:20.760 make a

00:47:21.180 measurement

00:47:21.580 you have

00:47:23.380 to have

00:47:23.860 a ruler

00:47:24.980 with a

00:47:25.480 certain

00:47:25.680 scale

00:47:26.140 and you

00:47:27.780 have a

00:47:28.040 certain

00:47:28.220 resolution

00:47:28.780 so you

00:47:29.420 might

00:47:29.720 measure

00:47:30.180 someone

00:47:31.000 might

00:47:31.400 measure

00:47:31.780 a

00:47:32.080 boundary

00:47:32.480 using

00:47:33.180 a

00:47:33.760 resolution

00:47:34.200 of only

00:47:34.620 one

00:47:34.860 mile

00:47:35.200 so you're

00:47:35.660 measuring

00:47:35.940 something

00:47:36.320 that's a

00:47:36.680 thousand

00:47:36.940 kilometers

00:47:37.380 long

00:47:37.820 you only

00:47:38.680 care

00:47:39.080 a

00:47:39.400 resolution

00:47:39.840 of a

00:47:40.180 mile

00:47:40.440 but you

00:47:41.040 might

00:47:41.200 have

00:47:41.360 someone

00:47:41.840 that

00:47:42.600 has a

00:47:43.060 resolution

00:47:43.500 of 10

00:47:43.960 miles

00:47:44.380 and someone

00:47:44.900 else

00:47:45.200 might have

00:47:45.560 a resolution

00:47:46.160 of you

00:47:47.600 know

00:47:47.860 I don't

00:47:48.300 know

00:47:48.460 he could

00:47:48.800 even be

00:47:49.180 a meter

00:47:50.660 in principle

00:47:51.520 and you

00:47:53.340 can

00:47:53.540 immediately

00:47:53.960 when you

00:47:54.400 start

00:47:54.620 thinking

00:47:54.940 about it

00:47:55.300 you realize

00:47:55.720 what the

00:47:56.060 problem

00:47:56.360 is

00:47:56.700 that if

00:47:57.600 you

00:47:57.840 measure

00:47:58.420 a boundary

00:47:59.500 which is

00:48:00.000 a squiggly

00:48:00.580 line

00:48:01.120 and you

00:48:02.160 put a

00:48:02.580 ruler on

00:48:03.140 it

00:48:03.340 where

00:48:03.780 the

00:48:05.420 resolution

00:48:06.700 is

00:48:08.020 say

00:48:08.960 10

00:48:09.380 kilometers

00:48:09.880 then anything

00:48:11.180 below 10

00:48:11.980 kilometers

00:48:12.480 you miss

00:48:13.440 but below

00:48:14.580 that 10

00:48:15.040 kilometers

00:48:15.440 the line

00:48:16.260 the boundary

00:48:16.860 may be

00:48:17.360 squiggling

00:48:17.800 around

00:48:18.240 so you

00:48:18.740 miss

00:48:19.100 you measure

00:48:20.000 that as

00:48:20.340 10

00:48:20.580 kilometers

00:48:21.060 but someone

00:48:22.040 else

00:48:22.360 with a

00:48:22.680 resolution

00:48:23.180 of

00:48:23.520 one

00:48:23.740 kilometer

00:48:24.200 would

00:48:24.940 measure

00:48:25.280 it

00:48:25.520 as

00:48:25.940 25

00:48:26.800 kilometers

00:48:27.340 for example

00:48:28.160 so that

00:48:29.780 was the

00:48:30.120 problem

00:48:30.480 he realized

00:48:31.340 and in

00:48:31.840 fact

00:48:32.180 then he

00:48:32.680 discovered

00:48:33.140 that this

00:48:34.380 followed a

00:48:35.100 very regular

00:48:35.780 pattern

00:48:36.260 that if

00:48:37.080 you plotted

00:48:37.620 the length

00:48:38.780 that's measured

00:48:39.580 or reported

00:48:40.280 versus the 0.55

00:48:41.420 resolution

00:48:41.980 there was

00:48:43.220 a very

00:48:43.780 simple

00:48:44.340 mathematical

00:48:44.900 relationship

00:48:45.560 and

00:48:46.440 amazingly

00:48:47.260 that relationship

00:48:48.160 is just

00:48:48.780 like the

00:48:49.240 relationships

00:48:49.760 I talked

00:48:50.380 about

00:48:50.660 in terms

00:48:51.260 of things

00:48:51.640 like

00:48:51.940 metabolic

00:48:52.700 rates

00:48:53.300 and all

00:48:53.720 the other

00:48:54.160 characteristics

00:48:54.780 of

00:48:55.120 organisms

00:48:55.620 and

00:48:56.580 that's

00:48:57.360 where the

00:48:57.760 connection

00:48:58.100 was to

00:48:58.680 fractals

00:48:59.320 and it

00:49:00.460 was Benoit

00:49:01.160 Maudenbrot

00:49:01.760 who realized

00:49:03.380 that not

00:49:04.280 just that there

00:49:05.000 was this

00:49:05.360 phenomenon

00:49:05.800 of the

00:49:06.680 problem

00:49:07.080 of making

00:49:08.100 measurements

00:49:08.740 and resolution

00:49:09.660 but that

00:49:11.340 in fact it

00:49:12.180 was self

00:49:12.620 similar

00:49:12.940 boundaries are

00:49:13.860 approximately

00:49:14.580 self

00:49:15.020 similar

00:49:15.360 so if

00:49:16.760 you look

00:49:17.140 at one

00:49:17.520 scale

00:49:18.020 and then

00:49:19.140 scale

00:49:19.560 that up

00:49:20.040 it just

00:49:20.600 looks like

00:49:21.260 what the

00:49:21.940 boundary

00:49:22.240 would look

00:49:22.680 like at

00:49:23.040 the bigger

00:49:23.320 scale

00:49:23.720 and so on

00:49:25.020 and so forth

00:49:25.580 as you said

00:49:26.140 this is a

00:49:27.040 genuine mystery

00:49:27.960 why this

00:49:28.600 wasn't discovered

00:49:29.400 literally thousands

00:49:31.300 of years ago

00:49:32.000 this is one of

00:49:33.160 those things

00:49:33.600 that was just

00:49:34.260 staring everyone

00:49:35.080 in the face

00:49:35.780 and most of

00:49:37.760 science is not

00:49:38.380 like that

00:49:38.820 does anyone

00:49:40.760 understand why

00:49:42.140 this wasn't

00:49:43.400 discovered

00:49:43.860 before

00:49:44.700 if you'd like

00:49:51.120 to continue

00:49:51.540 listening to

00:49:52.000 this conversation

00:49:52.700 you'll need to

00:49:53.680 subscribe at

00:49:54.380 samharris.org

00:49:55.300 once you do

00:49:56.520 you'll get access

00:49:57.100 to all full

00:49:57.780 length episodes

00:49:58.380 of the making

00:49:58.800 sense podcast

00:49:59.600 along with other

00:50:00.600 subscriber only

00:50:01.380 content

00:50:01.820 including bonus

00:50:03.080 episodes

00:50:03.680 and amas

00:50:04.740 and the

00:50:05.300 conversations

00:50:05.780 i've been having

00:50:06.340 on the waking

00:50:06.780 up app

00:50:07.260 the making

00:50:08.200 sense podcast

00:50:08.820 is ad free

00:50:09.700 and relies

00:50:10.600 entirely on

00:50:11.320 listener support

00:50:11.980 and you can

00:50:12.960 subscribe now

00:50:13.760 at samharris.org

Making Sense - Sam Harris - July 14, 2017

#86 — From Cells to Cities

Episode Stats

Summary

Transcript