Cesar Hidalgo on Why Information Grows
Oct 26 2015

How%20Information%20Grows.jpg Cesar Hidalgo of MIT and the author of Why Information Grows talks with EconTalk host Russ Roberts about the growth of knowledge and know-how in the modern economy. Hidalgo emphasizes the importance of networks among innovators and creators and the role of trust in sustaining those networks.

RELATED EPISODE
Yuval Harari on Sapiens
Yuval Harari of Hebrew University and author of Sapiens talks with EconTalk host Russ Roberts about the history of humanity. Topics discussed include the move from hunting and gathering to agriculture, the role of fiction in sustaining imagination, the nature...
EXPLORE MORE
Related EPISODE
Cass Sunstein on Infotopia, Information and Decision-Making
Cass Sunstein of the University of Chicago talks about the ideas in his latest book, Infotopia: How Many Minds Produce Knowledge. What are the best ways to get the information needed to make wise decisions when that information is spread...
EXPLORE MORE
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.

READER COMMENTS

Joseph Crivelli
Oct 27 2015 at 5:18am

You touched on this point: Why China or the Chinese can make things well?

I recall reading an unpublished essay/manuscript in the 90s that was attempting to link a Confucian ethic that permeates China and Japan in the same functional way a Protestant ethic so well formed the basis of capitalism in the West. I think this addition could help inform how information is conserved and grows within geographical markets.

Ken Bertagnolli
Oct 27 2015 at 11:01pm

Having worked in manufacturing for the past 20 years, I completely agree with Hidalgo’s idea “that what is most valuable in an economy is your capacity to make.” The capacity to make requires tremendous problem-solving skill. Plan rarely, if ever, equals actual in manufacturing. Therefore workers have to be able to solve problems, at the worksite, at high frequency. I believe that China became a centerpiece of industrial production because the workers were able to solve problems. Even though their productivity may be “coming from the crystallized imagination of other people,” a Chinese factory worker (or any worker, for that matter) cannot simply follow instructions and achieve high levels of productivity. Instead they must generate new knowledge and know-how in order to solve problems related to doing the work. The workers are not the same over time; they do literally get smarter by making things. Likewise the knowledge and know-how of these workers accumulates in networks that populate those countries, cities, and regions, making the network smarter over time as well. Perhaps a lack of problem-solving skill or experience is what keeps places like Ethiopia and Nigeria from doing what China has done.

Michael Byrnes
Oct 28 2015 at 7:35pm

Great comment, Ken Bertagnoli.

Didn’t Adam Smith emphasize the importance of tacit knowledge/knowhow?

Nonlin
Oct 28 2015 at 8:54pm

Network effects are important in the context of market competition. But the market leader is not the only one that has the knowledge. If Hollywood suddenly disappeared from the face of the Earth, we would be watching European or Indian movies for a while and we would be just fine.

Knowledge spreads over time. It used to be that China was the exclusive supplier of silk, or that Apple was the only company that knew how to make smart phones. Not anymore. The shipping industry power center migrated from ??? to Greece, to ??? to the UK, to the US, to Japan, to Korea and now is rising in China. Nigeria and Ethiopia are not doomed to eternal third world status.

Nonlin
Oct 28 2015 at 8:58pm

Read Shipbuilding industry, not Shipping industry.

Russ, the comment section should allow at least one Reply level. It would enhance the communication. Thanks.

Ak Mike
Oct 28 2015 at 10:44pm

Sorry not to be commenting on the podcast but rather on Nonlin’s request to allow a reply level: Please do not allow a reply level or threaded comments. It is far easier to read a single chronological list of comments than to dodge around and try to figure out what has been read before and what has not. It’s not difficult to make it clear in a comment to whom you are responding.

[Folks: The comment section is for comments on the substance and content of the podcast episode. Please use email or perhaps a twitter hashtag for suggestions and critiques of external matters such as the formatting. We appreciate your thoughts on these matters, but when we consider changes, we can’t locate and pull together your ideas if they are scattered around in comment sections. You can reach Russ and me at either mail@econtalk.org or webmaster@econlib.org. P.S. We do already know the matter of comment threading is one of strongly divided arguments. –Econlib Ed.]

Robert Swan
Nov 1 2015 at 4:23pm

I’m not too keen on Prof. Hidalgo’s definition of “information”.

Listening to the early part of this podcast (the milk in coffee part) reminded me of a comment I recently made at bishophill.com (climate sceptic blog). It was regarding Matt Ridley’s latest book and I used the Earth’s atmosphere as an example of how much order can come, apparently randomly, from the bottom up. You have all the molecules of the atmosphere trying to continue on their straight paths, but impeded by bumping into all the other molecules. You have gravity trying to pull them back to Earth. And you have the sun heating up the half facing it. If you didn’t know better you’d think this would lead to outright chaos, but there is actually lots of coherence in the atmosphere: hurricanes, winds, storms and rain, and we make plausible efforts at predicting them.

Just as an aside, and something I didn’t think of when writing the comment at Bishop Hill, in ancient times, the weather seemed intelligent enough that people ascribed it to one god or another. You could say that Earth’s weather has already passed a kind of Turing test.

But I would say that the interactions in the atmosphere (or the milk in coffee) reflect great complexity, but little information. Like the Mandelbrot set, the complex weather is the product of simple rules.

jw
Nov 1 2015 at 10:01pm

Wow! Wow and Wow! What a great podcast!

As usual, I was listening to the podcast while driving at a reasonable speed above the 70mph limit when I found myself clapping at the birth canal observation. Brilliant!

I had originally thought that podcast was going to be about bits and bytes, but fundamental information is so much more interesting! Where to begin?:

– There is no reason for the universe to allow localized entropy reductions. Why does it? Are we just lucky?

– We create new information at a dizzying pace due to our sentience (by the targeted use of energy). How was information created before? Random energy confluences like solar/lightning/waves in the primordial soup? In less than a billion years? Hard to believe. (Yes, I realize that this is just a variation on some creationist arguments.)

– If we are creating information constantly (either by sentience or evolution – BTW, BOTH require the standing on shoulders of giants…), why can’t information also be destroyed? Not just in the Susskind/Hawking sense, but by a culture/society/people losing collective information? It may still remain in a book, but if less people know it, total information has been destroyed. More on this later.

– Say it takes 50 people to design a widget and 1,000 to make a million widgets. Even if, as Russ postulates, the designers are products of their education and experience, think of Linkedin. They are all coming from DIFFERENT educations and experiences. So that means 50^500 (Linkedin relationships) of “information” quanta that the collective brings to the firm. Conversely, the 1,000 makers making millions constantly refine the process (the learning curve) to increase information as well. Great concepts (although I still place a higher value on creation…)

– Thank you for the comment on institutions suppressing/impeding information growth rather than abetting it. I absolutely agree! Now, with respect to China, they are moving in a direction of less suppression and the US (for example) is heading in the direction of MUCH more suppression. Is the 2015 Chinese tennis shoe maker making $2/hr now inherently better than a US shoe maker of 1915 when they were making $2/hr? Absolutely not. Give the 1915 US maker assembly lines and robots and computers and he would bury a modern Chinese maker because the US institutions were MUCH less suppressive. This goes back to both comparative advantage price/markets. It also may say something about the loss of information. Has the US lost the information to make via institution suppression?

– I look at the US, the Fed, the Eurozone, and I fear that we have destroyed a great deal of collective information and learning. It may remain in “WoN”, “ToMS” and “The Road to Serfdom”, but it does us no good if it remains in books. I fear that we are climbing down from the shoulders of giants to listen again to the Siren songs of the Bernie Sanders of the world (only to be crushed on the rocks once again).

– However, when it comes to innovation, he mentions LA, SF, NY, and Boston as bastions of innovation. The fact that these cities are the most expensive in the US is directly related to the huge value created by that same innovation. And they aren’t making tennis shoes in those expensive new skyscrapers in Shanghai (I’m not sure they’re doing anything in them, but that is another story.)

I am sure I will think of more as I digest this densely packed (with information) podcast, but this is a start. Prof. Hildago’s book is on my Kindle now. Thanks for another podcast that exposed me to more information that I didn’t know that I didn’t know…

jw
Nov 2 2015 at 6:57am

R. Swan,

If you drop a single drop of cream into a cup of coffee, we know exactly what the result will be after it is completely stirred and evenly distributed, we can calculate this with ease.

However, with all of the supercomputers in the world, we cannot predict the exact pattern that it makes as it dissolves in the first microseconds after it hits the coffee. The fluid dynamics are too complex and chaotic. We can get a lot closer than we used to because we have much better computers and algorithms, but it is still unpredictable for a single drop (no matter how precisely you set up the experiment, it will never repeat).

The air, water vapor, and oceans are all absorbing and distributing heat in a fluid manner. They are not static. It is currently impossible for us to predict how they interact very far into the future with any precision.

As to the Mandelbrot analogy, think of pi. The millionth decimal point has exactly the same information as the first decimal point. It is not nearly as USEFUL to us as as the first decimal point (as without the first decimal point, things would fall down…), but it’s value as pure information is the same.

So this, in a roundabout way, leads to the Butterfly Effect. There may be simple laws of physics that rule weather (and the Universe), but incomprehensibly minute changes may lead to devastating outcomes. We cannot predict when and where next year’s hurricanes will occur (or climate a decade or century from now).

I believe that we are agreeing, I just wanted to add to the case.

Robert Swan
Nov 2 2015 at 9:15pm

jw,

Yes, I think we agree on the whole, but I’m left a little perplexed at the enthusiasm of your earlier comment.

I think you’re right that Prof. Hidalgo is sometimes using “information” in the technical sense used in Information Theory (as in your pi example). But in that sense it’s the fully mixed coffee that contains the most information (to nitpick your first statement — it’s trivial to get the average distribution of the milk in the coffee, but this is a lossy representation; we can’t derive where all the molecules will actually end up).

As a different example, it is not possible to compress a file of random values. In this sense a random file contains more information than a highly organised one. In a non-technical sense, if it’s truly random, it contains no information.

So it seems to me that he’s playing word games. Perhaps this is a new theme at EconTalk: last week we had Harari’s misuse of “fiction”, now we get Hidalgo’s misleading “information”.

Robert Swan
Nov 3 2015 at 4:39pm

I listened to the podcast again and I’d like to backtrack on my last comment (though I still stand by my first). I think he’s consistent in his use of “information” and I don’t think he’s simply playing word games.

However I still don’t see that his observations help our understanding. It would be exciting if he could propose physical laws of information (parallel to, e.g., the law of conservation of energy) but Hidalgo’s “information” seems more akin to Dawkins’s “gene” (as in The Selfish Gene). It might affect the way we describe something, but it doesn’t actually change what we see — nothing new is revealed.

Comments are closed.


DELVE DEEPER

EconTalk Extra, conversation starters for this podcast episode:

About this week's guest:

About ideas and people mentioned in this podcast episode:Books:

Articles:

Web Pages and Resources:

Podcast Episodes, Videos, and Blog Entries:


AUDIO TRANSCRIPT

 

Time
Podcast Episode Highlights
0:33Intro. [Recording date: October 15, 2015.] Russ: Before introducing today's guest, I want to mention that over on our website, econtalk.org, Amy Willis of the Library of Economics and Liberty, which is our host, has started posting what we call Extras. Essentially they are blog posts with provocative questions to encourage follow-up conversation and learning related to each episode. A lot of you asked for that in our surveys. So, I encourage you to check those out; and follow me on Twitter @EconTalker to find out more. And I usually mention when they get posted. Now for today's guest....
1:34Russ: So, let's start with the basics. You mean something very specific by information that's not necessarily the everyday use of the word. So, why don't you explain what information is to you. Guest: So, when people think about the word 'information' usually they think about messages and they think sometimes even about the meaning of those messages. So, one of the things that I do in my book is, in the beginning, I'm very clear about what I mean by information: it's just not messages, it's not the meaning of the messages. But it's the physical order that we use to transmit messages; but also to transmit the practical uses of knowledge and know-how in products. So, I take a very general definition of information that involves all forms of physical order, where that order is embodied in a brand new car, in a strand of DNA (deoxyribonucleic acid), or in a sequences of words in a book. Russ: And you're a physicist originally, by training. Guest: Yep. Russ: And you embed this conversation and analysis of information in physical context; and you start with entropy--which is not going to be our main focus, but I found it fascinating. It's always intrigued me. Entropy is supposed to be increasing in the universe; and yet, so in the aftermath of the Big Bang, and things very hot, then it started to get cold, and you'd things would kind of die out and we'd just have an empty universe with nothing in it. But as you point out, and as scientists have wondered about for a while, the Earth is a pocket, a small corner of the universe where it appears at times we reverse entropy. So, talk about the physics of that for a minute and how the best way to understand how that reversal happens. Guest: Yep. So to understand information from a physical perspective, one thing that is very important, and that has been emphasized by a lot of scientists before me is the fact that information always needs to be physically embodied. Okay? So, when you think about information from a physical perspective, you are thinking about the different states or configurations that a particle or molecule can take. For instance, if you would talk to a very modern quantum physicist, when he would be describing or she would be describing a collision between a couple of particles, it would not be so much about like the exchange of momentum on energy, but the exchange of information that is embodied in those states. Okay? So, the first thing that we need to understand there is that information is always physically embodied. Now, the thing is, there are many different ways in which information can be physically embodied, because, you know, physics actually has a great diversity of ways in which you can store order. Some of the ways in which sometimes that sequence, that type of order is embodied is in thermal states. These are just like random fluctuations that exists that is at non-zero temperature. And the problem of those thermal states, even though there are many--actually most of these states are just, you know, based on these fluctuations--is that you cannot then extract that information. You cannot use it for anything useful. Because in some way, these are states in which everything has already averaged out. So, the information that we have in the universe that is useful for us is the information that is not necessarily embodied just on these thermal states, but is the information that is embodied on things that are highly ordered, that are actually the opposite of that. Which would be like information that is embodied in a strand of DNA, or in a molecule or sugar. Which is something that is not random, and that eventually we can make use of. So, the universe has a natural tendency to take systems that are highly organized and eventually average them out and put them in a state in which, you know, most configurations that are equally likely. And there is not that much order. Russ: And you give the example in the book of--you put some dye into a liquid; it starts off in one little place; it starts making little patterns. And then eventually it disperses throughout the liquid and at that average level that's not very interesting. Guest: Exactly. So, putting milk in coffee is another example that I think most listeners will be familiar with. When you put milk in coffee, at the beginning you have, like, a beautiful dollop. You have a swirl. But after a while it's just a hazy cloud in which, you know, the particles of milk are mixed with the particles of coffee and water. And there is no structure any more. So, the thing is: How do we get to live on a planet where there is so much structure, if the universe has a tendency to make structure disappear? And in the beginning of the book I explain the tricks that the universe has, to allow information to exist and to grow, despite this tendency that the universe has to be hostile to order. And the tricks are the existence of systems in which you have flows of energy. So this is the work of [?Ila Pergosian?], that helps us understand that information is not free and order is not free. You need to spend energy to produce it. There is a cost. The other ingredient that you need is you need to have ways of saving information. And that's what solids are good for. The DNA in your body, that's not need to recreate the information that it has embodied, need all the time, because it's very stable to thermal fluctuations, so it can store information for long periods of time. And finally, these two things combine to generate systems like biological cells or humans or societies that embody the capacity to compute. That it's that they can use energy to actually beget new information out of old information, by recombining it, by copying it, by duplicating it, and so forth. Russ: And in the case of the Earth, the ongoing energy from the sun, is a key part of that story. And of course, for our daily lives, it's, as you call it, our software--that we have in our bodies. Which is our brain. Which allows us to increase order even though the natural tendency is chaos or disorder. Guest: So, in some way you can think that a lot of what the book is about, is about like the struggle between developing computational systems that are enabled to create information--whether these are cells, humans, or economies--and how these computational systems exploit energy to create sequences of information that they can use in the future. And these could be proteins that they use in the operation of a cell. They could be texts that we use to communicate among humans. They could be objects, like the cellphone that I'm using to communicate with you nowadays. And these objects are solid. They are stable to the [?] fluctuations, and therefore we can use them to communicate and to help this collective[?connected?] computer keep on functioning. Because, you know, we don't need to spend energy to keep on creating them all the time. So, we have a, like, in a more economic language, like a [?savings?] rate that is allowed by the universe, by existing at the range of temperatures in which we can create objects that are stable to thermal fluctuations.
8:39Russ: So, before we get into some of the more economic applications of this, I just want to read a short paragraph that I enjoyed a great deal. And if you want to comment on it, you can. I'm not quite sure--I don't remember why you told the story, but it's such a nice story, so I'm going to read it. It's about the birth of your daughter, when your wife gave birth to your daughter, Iris; your wife is Anna. So, here's--sorry about all that mess-up--the quote:
It took only 26 minutes for Anna to push Iris into the hands of the nervous but focused medical student who received her. Twenty-six minutes sounds like a short time for delivery. And it is. And I will argue that the trip that Iris made that night was not a 26-minute trip down a few inches of birth canal, but a hundred thousand year journey from a distant past to an alien future. In 26 minutes, Iris traveled from the ancientness of her mother's womb to the modernity of 21st century society. Birth is in essence time travel.
Why do you mention that? I kind of remember actually. Do you remember? I'm sorry to surprise you with this. Guest: No, absolutely. That is in some ways the opening story of the book. And it helps illustrate one of the main points of the book. Which is, when we think about the development of our society or our economy, in some ways what we are looking at is the development of changes in which we have ordered a fixed set of atoms that we have in the world. The difference between the civilization that we have now and the civilization that existed in our planet 100,000 years ago is not in the atoms that exist in our planet. Because those are conserved: people a hundred thousand years ago also had access to energy; if you think about it the sun was obviously there; and there were fossil fuels that they didn't know how to exploit. But they were available as well. You know. So what has actually changed in this 100,000 years is the way in which it[?] have arranged those atoms to create a world that allows us to live at a much higher standard of living. But what's interesting about the way that we arrange atoms--unlike other species--is that we arrange atoms by first figuring out configurations of atoms that are useful to mental processes. And then embodying that imagination into objects. So, for instance, if you think of a world that is populated only by ants: Ants have a very fixed vocabulary that they can use to communicate that is based on pheromones. And that is encoded in their genes. But we have a vocabulary that is ever-expanding because we are able to criticize[?] information that comes to our brains. So, when my daughter was born that day, what happened is that she encountered a world in which the light that was filling up that room that night was coming up from incandescent light bulbs that were imagined by people before they were constructed. You know--the music that filled up the room was music that was being played by a tablet computer that was actually running[?] an algorithm that would recommend some [?] and would play them for us. So, she was born in a world that was very different from one in which our species evolved. In big part because our species would have none. And the economy, what my interpretation is, is this ability to transform information into reality. So, this story[?] thing helps illustrate powerfully the idea that, like, what makes our society prosperous is not financial measures that are, I would say, more short term interpretation of what the economy is. But, you know, the ability that we have had a suspicious to create imaginary objects in tangible forms that are useful, and that allow us to access capacities that allow us to be not otherwise possible to access without this ability to embody imagination through tangible objects.
12:22Russ: And, you talk about this in different ways; and it made me think about it as well, which is that: It's not just that we have all these clever ways to use information. We preserve them. And we--so you emphasize a lot our connections to each other at a point in time. But I kept thinking, and as you talk about it now, I keep thinking about our ability to stand on the shoulders of giants. People have come before us who have thought of things, and we don't lose those. That's really what's extraordinary. You talk about it some. But I think that part of the human condition, we don't appreciate it. We obviously understand that there are a lot of things in our lives that we enjoy that we don't understand--we don't know how they work. But the idea that insight in what you call crystallized information, the creativity of past human beings, can still be enjoyed by us today is really a rather remarkable thing. Guest: Absolutely. So, I'm probably guessing[?] here. You might have heard of the work of Robert Boyd, the cultural anthropologist. Russ: No. Guest: He actually has a bunch of like great books. You might consider inviting him to your program. Some of these books are called Not by Genes Alone, and what he talks a lot about is this idea of cumulative culture--which is the fact that, as a species, one of the things that makes us successful is that we are able to accumulate volumes of knowledge and information that are much larger than any individual could, by distributing that accumulation in our society, by being able to perform that accumulation of knowledge and information through generations. So, once there is that [?] he sometimes used to motivate that idea is he make the observation that when people from England tried to cross the Arctic and died, there were people that were relatively sophisticated for their time, they were like 19th century explorers that were coming out from the most developed country in the world, but they died in an environment in which people that were "more primitive" like the Inuit were able to survive for thousands of years. And the question is: Why? And Boyd's answer to that is, the Inuit had an enormous amount of knowledge of how to survive in such environments. And that knowledge had been accumulated over many generations; it was very hard to codify; and therefore, the English men that tried to cross the Arctic in that expedition were unsuccessful at doing so because they didn't know how to do it. And it would have been hard for them to acquire the knowledge of how to do that simply because that knowledge is cumulative; it develops over large amount of generations, and it's accumulated at a collective level--which makes it very hard to transfer, but at the same time very crucial for you to be able to survive in that environment, or [?] work to create things that are relatively complex. Russ: The other thought I had is that we are able to do that relatively easily, partly because we can read books--obviously. But as your example points out, the Inuit couldn't--they didn't write books. If they had, reading it probably wouldn't have helped the English. That wasn't exactly the problem, because a lot of it, as you describe in your book, tacit knowledge. It wasn't necessarily things you can write down. It's things you come to--I don't know--embody and feel without being able to totally explain them. Some of that comes from imitation; some of it comes from trial and error; some comes from a lot of different ways. But I suspect a lot of the different ways that we are able to access the information from the past is through the technology and the machines of the past that we take and just improve. We don't think of it as part of our inheritance. But, of course, it is. So, cars get better and better: nobody has to figure out from scratch how to make a car because it's not like the last generation ends and we say, 'Tell us what you learned about cars so we can make one.' It's this ongoing overlapping process that the knowledge gets preserved through--communication as well as the devices that are already there. Guest: Absolutely. So, 'knowledge' is one of those words that in some ways can only be properly understood if you make some distinctions. And one distinction that is very important, as you mention, is the one between tacit and explicit knowledge. Tacit knowledge is the one that Michael Jordan had when he was playing, you know, basketball in the 1990s. And I think here we could both agree that we could have talked with Michael Jordan for days, hours, months, years and we would have never been able to develop his capacity to put a ball inside a hoop. Russ: Yeah. It's easy! Guest: Yeah. Exactly. He can tell you, 'Yeah, you just like shoot it.' That's how you stand. But at the same time, you have that tacit knowledge that is embodied in individuals. And sometimes there's tacit knowledge that embodied in teams. Sports is a great example of that, because sometimes you see teams in which the players might individually not be that great, but they play collectively in such a coordinated way that they know exactly when another player is walking behind them so they can give them the pass, or fool the defensive maneuvers of an adversary team--that they are embodying knowledge at a collective level. And that is very much true in firms. Even, I see it, in my group and in my field[?] and in my startup, which, we develop this complex software project that are these large data utilization[?] engines, and by now we have a relatively well-defined division of knowledge in which I know that there are some guys that know about back end, some guys that know about front end, some guys that know about design. And that it would be very hard for me to accumulate that knowledge or even to transfer that knowledge to a lot of people without adding them slowly to the group. Which is the only way I am able to expand the number of people that have that capacity. Simply because there is tacit knowledge that is also embodied in groups of people. And firms are a great way for society to embody that tacit knowledge. Because they are made of overlapping generations in which people pass on some practices that can be qualified[?] but many of which cannot be qualified[?].
18:37Russ: So, in the 1990s--I think it was the 1990s--there was a big fad--at least I think it was a fad--in management education about trying to utilize the knowledge that was in a firm that was greater than the knowledge of any individual person in it. Or, to think about how to maintain the knowledge when there was turnover. So, if you were to leave the lab, you have lots of knowledge not just of your own insights and expertise: you have a huge amount of tacit knowledge about the capabilities of the people in the lab--their strengths, their weaknesses--that also can't really be written down and can't be preserved, I don't think. I think it's unique to the team and its constituents at any one time. Guest: Exactly. And that's a little bit of the theme of the book: How do we get to accumulate this vast amount of knowledge? And the idea is that we are able to accumulate this vast amount of knowledge, you know, because we have tricks that allow us to transcend or finite capacity to do those as units. So, if you think about the world, you know, 3 billion years ago--which is kind of like a time scale that is longer than the one that economists usually like to think about, but the [?] was here-- Russ: [?] Guest: Exactly. And it was populated by, you know, mostly unicellular organisms. And at that time, the type of things that those uni-cellular organisms could achieve were relatively primitive for today's standards. And the thing is that those unicellular organisms had a finite capacity to compute, because any system that is finite in the amount of matter and energy that they are able to embody is going to be finite in its capacity to compute, too. There are actually physical laws that tell you, given the amount of matter and the amount of energy available to a system, how much computation that system can complete. But that finiteness was not the end of the growth of order and information in our planet. At some point, those unicellular organisms figured out how to become multi-cellular. And those multicellular organisms were able to accumulate amounts of knowledge that were much larger than one of the unicellular organisms could ever achieve. By the same token as humans, we are also finite in our capacity to accumulate knowledge and knowhow. Actually, each one of us knows very few things; and when you talk to a person who comes from a different field, it's very easy to find areas of knowledge that you might know and that person doesn't, and vice versa. We know very few things. And the way that we transcend our finite capacity to accumulate knowledge is by accumulating that knowledge collectively. So, the fact that knowledge has to be embodied in units that are of finite capacity is what allows us to accumulate knowledge and knowhow and eventually to beget[?] information. But it's also what makes it difficult. Because it transforms the problem of knowledge accumulation to a problem of network creation. If you are not able to sustain the network that allows you to accumulate knowledge, you are going to eventually be stuck with a ceiling--at some point you are no longer going to be able to accumulate any more information simply because, you know, you have reached the capacity of the units. Russ: Yeah. Why don't you talk, before we go on, about the distinction between knowledge and know-how, which you mention in passing. Guest: So, knowledge and know-how is the language that I use to describe most simply the difference between explicit knowledge and tacit knowledge. So, I use the word 'knowledge' to refer to the knowledge that can be qualified[?]--the one that can be expressed in books, the one that we can talk about. Know-how is the shorthand that I use to talk about tacit knowledge, which is the knowledge that is embodied in an expert violinist, that might be able to tell you something about how to play the violin, but when push comes to shove, that's a type of knowledge that can only be accumulated by practicing in the right context, the activity that you want to do. And many of that knowledge sometimes can't be qualified at all. A lot of simple examples of that. For instance, you know how to see; but you probably don't know how you see. Because if you knew how you see, you should be able to create like some very sophisticated computer vision algorithms that would match what humans are able to do when processing visual input. So, the words 'knowledge' and 'know-how', I use them to refer to tacit knowledge and explicit knowledge. Know-how being the definition of tacit knowledge; and knowledge being the one I use for explicit knowledge.
23:17Russ: So, let's take the next leap. As you say, a lot of what we need to then, as human beings we need to cooperate; we need to work with other people. We need to leverage not just the information of others, not just the knowledge of others, not just the know-how of others; but we need to interact with that. It's very complicated, obviously; and your book spends a lot of time trying to get at that. So, what are the keys, as you see them, in making those networks possible? Guest: So, in that context, what I like about thinking of societies[?] and economies in terms of computation, is that by understanding that computation needs to be embodied in networks, we can start quickly incorporating ideas of institutions into our understanding of economies, in a very natural way. Because at the end of the day, institutions help us shape the ability that we have to connect with one another. One institution, for instance, that people in the social capital literature have focused a lot on is the idea of trust. Okay? In a society in which you would have high levels of trust, and people would not be very skeptical about one another and they wouldn't think that they are going to very easily betray them, they are going to be able to have a larger number of interactions because each one of these interactions has a lower cost. And it's going to allow them to create larger networks that can accumulate a larger computational capacity. In a society that the levels of trust are relatively low, you are going to have a society in which people can have very few links, because each link is very expensive. And consequentially, they are going to be able to accumulate a relatively small computational capacity. So, trust is one of those institutions. Obviously there are many others. You know, there are changes in sometimes transportation, communication technologies that change the cost of interactions. You know; there could be also some formal institutions, some certain laws, insurance schemes that can help us reduce the transaction costs that we have between people or reduce the risks of performing certain types of transactions that also help us catalyze the formation of these networks. So, that's, in summary, like how the networks would help us incorporate this idea of institutions. And how, by incorporating institutions into networks we can explain how societies with different institutions have different computational capacities. Russ: Well, we'll get to that in a sec. But I wanted to challenge something I thought was--I didn't see in the book and get your reaction. So, you don't make much of a distinction between innovation and production. Or creation and production. So, when I think of a small firm--or a large firm, even--there's a small number of people. They work together; they interact in all kinds of inexplicable and hard-to-quantify ways. But a great firm, a really innovative company, is doing exactly what you are talking about: there are people who come to trust one another; they share ideas; they stimulate each other's thinking. And they come up with fabulous products and fabulous ideas. But the production of those products is a different challenge; and it also involves networking. So, because we don't specialize--right? Typically there is not vertical integration of a product. So, a company, as you mention, like Apple, will produce its products all over the world. Different pieces come from all kinds of different places. It's not just the know-how, though, that comes from different places, or the knowledge. It's also the physical need to produce those in different places. And you didn't talk much about that. Or, do you think of that as just part of the whole process? Guest: So, I agree that the links[?] that you have with people to invent something or to scale the distribution of something that you have invented already are quite different. You know? The links[?] that you would use to invent something tend to be more intimate; they have to be much stronger; they have to have a high level of trust because people need to actually challenge each other and in ways that would allow them to do those creative activities. But at some point, you know, when the product that is being made has been figured out, you can scale that to a different type of network in which you might have a lot of repeated type of knowledge and know-how. For instance, you know, you can have like a large sales team in which many people more or less know about the same; but you need a large number of them--not because you are accumulating a vast amount of knowledge, but simply because you are trying to have connections to a larger number of people in the context of a commercial operation. Russ: The world is a big place. Guest: Exactly. Or, by the same token, when you are manufacturing a product that you have already figured out how to manufacture or how it looks, like you have already designed it, to scale that manufacturing you are going to need to include more people that are doing exactly the same as other people in that firm. So, I do think that there is a strong distinction between the creative links that you use to invent. You know, I like to think of invention as very different from innovation. Invention is when you create something new. Innovation is when you improve something that already exists. And then the links that you would use to scale a certain operation, like the ones that you would use to bring manufacturing to scale or to have a big sales effort by having a large sales team. Russ: I guess the other thought I had is that--I'm just--I'm sort of challenging the social network part of this, at least for the creation of knowledge and creation of know-how. It's usually a small group of people who create a new product. They do rely on immense numbers of people who have come before them. But that kind of knowledge that they are using is often passed on in the form of books, in the form of teaching, in the form of going to graduate school or other ways. You don't have to trust those people. That knowledge is just--you can grab that. The really hard part of the people is the knowledge that is created through the face-to-face interactions. And that's a relatively small network, it seems to me. Guest: Yeah, exactly. It's a relatively small network. But it's a crucial network, because it's the one that recombines all of that codified[?qualified?] knowledge, and then transforms it into something new. But at the same time, the capacity of people to have intimate connections is quite finite, just because of all these time constraints. If you think about it, how many people you can actually have a meaningful meeting on a given day? Maybe 6, 8 meetings a day; and you are super-tired at the end of the day. So, despite the fact that those networks are relatively small, they eat most of the banquet[?] of an entrepreneur. The entrepreneur is focusing on participating on the creative process. So, the fact that we can qualify knowledge and we can transmit it sometimes in books, sometimes through education, and sometimes through objects--because the guys that I build software with, they don't build the hardware that we use to embody our software. But they access, you know, the knowledge--not the knowledge needed to create the hardware, but the practical uses of the knowledge that went into the creating of the hardware to those objects. So, I wouldn't feel like any contradiction: the fact that the networks are small--I don't see it as something that would in some way contradict what we are have been talking about before, but rather that the capacity that we have to create, you know, meaningful networks in which we can be creative is quite finite. We have, like important constraints that come from the amount of time that we have, the speed at which we communicate with one another, you know, and the length that it takes to form a meaningful relationship with another person. Which is also something that is not easy.
30:50Russ: I guess I brought it up because I think about Adam Smith, and I think about growth from the levels of, say, 1500, 1400--50, 80, 120 years in the Common Era. There's been a very flat time, and then we had an enormous explosion over the last 200, 250. And a lot of that comes through, I would suggest, non-planned, undesigned cooperation. The social trust--a lot of that is face-to-face, and it's planned. A group of people get together in the modern world to do something. But a lot of our cooperation, I suspect, that's important, takes place mitigated through the price system, that's not intended; that just emerges as, you know, supply chains and other forms. Guest: Yeah; so I think, you know--and here, I challenge you back. Russ: Go ahead. Yeah. That's the goal. Guest: Exactly. In some way, I understand that it's very attractive to think of a world in which everybody more or less is driven by self-interest, left to their own devices: They buy inputs through the price system; they recombine them, and they sell a certain output, you know, that also goes into markets and the price system. And it's attractive, you know, from, I think that's standard economic theory, to think of [?] the world that way. But most people, for a fact, they don't work by themselves. So, first of all, all of these units that are getting inputs in and they are producing outputs, they tend to already be collaborative. And they--you know, even a bakery might have, you know, a couple of bakers; they might have, you know, someone that is sitting at the register. If you have a larger operation like a car manufacturing plant, they might buy inputs and they might sell outputs. But internally, those networks need to exist. So, on the one hand the fact that you have a market doesn't preclude that you have other instances in which you need to have more intimate and other formal collaboration that is not driven by prices. That's a little bit of what motivated Ronald Coase to come up with his theory--you know-- Russ: Yep, couldn't agree more. Guest: Coase's theory of the firm in which Coase, in the 1930s goes to a presentation from a professor in the department of commerce in LSE (London School of Economics) and this professor says, 'Well, the normal economic system works itself out' and Coase, as he said in his Nobel Prize acceptance speech, thought that that was total BS because he thought that in reality the world was made of a lot of islands of conscious power, that were firms, that work in some form of, you know, centralized planning, which the price system was not playing such an important role. The price system played a role between firms. But within the firm, that important heritage or that important networks that are formed that had to make decisions not based on the supply and demand that is signaled by price systems but based on more social mechanisms that might involve that knowledge that each person has about the capacities about someone else, the way that they will react to other things, you know, and ultimately those units are fundamental. Because if you would destroy each one of the firms and we would go back to an economy in which everybody would be by themselves, getting inputs and trying to build outputs, I'm pretty sure that our economy would be probably less sophisticated than that is in the Mad Max movies--in which people are more or less by themselves in the desert. So, the price system--I agree[?] it helps reveal information about supply/demand things. But there is much more to the economy than that. It's not just about making commercial transactions. What makes an economy work at the end is that we figure out things that are worth transacting. And that's a part, you know, of the economy that I haven't seen that much explaining the theories in which the things that get produced, more or less are assumed to be there because they are either widgets or they are, you know, some sort of object that we assume that it's there. And the focus has been most on the commercial transactions among those things that exist, rather than on the interactions that have to happen among people to bring things into existence. Russ: Well, first of all--I think you are right as a general criticism of at least textbook economics. It is one of the stranger--especially the theory of the firm--it is one of the stranger aspects of the way we teach economics: that the biggest decision a firm has to make is how much to produce. Which is the last thing in the real world that firms worry about. In the real world, firms have to figure out what market they want to be in. Whereas in the textbook, they take the market they are in as given; they take the price as given, often. But in the real world, they have to figure out what market they are in, which is going to affect what they can charge; it's going to affect how they produce and what they produce. So I totally agree with you. But I think it's a bit of a straw man. Adam Smith wrote two books. One is about how we interact with each other--that The Theory of Moral Sentiments-- Guest: Yep. Russ: when we are face to face. The other is the book about markets, which is the Wealth of Nations. They are both important. Couldn't agree more. And I couldn't agree more that Coase was a challenge at least to the textbook version of--although I think you might have used a different word than the one you suggested when he was seeing the traditional theory. So, I totally agree with you as a bottom line about a richer approach is very important. But I do think that it's important to think about the role of cooperation that occurs across space, and across time--that prices allows. Because we can't love everyone. We can't be face to face with everyone. As you say, we have a limited amount of time; we have a limited amount of resources. And it seems that prices play a key role in allowing those things to be overcome a little bit. Guest: No--I would agree; I would agree 100%; and the only thing that I would say is that since that had been say--you know, like Hayek said it, and basically it gets regurgitated even nowadays in every common section of every newspaper more or less when people are discussing the economy--since the role of prices in revealing information about supply and demand has been said, I tried to focus on aspects of information in the economy that had not been so much discussed in the literature. Because I wanted to make a contribution, even if it's a little bit controversial, by bringing in something that was a little bit different: which is the figuring out of what eventually you make, and why the things that we make are the essence of what makes our economy prospers, rather than just the commercial interactions that we have.
37:28Russ: So, let's move to that. You asked the following; you say, "Why is our ability to create refrigerators, jet engines, and memory devices concentrated in a few parts of the world? Why do many countries know how to make and export shoes, but only a few know how to make and export helicopters?" So, in that section of the book you are trying to explain what might be called, what used to be called comparative advantage. Sometimes, I think confusingly called competitive advantage. But it's clearly an ongoing key part of our lives, the commercial part of our lives. What's your answer to those questions? Guest: So, one thing that is a basic observation is that in some way our ability to produce products have what an ecology[?] would call a geographic range. Yeah? There are certain parts of the world in which that ability exists; and there are other parts of the world in which that ability does not exist. Okay? So, from the perspective of the theory that I advance in why information rose, the term is the existence of that ability first and foremost is the knowledge and know-how that is embodied, you know, in the networks of people that reside in a place. And that ultimately is what makes that knowledge very hard to transfer. Because, let's say, you take Hollywood, and you take all of the knowledge and knowhow that is there about producing, like, blockbuster movies. That knowledge obviously is not stored in a single individual likes James Cameron or Steven Spielberg. But it's stored in large networks of people that get to collaborate, you know, through markets and through social relationships to create, you know, those films. And ultimately, there are very few parts of the world that can imitate that, because accumulated that knowledge is very hard because you are going to have to accumulate it in a network because of the simple reason that it is too much knowledge to fit on a relatively small group of individuals. So, going out from like a very Ricardian old theory of comparative advantage in which, you know, there would be certain factors that a population is endowed with, you know, here, the idea of the factors, which is a more modern way of looking at it but this is also present in not only my book but other works in economics is that this comparative advantage is much more dynamic because they hinge primarily on the computational capacity of social systems. And when you are looking at different countries or different regions or different cities producing different mix of products, what you are looking indirectly is at the knowledge and know-how that has been accumulated in the networks that populate those countries, cities, and regions. Russ: Yeah, well, I like your contrast, Ricardian view, because if you weren't careful you might think that the reason great movies come out of California is because they have mountains in the distance and lots of pollution. But then it would be easy to make movies in Santiago, Chile-- Guest: Exactly-- Russ: which--I haven't been there since 1981, but it used to remind me a little bit of LA (Los Angeles). But, as you point out, what makes Los Angeles extraordinary is the people: not the mountains, not the water, not the ocean. Those help; they make things pleasant for people to live there. But once they are there, there's an incredible power to those networks. And they are networks of people. Not natural resources. Nothing inherent about the geography. Guest: Yeah. And those networks of people are just like the hardware, with ability to make things need to be embodied. Because all of our knowledge, know-how, all we need is to be embodied, just like information is to be physically embodied on DNA or photons or books or brains. Knowledge and knowhow needs to be embodied on people and networks of people. And, you know, that's what, you know, LA has. And that's what San Francisco has, you know, in the tech sector, that's what Boston has in robotics. That's what New York has in finance and, you know, publishing. So, these networks that ultimately, you know, the hardware where we ultimately accumulate our capacity to make things and are both the enabler of our ability to create things that make our life prosperous, but also the curse of our inability of moving that knowledge and know-how around. Because the knowledge is trapped in the networks that embody it. Russ: Yeah. So, when you talk about transactions costs, we had a recent episode with Adam Davidson talking about the making of a Hollywood movie, and one of the things that struck him was that when you arrived on the set, nobody was bossing anybody around, because everybody knew their job. There were 250 people who had done this many times before, and they just knew what had to get done. And what an incredible advantage that is in terms of transactions costs; and, coordination--it's an incredible thing. So then the question is: Why is it that some countries excel at this coordination and trust and networking, and others struggle? And how do they--let's think about any lessons we might have for development and helping increase the standard of living in other places. It's obvious that you can't take a poor country and say, 'Why don't you become good at making movies? That way you'll be rich--because those are good-paying jobs.' And I think a lot of people make that mistake. But then the question is: Okay, that doesn't work. What might work? Guest: So, the way that I see economies as very historical[?] systems. Which is a little bit different than the more neoclassical ways of looking at economies, that look at them as a-historical systems in which a set of incentives would ultimately [?] the system, and the system finds a configuration that matches those incentives and figures everything out. The way that I see it is that they are very historical systems, so when you see an economy that has a capacity to produce a certain good, it's not that it was developed overnight through like some sort of incentive structure; it's that it was developed over a long period of time in which people accumulated in the context of a network all of the knowledge and know-how that they need to do that. So, the way that I see these historical processes is a way in which the mix of activities that you do actually helps shape a lot of important economic institutions. Recession, for instance, we finished a paper that we published as a working paper and we are now sending to submission in which we find a very strong correlation between the mix of products that a country makes and the income inequality that they have. You know? And the way that I interpret that relationship is that in some way, the economic activities that a society performs has to shape the type of institutions that you have, because ultimately people don't learn how to interact with others in sort of like the vacuum of their national identity, but they learn how to interact with others by going to work every day. And if they are going to work every day at a software firm in Silicon Valley where the structure is very flat and people are graded but you have to basically be very direct when it comes to feedback, you are going to have a different set of norms and expectations. And you are going to have to [?] a different set of institutions that, if you are going every day to a mining operation in which the structure is very hierarchical, it's all around safety and it's top down and basically contributing your creative input is not encouraged. So, different countries have had a historical different practice structures. These practice structures have helped shape the institutions that they have. These institutions [?] they have helped shape the mix of products that they are able to make in the future. I think this is always like co-evolutionary[?] process. And then eventually those structures can only adapt ever so slowly. You know? And that's what we observe in the data. If you look at the data, we see that if you know the mix of products that a country makes and you try to make a machine-learning algorithm that would predict what they are going to be making in 20 years from now, the predictions are very accurate because fruit doesn't fall that far from the tree. Countries that are producing a certain mix of products usually end up being successful in the future making products that are rather similar, and that tells you that there are a lot of things that go into the production of a product that can only be accumulated very slowly, and they can only be accumulated to things that are relatively similar to the ones that you already know.
45:36Russ: So, let me use an example from the book. A lot of manufacturing in the last 25 years has been done in China, for the world. China is the world's factory. Somewhat less so, maybe, than it has in the past in terms of as an employer because productivity is growing in the factories there just as they are growing everywhere and that means fewer people are needed to make any level of production. But if you think about why that's true--and you give an answer. So, give your answer, to the book: Why did China become a centerpiece of industrial production? And as a result, improved their standard of living quite dramatically over the last 25 years? What do you think--how do you relate the concepts of your book to that story? Guest: So, the way that I think about the particular case of China, which in some sense has been a little bit puzzling, we can say-- Russ: yep-- Guest: is that on the one hand, when you think about institutions and the role that institutions have on shaping the capacity of an economy, I think institutions can act better as a brake than as an accelerator. So, if you have a country that has a certain capacity, you can slow it down significantly or you can even destroy this capacity through[?] institutions, but you cannot jumpstart the development of those capacities very easily by having good institutions. China is a country that historically has always been very productive, creative, has a rich history and has not been devoid of inventiveness or the generation of cumulative culture; and it had a very rough transition in the 20th century because China didn't go through a period of Renaissance like Europe did but moved more from a medieval society to a modern society over a period of only a hundred years, maybe. But even if you go back to the 1950 and 1960 when China was the poorest country in the world, they had a leap that had 80 million people, which is the size of the population of Germany, that was sophisticated enough to have been able to produce atomic bombs and to produce relatively sophisticated things internally. So, I think China is a country that in some ways has had knowledge for a long period of time; it has a lot of capacity and as the institutions became more inclusive of the exportation[?] of that knowledge and know-how in productive activities, you know, China was a country that was seen to grow. But other countries are not in the same position because they might have the bad institutions that, once they are removed, they are not liberating a population that has accumulated a relatively good capacity to make things. So, for instance, when we look at our data, when we look at economic complexity in this which is this formula that I created to estimate the computational[?] capacity of a country, even if you go back to the 1960s, China was above average in the world, despite being the poorest country in the world. So, it's telling you that the ability to make things was never that bad, and that's why it's not that surprising that mainland China has followed the footsteps of other Chinese-based economies, like Taiwan, Hong Kong, and Singapore. Russ: But it's interesting, because so much of their productivity it seems is coming from the crystallized imagination of other people--a factory, an American corporation, an international corporation locating a factory in China versus somewhere else. Once they've put it in China, those workers are suddenly very productive. But they are the same workers. They didn't literally get smarter. They don't have any more knowledge; they don't have any more know-how. But they are augmented by those machines. So they-- Guest: Yeah. But, like, if that were true--Ethiopia has 80 million people; Nigeria a hundred million--we should be able to do that there as well. So, I think that thinking that in China maybe there was little capacity-- Russ: Maybe we can. Guest: I don't think so. Russ: I'm not sure. It's a good question. So, that's your question for me: Why can't we? And my answer would be, maybe we could if the governments and institutions of Nigeria and Ethiopia were a little more open to trust. Guest: So the way that I think is that China is a country that has been relatively high in capacities, you know, has had like a long tradition of history, art, sciences, you know, and that it's different from that of sub-Saharan Africa or Latin America in that case. You know? And the story that people tell in the West is that China basically is like copying and imitating everything and they get very little credit for they actually have achieved. But in my view and according to the data that I have, which is global data on international trade interaction[?], China is a country that always has been relatively sophisticated in its ability to make things; it has actually increased over the last 40, 50 years; but not as much an increase from like being in the 40th number of the rank to being like 20th in the rank that we have of economic complexity. And what explains for me the growth of China is not low prices, but the relatively low price of high capacity. Okay? Russ: Agreed. Guest: Because one thing is to be able to pay a person $2000 a year, you know, and asking it to be able to pay $2000 a year when that person is able to produce something that is equivalent to what is produced in a country that people are making $30,000 a year. Russ: I totally agree. I think again, I think again, I think the challenge is: Is that an institutions problem? Is it a people problem? Is it a governance? Is it a private property, rule of law problem? There's a lot of things going on there. But the part I totally agree with--you have a great sentence here. You say, "Economic development is not the ability to buy, but the ability to make." Guest: Exactly. Russ: And that has to be the starting point. And I think that's very hard for people to accept. Guest: Yeah. So, I just published today a story in the Harvard Business Review that makes that point. Did you ever hear about Oswald the Lucky Rabbit? Russ: No, I missed that. Tell me. Guest: Yeah. So, Oswald the Lucky Rabbit, in the 1920s was more famous than Mickey Mouse. So, he was a character that was developed by Disney; and Iwerks--Ubbe Iwerks was kind of like the Steve Wozniak of Disney being the Steve Jobs. He was actually the guy that made the drawings and was a great designer; and he created this Oswald the Lucky Rabbit character that became very popular in the 1920s but was owned by Universal Studios. So, when Oswald became very popular, Universal Studios threatened Disney to approach all of their animators if they didn't reduce their production costs. Okay? Why? Because basically Universal [?] saw Oswald as what was valuable. You had what you could sell. What Disney and Iwerks did: on the train back from New York, they created a new character. That new character was Mickey Mouse. And obviously Mickey Mouse left Oswald the Rabbit in the dust relatively quickly. Because ultimately what was valuable was not what you had made, but your ability to make it. And time and time again, if you look at the history of firms and the history of creative activities, you'll find that the people who are doing creative activities, they get ripped off, once, twice, three times. But they are always able to bounce back. Because you can take property away from someone, but you cannot take that tacit knowledge and know-how away from someone. As long as the person has that ability to create something new, they are going to be able to create something better than what they did before, because they are even more experienced. And they learn from the bad experiences as well. So, I completely subscribe to the idea that what is most valuable in an economy is your capacity to make. It's not the book that you wrote, but your ability write books. Not the film that you made, but your ability to create great movies. Russ: Yeah. You can steal my book, but you can't steal my brain. Right? Guest: Exactly. Russ: It's an important--it's a very important point, actually. It sounds like a tautology or just a cliche, but it seems to me it is very important.
53:47Russ: A lot of your books about the augmenting of human potential with other people's crystallized information, other people's knowledge: Some people are very afraid of that because of what's coming with artificial intelligence. Where do you stand on that? What are your thoughts on what's going to be "left for people to do" after we figure out ways for robots to do everything? Guest: So, I think that like some short term and long term views on that, I think are important to discuss. And the question is, if I [?], it requires me to do a little bit of a value judgment. Yes? Because in some ways it's a question that the answer is going to be good or bad, not yes or no. So, in that context I have a very simple modal axiom that I subscribe to. And I have very few of those. And in this case, it is that you should never morally judge people from a different time period than yours. Because you are not prepared to do so. So, for instance, I wouldn't want a person from the 1500s to judge my lifestyle, because probably they will find it deeply immoral. If in the moral code that they had at that time, I am an atheist[?], I had sex before I got married; I did a lot of things that would be morally reprehensible according to the moral standards of the 1500s. By the same token, we might have grandchildren that are going to start driving[?dragging?] U.S.[?] deporting to their brains directly, you know, to augment themselves in ways that might feel scary to us. But firstly I think that's like Puritans judging people from San Francisco in the year 2015. We haven't made that moral judgment; and in that case, I have nothing but trust in that society knows better than I do. And if that's the way that we're going to go and that's the way that our economy is going to be in the future, in which we are going to embody ourselves in electrons rather than in, you know, like complete carbon atoms like we do right now, that's something that I'm willing to let them decide. You know? Russ: So, you're an optimist, more or less. You think it's going to be okay. Guest: Yeah; I think that eventually we will figure things out, and that doesn't mean that we find problems in between. But I'm a little bit of an optimist. I do think for example my friend Steven Pinker, he always treats, you know, that optimism that in some sense comes from his book, The Better Angels of Our Nature, that society despite all it's problems, has been getting better. There is less violence now than it used to be. And always there is a lot of horrible things that are still happening in the world, but all in all, as we move forward we have been able to become better at many things. Where, for example now, we are much more environmentally conscious than we were 30, 40 years ago. That's a type of progress. Like, Europe had an extended period of peace. Despite all of the economic problems they might had during the last decade. It has not--the idea of a war between France and Germany, which was a recurrent thing, or a war between Spain and England or whatever, a recurrent thing in history--is not something that I would say is on the horizon. So I think that we figure out how to improve our society collectively; and that over the long time scale, the arc of time bends toward things that are positive. Because the things that are negative, they tend to be a little bit more transient. While the things that are positive tend to grow more slowly, but be long-lasting. Russ: It's interesting. I'm not sure I am as optimistic as you are. I'd like to be, though. So, I'll just say that. Could be true.
57:36Russ: What is next in the research on these kind of topics? We are almost out of time, and we didn't get to talk about the empirical side of this. You do a lot of interesting visualizations and measurement of complexity of economies. I didn't fully understand them and I look forward to reading more about them; and people can read about them in the book or in papers that you've written with colleagues. But even at the level at complexity that you are looking at, there is still a bit of a black box aspect of this. Just as there is in the standard economic approach: We don't really have a very good understanding of what's happening below the surface. And I suspect that's true of your computations as well. Do you want to try to make them more refined? Do you hope to--what do you see as coming next in terms of thinking about these--the growth of knowledge and information? Guest: Yeah. So, I would answer this with like two things that are now my research agenda. On the one hand, definitely I want to develop better metrics of economic complexity, that, on the one hand might be more transparent and better theoretically founded, and on the other hand, they are empirically better at predicting the aspects of the economy that we have found as measures which are good at predicting, which is future economic growth, income inequality, and the mix of products that a country is going to be able to make next. You know? On the other hand, one of the things that I've--one of the errors that I've gone through during the last year is I started to think about the world not only in terms of the mix of products that people make, but also in terms of the type of messages and cultural products that we produce. So, a couple of years ago, we released a project that is similar to the [?] of [?] complexity but that is [?], that looks at the data from more than 11,000 biographies of globally famous people who tried to map out how our collective memory has changed over time. Because I become convinced that we want to go beyond our understanding of economic growth that comes from thinking of measures of consumption like GDP (Gross Domestic Product). We have to look at our ability to compute and our ability to accumulate information. And we find like very interesting things already in that new data set that focuses on like the information that we have recorded in our history. For instance, we find that there are important discontinuous transitions in the rate in which we remember people that occur when you have changes in communication technologies. So, we have a paper now that we are finishing in which, if you look at the period before the printing press, 2000 years between the printing press and if you look at the number of globally famous biographies that we find in two comprehensive data sets and you divide that by the population of the world at a given point in time, you find that the fraction of people that we remember is a constant number, is a constant fraction of the global population. So, if you look at the number of people that we remember from the year 200 B.C., you know, it's the same fraction of the global population that the people we remember from the year 1200. That number didn't change. And when the printing press came along, that number jumped; and it became constant again for like 300 years, until we have changes in the format of printing--the dimension of journals, you know, then eventually we have mechanized printing. Then, collective memory accelerates. We find also that who you remember has changed enormously, as you change communication technologies. For instance, before, you know, the invention of printing, most of the people in the biographical records that we have are either political leaders or religious leaders--they are kings, emperors, and prophets. After the printing press, you have a new elite. History is not reserved to, you know, the leaders of the institutions any more, but to this creative class that also includes artists and scientists. Then you have the introduction of [?] and radio, and you generate an elite of performers that did not exist before. So, one of the things that I've been interested in is to understand how what are related to recorded, transit[?] of information changes over long periods of time. Because I do have [?] GDP might have been growing only for the last 200 years or so, but you know, our ability to process information has been growing for longer periods of time, even though it may not have been expressed in increases in GDP per capita.