Why Information Grows: The Evolution of Order, from Atoms to Economies by César A. Hidalgo (PDF)

Description

“Hidalgo has made a bold attempt to synthesize a large body of cutting-edge work into a readable, slender volume. This is the future of growth theory.” — Financial Times What is economic growth? And why, historically, has it occurred in only a few places? Previous efforts to answer these questions have focused on institutions, geography, finances, and psychology. But according to MIT’s antidisciplinarian Cér Hidalgo, understanding the nature of economic growth demands transcending the social sciences and including the natural sciences of information, networks, and complexity. To understand the growth of economies, Hidalgo argues, we first need to understand the growth of order. At first glance, the universe seems hostile to order. Thermodynamics dictates that over time, order-or information-disappears. Whispers vanish in the wind just like the beauty of swirling cigarette smoke collapses into disorderly clouds. But thermodynamics also has loopholes that promote the growth of information in pockets. Although cities are all pockets where information grows, they are not all the same. For every Silicon Valley, Tokyo, and Paris, there are dozens of places with economies that accomplish little more than pulling rocks out of the ground. So, why does the US economy outstrip Brazil’s, and Brazil’s that of Chad? Why did the technology corridor along Boston’s Route 128 languish while Silicon Valley blossomed? In each case, the key is how people, firms, and the networks they form make use of information. Seen from Hidalgo’s vantage, economies become distributed computers, made of networks of people, and the problem of economic development becomes the problem of making these computers more powerful. By uncovering the mechanisms that enable the growth of information in nature and society, Why Information Grows lays bear the origins of physical order and economic growth. Situated at the nexus of information theory, physics, sociology, and economics, this book propounds a new theory of how economies can do not just more things, but more interesting things.

User’s Reviews

Reviews from Amazon users which were colected at the time this book was published on the website:

⭐This is a very ambitious book that unfortunately is marred by some significant flaws. Its goal is to unite economics with biology and physics, through the common thread of information theory. But along the way it makes some excessive claims, particularly about “order” and humanity’s uniqueness in the cosmos, that undermine the logical coherence of the argument. The author (CH) also adopts an unusually narrow definition of an economy, excluding not only production of goods for domestic consumption but the entire services sector, even though the latter accounts the majority of GDP in most developed countries. The book also presents a graph that seems to conflate exchange rate differences with economic growth (unlike the peer-reviewed paper on which it’s based); see the discussion of my correspondence with CH on this point in 2(c) below. Finally, and most disturbingly, while celebrating the production of ever more complex physical products as the fulfilment of a natural law of the universe, the book is entirely oblivious to the environmental impact of making so much stuff — a reckless omission for a book about physics and economics written in the early 21st Century. [NOTE ADDED 2017 March 22: the criticisms and page references below are also pertinent to the paperback edition.]In the long review that follows, I’ll first discuss some problems relating to the discussion of physics and information theory (section 1), then focus on some issues relating to economics (section 2), before discussing the book’s synthesis of these topics (section 3) and concluding with some observations about style (section 4).1. PHYSICS ISSUESThe book opens with a dramatic discussion of how Ludwig Boltzmann’s formula for the statistical mechanical (SM) entropy is “identical” to Claude Shannon’s formula for the information theory (IT) entropy of encoding a message (@xvii, 14). Shannon, CH says, equated “information” with “entropy” (@17). CH overlooks that Shannon called the same expression the “uncertainty” when it comes to *decoding* the message — our usual position vis-à-vis the information content of nature’s messages. This notion of “uncertainty” leads many writers to compare the IT entropy to *missing* information (e.g., physicist Arieh Ben-Naim, who wants to ditch the word ‘entropy’ altogether).More common, though, is to confuse entropy with “disorder.” CH correctly points out that error (@17), but while doing so seems to have fallen into a new one. Throughout the book, he uses ‘information’ and ‘physical order’ to mean the same thing, often using both in apposition within the same sentence (e.g., @ix, xiv, xx, 5, 7, 8, 12, 18, 19, 23, 24, 184, 196, to name a few). Noting a “tradition” among “the general public” of “equating information with something more than bits” (@18), he seems to reason as follows: disorder ~ SM entropy ~ IT entropy ~ _missing_ information => information ~ physical order.I’ve checked stacks of books and articles, but so far I’ve yet to find another physicist who endorses this correspondence. The presence of ‘order’ is very context-dependent. Consider a 1024×1024 screen with about half black pixels and half white pixels. Suppose it’s organized into 4 stripes of even width and alternating color: that’s highly ordered, with low IT entropy. Now suppose it looks more like the “snow” on a TV screen — but suppose I’ve painstakingly encoded a message into the million pixels, one at a time. In that case, it’s highly ordered, with high IT entropy. Finally, suppose I based it on a snapshot of a particular instant of real TV screen snow: in that case, the pattern is random, i.e. *disordered,* with high IT entropy. Maybe you could distinguish the ordered, message version from the random version using a tool called “mutual information” (never discussed by CH), but the point is: order can have high or low IT entropy, and high IT entropy can be associated with both order and disorder. Regardless of whether you use Shannon’s or CH’s interpretation of “information,” there isn’t a monotonic association between information and order. (Although he is focused on the meaning of “complexity,” Peter Grassberger makes a similar point in discussing Fig. 3 of his 2012 paper, “Randomness Information Complexity,” on the arXiv.) Moreover, many physicists regard new information as coming from such events as quantum fluctuations and quantum decoherence — events that seem to me to be more in the “disorder” camp; see, e.g., the article by Seth Lloyd in “Complexity and the Arrow of Time” (Cambridge UP 2013). CH never mentions either of these types of event.CH invokes Nobel laureate Ilya Prigogine as the inspiration for several claims, among them: (I) “[I]nformation emerges naturally in the steady states of physical systems that are out of equilibrium” (@28), (II) the whirlpool that forms over your bathtub is drain is an “information-rich state …that we get for free in an out-of-equilibrium system” (@29), and (III) “[T]here is no past, and no future, but only a present that is being calculated at every instant,” a view CH calls the “instantaneous universe of Prigogine” (@40). Prigogine is a divisive figure among physicists, and also among the broader community of researchers studying complexity; but as a dilettante in that field I’m probably more sympathetic to his views than not. Unfortunately, despite my good will, when I looked through four books by Prigogine I found evidence only for statement (I). The closest I found regarding (III) appeared to be a *rejection* of Henri Bergson’s “instantaneous” view of time (Prigogine & Stengers @92). Regarding (II), CH claims “information-rich” refers to states with “correlations that give ‘information’ its usual meaning” (@18), though in (II) CH evidently means the same sort of information as used in (I). I consulted with a nanotech theoretical physicist who had recently published a paper with “information-rich” in the title, who suggested its use in the bathtub context arose from a confusion about information and entropy. Cf. Cross & Greenside’s textbook on non-equilibrium pattern formation (CUP 2009), which doesn’t invoke information at all; ditto for the Sinha & Sridhar text on patterns in excitable media (CRC 2015). Still, for the sake of argument, I’m content to give CH the benefit of the doubt on this, but since statements (I) and (II) play a prominent role in his synthesis, I postpone further discussion of them until Section 3.2. ECONOMICS ISSUESa. What is the economy?:CH’s characterization of the economy is limited to the conception and production of solid objects. His discussion of “economic complexity” is further restricted to sales of such objects (and their physical inputs) in market transactions in international trade (e.g. @154-155). To say that’s a limited view of the economy is an understatement. It excludes services, which have accounted for >50% of the GDP of the US since the 1950s and of Japan since the 1960s, to mention but two examples. It excludes domestic production of goods, which may incorporate considerable knowledge and knowhow. For example, Japan has the world’s largest number of companies over 100 and 1,000 years old. Many of these are providers of local goods and services, such as foodstuffs like yuzumiso, products like the nose cones for Japan’s bullet trains (which are shaped *by hand*), and services such as repairing ancient religious architecture; the skill required for each of these is tremendous.There are many types of knowledge and knowhow that aren’t involved in market transactions, too. But while the book’s long discussion of “social capital” (Ch. 8) veers close to that topic, it focuses resolutely on economic impact. This myopia leads an expression of surprise at discovering society: “Yet there are also good reasons to believe that social networks predate modern economic activity. … As [a] team of researchers puts it, ‘Certain elements of social network structure may have been present in an early point in human history’ — a point that, of course, might be earlier than the emergence of modern markets” (@117). Boy, had I been messed up: I’d thought it was small bands of intrepid bond traders who were the first to kill wooly mammoths for food and protect their families from hungry sabre-tooths.I stray, though, in mentioning bond traders, because in this book’s view the financial services industry doesn’t exist. And anyway, there isn’t any need to worry about those members of the 1%, because they aren’t actually rich: “Our ability to crystallize imagination teaches us an important lesson about the complexity of economies: markets do not make us richer but *wiser* , since they produce wealth as long as they give us indirect access to the practical uses of the knowledge and imagination that our species has been able to accumulate” (@69, emphasis in original).” How broad is the meaning of “indirect”? Broad enough to take in short trading, front running, and outright market manipulation?And is *producing* wealth all that markets do? Might they not have a role in *transferring* it, e.g. via what the classical economists called rents? Whether it’s because of his own ideological commitment or simply because he’s too focused on his field of research, CH’s narrative effectively masks not only most of the economy but the economy’s impact on society, and risks turning the book into a sketch for a neoliberal Theory of Everything.b. Aggregates and advantages:Turning to more technical issues, CH does make a worthwhile point in championing disaggregation of types of production, updating the work of Wassily Leontief by using network and graph theoretical techniques. That aspect of his work seems both interesting and solid. Nonetheless, he swallows whole the notion of using production functions, never so much as mentioning the critique of them by Piero Sraffa et al. in the “Cambridge controversy” between U. Cambridge economists and MIT economists (a controversy that MIT economist Robert Solow, having outlived his antagonists, has for decades simply asserted that he won).It’s ironic, though, that CH offers us long and very reasonable endnote (@200n12) deriding the circularity of “revealed preferences” — the neoclassical notion, developed by MIT economist Paul Samuelson, that the fact that we buy something means that we really wanted to buy it. He omits to mention that his own published research relies on something called “revealed comparative advantage,” a term that suggests that because a country sells a disproportionately bigger amount of a product than other countries it’s got some Ricardian advantage in selling it. In the real world, there can be many reasons why we buy stuff from one country and not another — oil from the Saudis and not from Iran, Cold War era goods from Japan and not from the Soviets — that don’t have anything to do with quality, productivity, etc. Similarly, despite the fact that the data he presents seems to start in the 1980s, CH never once reflects on whether peculiar institutional conditions promoting international trade since that decade — e.g., WTO, the IMF, the World Bank, etc. — might render his results to be less representative of a force of nature and more representative of a particular socially- and politically-constructed form of capitalism in our present era.c. Predictive accuracy of “economic complexity” metric:In Figure 15 (@160), CH presents evidence that the metric of “economic complexity” he published in 2009 with Ricardo Hausmann (Proceedings of the National Academy of Sciences 106:10570-10575, available online for free, along with supporting material) predicts future economic growth accurately, albeit on time scales of ‘ 5 years. The graph shows a scatterplot of annualised GDP per capita growth rates for various countries from 1985 to 2000 (i.e., the calculated annual rate that would take you from GDP per capita in 1985 to GDP per capita in 2000), with the X-axis values being the rate predicted by CH’s metrics and the Y-axis being the observed rates.Look at the book’s graph and you’ll immediately see something striking: there are about 15 countries where the supposedly observed annualized rate was above 8%, and in some cases above 10%. Reality check: 8% annual growth of GDPPC for 15 years would more than *triple* it, and a 10% rate would more than *quadruple* it. One of these supposed super-growers is Japan, which for most of that historical period was famously experiencing its low-growth “Lost Decade,” so this data struck me as downright bizarre.I followed a link from the American Economics Association and checked the Conference Board’s Total Economy Database. Using PPP (purchasing power parity) GDP per capita data, I computed the annualized endpoint-to-endpoint rates over that interval for the 15 countries involved. Here’s what I got: Chile 4.98%, China 5.56%, Dominican Republic 3.29%, Hong Kong 3.54X, Ireland 5.75%, Israel 2.60%X, Jamaica 1.28%X, Japan 1.95%X, Korea 6.70%, Portugal 3.45%X, Singapore 4.66%, Spain 3.25%X, Turkey 2.51%X, United Kingdom 2.40%X, Uruguay 2.74%X. *All* of these are way less than the 8% to 10+% alleged to have been observed, and those marked with an X — 9 out of the 15 — are also *way* less than the predictions of CH’s model as presented in the book. (Fig. 15 shows Japan’s predicted growth at around 8%, too.)Then I looked at the 2009 PNAS paper, which also uses PPP data. Figure 3F presents a test of the complexity metric for the interval 1985-2005 — but only one country has an observed GDP per capita annualized growth of over 8%, China. Another outlier was Korea at just under 6%; Japan is lost in a crowd of type just below the break-even line, but its predicted growth looks to have been between 2%-3%. All in all, this graph looked pretty good, and consistent with my own sample of PPP-based calculations.Thinking CH had made an unintended error by swapping in the wrong graph, I wrote to him to alert him to that. He kindly replied promptly, but defended the book’s graph, explaining he had changed the data set subsequent to the 2009 peer-reviewed paper. His new source was … Google, which uses current dollar value statistics from the World Bank World Development Indicators. He explained that “[t]his time series has much better coverage (it includes more countries for more years) so I chose to use that.” (I’ve modified this review to correct my prior misunderstanding about a possible inadvertent substitution of the graph.)When I checked Google, sure enough I found 7.24% annualized Japanese growth from 1985-2000. Since Google is always right, I might have ended my inquiries there, but I decided to check Japanese government statistics anyway. I couldn’t find inflation-corrected figures for the entire time interval, but using official GDP and population statistics, I arrived at 2.77% annualized growth in nominal GDP per capita from 1985-2000, and 2.31% for inflation-corrected GDP per capita from 1985-1998 (Google gave the mind-boggling result of 7.94% for that period). The government growth rates are slightly higher than the PPP-based rates (which are intended for international comparison), but still quite close.Digging deeper, the reason for the discrepancy appears to be simple: the yen-dollar exchange rate (XR). “Current dollar” values reflect XRs in effect at the time of the relevant data point; as this swings from year to year, GDP per capita expressed in this way will also be affected. If we use Japanese government data for population and nominal GDP, along with the annual average of the Federal Reserve daily XRs, we get results (4th Column) pretty close to Google/WB:YEAR—JPN-NOMGDPPC——-FRB ¥/$———-RESULT——-WB (Book)—-“Back Translation”1985……¥2,672,822…………238.471………$11,208.16…….$11,465.73……¥2,734,2441998……¥3,924,471…………130.818………$29,999.47…….$30,969.74……¥4,051,3022000……¥4,029,610…………107.821………$37,373.15…….$32,716.42……¥3,527,517Notice how the XR difference between 1985 and 2000 means Google/WB’s GDP per capita figure almost triples, while the government’s doesn’t even double. That XR difference works out to *an annualized rate of more than 5%,* meaning that most of the purported Japanese “economic growth” shown in this book’s Fig. 15 actually is due to exchange rates. WB apparently didn’t use FRB XR data, so I can’t be more precise. For the same reason, we can only get a very rough idea if we use the FRB data to “back-translate” the Google/WB data into yen. But doing so yields annualized growth rates in the 1.7%-3% range — in the ballpark for the rates from the Japanese government, my PPP results, and the PNAS 2009 graph. It seems to me that Fig. 15 simply forgot to factor out exchange rates at each data point. (Actually, this would be hard for so many currencies — a good reason to avoid constant-dollar stats.) If I’m right about that, then a good way to avoid this Macroeconomics 101 type of error is to use PPP-based time series, just as in the PNAS 2009 paper.Given that Japan was until recently the world’s #2 economy and remains #1 in complexity per Hidalgo & Hausmann’s metric, it’s surprising, to say the least, that neither CH nor any of the distinguished advance readers of the book who provided back cover quotes were familiar with recent Japanese economic history. Nor with the history of the UK, another of the 8% Google-growers: if growth were as spectacular as the book suggests, how in the world did John Major manage to lose the election to Labour in 1997? Nor did any of them notice that the shape and scatter of Fig. 15 are entirely different from the PNAS 2009 version, which would suggest some artifact in the new data set. I appreciate CH’s openness about his change in source. But I wonder if it might have been more prudent to stick with the PNAS graph after all.3. SYNTHESIS ISSUESHere I’ll focus on three particular discourses that connect “information” to economics.(a) Bugatti discourse:CH mentions that a Chilean paid $2.5 million for a Bugatti Veyron, and observes that of course no one would pay that for a Bugatti that’s been smashed against a wall. “This is another way of saying that the $2.5 million worth of value was stored not in the car’s atoms, but in the way those atoms were arranged. That arrangement is information. [¶] So the value of the Bugatti is connected to physical order, which is information …. [I]nformation is a measure of the minimum volume of communication required to uniquely specify a message. That is, it’s the number of bits we need to communicate an arrangement, like the arrangement of atoms that made the Bugatti.” (@12-13.) Later he refines this by specifying that there is an equivalence class of atomic arrangements that could have the same value, e.g., a Bugatti whose tires have been rotated (@20).For the moment, I’ll accept CH’s comment about information about atomic arrangements being the minimum volume of communication, etc. — i.e., information in its IT sense. Even so, the notion that the Bugatti’s value comes only from the arrangement of its atoms is nonsense.The value of the Bugatti comes not only from the arrangement of its atoms, but from social MEANING. Suppose a Bugatti Veyron has a lot of B’s and V’s on it. Now suppose those letters signified a Buick Verano: still $2.5M? Or suppose I tell a prospective buyer that the Bugatti was assembled in an area of Syria controlled at the time by ISIS. Will she still be so confident about the workmanship? Or suppose the car pops up in a society where no one has ever heard about Bugattis, and people use it as a planter for growing vegetables. Etc., etc.Aristotle teaches us that the prices of goods are arbitrary and established by social convention (Politics, Book I). The 19th Century neoclassicals, along with some of their intellectual forebears (like J.-B. Say), tell us the same: Léon Walras famously declared that the utility of the same phial of medicine may be greater to someone who plans to use it for murdering his family than to a doctor who plans to use it for saving a sick child. It’s not surprising that the book overlooks this, though: CH shows little sensitivity to the history of economic thought, as when he declares that around 1906 “[e]conomics did not need psychology” (@xiii). In fact, that was the absolute heyday of psychological utility, as in Walras, Edgeworth, Slutsky, Pareto and others.There’s also a more technical objection to this discourse: how can one communicate the positions of the *different types* of atoms in the Bugatti, distinguishing iron atoms from molybdenum, carbon, etc. using only 1s and 0s? Even if you’re talking about a single-element crystal like a diamond, how do you communicate the positions of atoms without some convention about word length, etc.? I.e., the issue of meaning comes up in this context too. It’s the same as running an object file: it will generally only work on a particular type of processor, and be meaningless on others. Shannon was concerned only with a signal, not a signal that represents something. But if you want to use a digital sequence to communicate an arrangement of something else, you’re smack in the realm of representation.The Bugatti discourse bears a striking similarity to certain views of Ayn Rand. In an essay called “The Metaphysical versus the Man-Made,” collected in “Philosophy: Who Needs It,” Rand says “The power to rearrange the combinations of natural elements is the only creative power man possesses. It is an enormous and glorious power — and it is the only meaning of the concept `creative.’ … Creation means the power to bring into existence an arrangement (or combination or integration) of natural elements that had not existed before. (This is true of any human product, scientific or esthetic: man’s imagination is nothing more than the ability to rearrange the things he has observed in reality.)” This theme got picked up by the economist Paul Romer, who gave several interviews where he spoke about obtaining value for economic growth from rearrangements of the elements from the Periodic Table, and the value of “recipes” for doing so. Indeed his 1990 version of endogenous growth theory is supremely John Galt-ish, being hard-wired so that the value added by labor pales in comparison to that added by “ideas.” CH acknowledges being influenced by Romer, and aside from its overall Prometheanism the book has plenty of passages extolling creative geniuses, including the story of poor Boltzmann, a chapter entitled “Out of Our Heads!” about genius inventors and the “favors” that Michael Faraday did for us, and even the Acknowledgments that chronicle CH’s violent and messy creative torments. To be fair, though, CH doesn’t give any sign of endorsing Rand’s culture of self-interest, so her influence may have been indirect only.(b) Cosmic whirlpool discourse:CH tells us several times we’re exceptional when it comes to information, e.g. “We can think of our planet as a little whirlpool of information in an otherwise vast and barren cosmos” (@30; see also xx, 181). I read this together with the statements (I) and (II) mentioned above in Section 1, in which an “information-rich” bathtub whirlpool emerges naturally from the steady state of a physical system that’s out of equilibrium (an SSOOE system).OK, so let’s consider: either the type of information that emerges from natural SSOOE systems is qualitatively the same as the information involved in the human economy, or it isn’t.First, let’s suppose they’re the same. Then CH is simply wrong in his cosmic assertions. There are SSOOE structures *throughout* the universe. Here are a few: The spots and bands on Jupiter and the other gas giant planets of our Solar System (and presumably in other solar systems as well). Sunspots and chromospheric granularity on our Sun, and presumably on many of the ca. 10^22 other stars in the visible universe. Spiral arms in our Galaxy and in a sizeable fraction of the other ca. 10^11 galaxies. The distribution of galaxies in our universe. (See chapter 1 of Cross & Greenside for a few more cosmic examples.) If our bathtubs are generating the same type of information as our economy, then so are these puppies. What we produce is pretty insignificant.Alternatively, suppose the information in our economy is qualitatively different from the stuff in our bathtub. Then all the physics mumbo-jumbo about Boltzmann, physical order, etc. is sheer empty pretentiousness.Take your pick.(c) What-the-economy-does discourse:Arguably, this discourse is distributed throughout the book in the “crystallized information” metaphor for manufactured products, but CH puts it explicitly early on: “Information, when understood in its broad meaning as physical order, is what our economy produces. It is the *only* thing we produce, whether we are biological cells or manufacturing plants.” (@xix, emphasis in original.)Wouldn’t the the practice of recycling old computer equipment, vehicles, and other products be an obvious counter-example to this claim? If assembling a Bugatti is producing information, what does stripping one amount to?But what really qualifies this statement as being not just nonsense but dangerous nonsense is that it ignores the Second Law of Thermodynamics, and the fact that all real production processes create waste. To be clear: I’m not alluding here to the argument of the late Nicholas Georgescu-Roegen that we can’t clean things up adequately because our stock of “low entropy” is becoming depleted. (For a critique of NG-R’s argument, please see my Amazon review of “The Entropy Law and the Economic Process” and references mentioned therein.) Rather, I mean that as a matter of practical fact, we do create waste and *don’t* clean it up.CH invokes plants and ecosystems for the sake of an analogy about networks of economic knowledge and knowhow (Ch. 11), but this book is completely silent about the non-metaphorical impact of economic activity on the environment. Many toxic metals, such as mercury, naturally occur in high concentration within ores, but thanks to economic activity have wound up highly dispersed in bodies of water and in soils, where they can have residence times of centuries. The burning of fossil fuels produces waste heat and gases that enter the atmosphere, and nuclear plants inject waste heat (and much else, thinking about Fukushima Dai-Ichi) into aqueous ecosystems. By no stretch is any of these cases one of enhanced physical order “a/k/a” information. For a 21st Century physicist to tell a popular audience that the *only* thing economic activity produces is “information” is tantamount to climate change denialism (actually, denialism about a whole slew of man-made environmental ills). That’s all the more the case when the production of information is presented as humans’ triumph over the barren universe. It’s not only physically incorrect, it’s deeply irresponsible.4. CONCLUDING REMARKSThe writing style is generally easy to read, all the more an accomplishment because CH isn’t native in English (I presume). Still, there are a fair number of unfortunate metaphors and over-the-top passages that his editors and other readers should have counselled him to prune back. E.g., “Our planet is to information what a black hole is to matter and a star is to energy” (@x): what does this mean? Black holes suck in matter — is that what our planet does with information? Or the concluding paragraph (@181), which speaks of “our planet continu[ing] its rebellious path marked by pockets that are rich in information” — this suggests the pockets are in the path, but the rest of the book suggests they’re in the planet. (And maybe there isn’t even a path: see (III) discussed in Section 1 above.) We are “enslaved by the growth of order”, so is it reasonable to expect we can also appreciate “the beauty of information” if order is the same thing? (@id.) And we’re told that a “godless creation … has now been bestowed upon us”: bestowed by Whom? (@id.) I won’t dwell on the Acknowledgments, other than to say that they were, ironically for this book’s theme, TMI. (Life imitates art: see Paul Theroux’s 1979 New Yorker story, “Acknowledgments,” anthologized in, e.g., “Sudden Fiction International” (1989)). The book has endnotes without any bibliography, but the notes section is brief, so finding sources isn’t as difficult as in some other recent books.I usually try to cut young researchers some slack when reading their books. Unfortunately, the quantity and quality of problems with this book limited the extent to which I could do so. I can only suggest and hope that CH will take somewhat more time and care to do better on his next attempt. There’s a good reason why an attempted grand synthesis like this is more typically the product of a scholar’s later years.

⭐I hesitate to call the following a review but I will comment on some aspects of the book, namely those having to do with information. I found myself highlighting and writing notes to attempt to understand what the author meant by ideas such as physical order and information are the same, the universe can compute, why solids are special for containing information, why the distinction between knowledge and knowhow (since they are both information), the distinction between individual capital and social capital (since these are really not separable), etc. It seems to me that these problems with the book are due to the fact that there is no standard definition of information and what the author uses to characterize information contains some truth but that it all needs to be worked out more carefully.I had intended to elaborate each of my notes but since most of these notes were trying to clarify the author’s ideas for me in terms of information I will just give my definition of information. See below. Then the author or you, the reader, can use it to better understand what the author is saying. This definition does not depend on any mathematics, it doesn’t refer to Boltzmann or Shannon. It applies to the information in or processed by computers, the information in human and other animal brains/bodies, and life in general, the information in all our artifacts, the information in all our institutions and organizations, and the information in nature.Here’s one confusing quote:“That something, which fascinates you and me as much as it did Boltzmann, is physical order or information.” —- Hidalgo, Cesar (2015-06-02). Why Information Grows: The Evolution of Order, from Atoms to Economies (Kindle Locations 236-237). Basic Books. Kindle Edition. Here Cesar Hidalgo clearly equates information and physical order. They are not the same. This confusion messes up the title too. The book is (trying to be) about the evolution of information.Although A.J. Sutter has already pointed out that Cesar Hidalgo uses ‘information’ and ‘physical order’ to mean the same thing throughout his book and that they are not the same, I suggest the author could eliminate almost all confusion regarding information if he would apply the definition I suggest consistently throughout his book because his book has great value in so far as it can link information and the growth of information, etc. to the economic concepts of human capital and social capital. Maybe he could get one or more of his students to do the grunge work. It wouldn’t be the first time students do a lot of the work.Information is much more than physical order. It is not a physical thing but it is inseparable from matter/energy. Some people question whether information exists. We use the word but I’ve never seen a definition except in a dictionary. Most books about information don’t give a definition of information. They may give a few examples and then they get mathematical.OK, then what is information? Let’s start with a dictionary definition of information. This is from Merriam-webster.com.Main Entry: in•for•ma•tionPronunciation: ˌin-fər-ˈmā-shənFunction: nounDate: 14th century1 : the communication or reception of knowledge or intelligence2 a (1) : knowledge obtained from investigation, study, or instruction (2) : intelligence, news (3) : facts, data b : the attribute inherent in and communicated by one of two or more alternative sequences or arrangements of something (as nucleotides in DNA or binary digits in a computer program) that produce specific effects c (1) : a signal or character (as in a communication system or computer) representing data (2) : something (as a message, experimental data, or a picture) which justifies change in a construct (as a plan or theory) that represents physical or mental experience or another construct d : a quantitative measure of the content of information; specifically : a numerical quantity that measures the uncertainty in the outcome of an experiment to be performed3 : the act of informing against a person4 : a formal accusation of a crime made by a prosecuting officer as distinguished from an indictment presented by a grand jury2b is what we want. Let’s see if we can make sense of it:The attribute inherent in and communicated by one of two or more alternative sequences or arrangements of something (as nucleotides in DNA or binary digits in a computer program) that produce specific effects.So, we must first of all have:“one of two or more alternative sequences or arrangements of something”Consider a traffic signal. A simple traffic signal as seen by a driver approaching an intersection has three lights. Only one light is on at any time. On the top is a red light, on the bottom is a green light, and in the middle is a yellow light. If the green light is on it means the driver can continue driving through the intersection if the intersection is clear. If the red light is on it means the driver must stop the vehicle. If the yellow light is on it means the red light is about to come on and the driver should stop the vehicle if it is possible to do so before it gets into the intersection.So there are three alternative arrangements of the traffic signal: 1) The green light is on; 2) The red light is on; 3) The yellow light is on. They are alternative arrangements of the traffic signal since any one light could be on but no more than one at any one time. At any instant the driver will see only one of the three alternative arrangements of the traffic signal. And the driver is expected to do different things depending on the color of the light seen. The different colors of the light produce the “specific effects”.The definition of information as applied to this traffic signal says the information is:The attribute inherent in and communicated by one of two or more alternative sequences or arrangements of something that produce specific effects.So what is the attribute inherent in and communicated by one of the three alternative arrangements of the traffic signal? It is the color of the light that’s on at any given instant, either green, red, or yellow.This definition of information says that information is the attribute inherent in and communicated by one of two or more alternative sequences or arrangements of something … What are some other attributes of something? We just saw that color is an attribute. Other attributes of a physical thing are weight, size, shape, texture, loudness, frequency of sound or light, in general physical qualities of various things, matter and energy, that are measurable, observable, perceivable, etc.What does it mean to say something is inherent in something? It just means it is essential to that something as sweetness is inherent in sugar. This is why we say that information is in or on something, as information is in a book, on a computer disk, in your body, or in your brain or on the internet. Whatever the information is in or on we call the substrate of that information.What is this “communicated by”? What does this mean? What is it telling us? In the case of the traffic signal we can say the attribute inherent in and communicated by the traffic signal is the color of the light which is on at a specific instant. Communicating is more than emitting. It takes two to communicate. For communication to occur there must be a receiver, a someone or something to receive, perceive, or otherwise react to the color of the light which the traffic signal is sending out. In our example it was a person driving a vehicle.To communicate is to transfer information. So something created the information (made a pattern of matter/energy) and something (else) reads it or perceives it and uses it; that something (else) interprets the information and makes the decision, the choice(s) inherent in the information. The something (else) sees an alternative and does or causes the specific effect associated with the alternative seen. How does it do this? A human already has information in its brain/body that he or she matches with the alternative seen, the external pattern of matter/energy. The receiver then does whatever it has associated with the alternative.This connects with the phrase “… that produce specific effects”. The color is communicated to the driver and produces specific effects in her or him. In particular there are three separate behavior patterns that were embodied in the driver’s brain/body long ago one of which will be triggered depending on which of the three colors she or he sees. So green produces one behavior, red produces another behavior, and yellow a third behavior.So information is not just alternative arrangements of matter/energy on some substrate. Someone or something must do something with it or be able to do something with it. By the above definition of information, information must be communicated to someone or something.We have not yet talked about the word “sequence” in the definition. A sequence is a particular kind of arrangement, so it’s not really needed in the definition and we could leave it out. But a sequence is a nice kind of arrangement. The things in a sequence are ordered. In a sequence there is a first thing, a second thing, and maybe a third, a fourth, and so on. For example a word is a sequence of letters. A sentence is a sequence of words and punctuation marks. Most discussions of the information processed by computers end up talking about binary numbers, such as 0100101111 and 11001. These are sequences of zeros and ones. This happens because the information processed by ordinary computers is or can be coded as binary numbers. We want information processed by computers to be included in any definition of information we consider, but our main emphasis is to understand information as it relates to human interactions, human communication and behavior. Below we can leave out “sequences” as long as we keep in mind that a sequence is a particular kind of arrangement.We have said that the “something” that is arranged means something in the physical world, matter and/or energy, and that we don’t need the word “sequence” since a sequence is just a particular kind of arrangement. Then we rephrase the definition as:The attribute inherent in and communicated by one of two or more alternative arrangements of matter/energy that produce specific effects.Now what’s the effect of the word “alternative”? It is to say that any of the arrangements could be used and the arrangements must be different and distinguishable by someone or something. How do we make sense of the phrase:“that produce specific effects”How can different arrangements of matter/energy produce specific effects? The definition doesn’t say. The phrase “communicated by” implies there must be something or someone which looks at, or manipulates or otherwise perceives or interacts with the alternative arrangements of the matter/energy and do different things (produce specific effects) depending on which particular arrangement is present at the time of the looking, etc.Could an arrangement of matter/energy by itself produce specific effects? Yes. Atoms and molecules move around, combine, break apart, and in general interact with each other. A system of atoms could be in multiple different arrangements, and such a system, operating according to the laws of physics and chemistry, will do various things, and produce specific effects, in the future, depending on its particular arrangement now.Since someone is something, we can drop “someone” to simplify the definition further. So each arrangement itself, or something else, must produce specific effects. So something must be able to do different things depending on the specific arrangements that occur. So the arrangements must be distinguishable by whomever or whatever acts on them, because if two arrangements were indistinguishable by whomever or whatever acts on the possible arrangements, the actor couldn’t do possibly different things depending on which arrangement the actor saw. If you are color blind and cannot see the difference between the colors red and green, then you cannot use color to determine whether you should stop or go at a traffic signal. So for you the color of the light is not information for you since you can’t use it; you can’t distinguish the red color from the green color. Red/green color blind people can use other information such as the fact that the red stop light is above the green go light.So we could add the word “distinguishable” but we don’t have to since distinguishability is implied by the actions someone or something must be able to take. So we have:Information is the attribute inherent in and communicated by one of two or more (distinguishable) arrangements of matter/energy that something uses to produce specific effects corresponding to the specific arrangement detected.Notice that two distinguishable arrangements could produce the same specific effect.Here’s another way to say it:Information is the attribute inherent in and communicated by one of two or more (distinguishable) arrangements of matter/energy such that if the arrangement is one way one thing is done, and if the arrangement is another way a possibly different thing is done, and so on. And there must be someone or something which distinguishes between the arrangements and acts accordingly.Notice that the above discussion about distinguishable arrangements and whether or not different arrangements could produce the same effects amount to questions about the associations between the arrangements and the specific effects. An association relates members of one collection to members of another collection. For example a collection of arrangements and a collection of specific effects. The simplest association is called a one to one correspondence: For each arrangement there is exactly one effect corresponding to it such that that arrangement produces that effect, and for every effect there is exactly one arrangement that produces that effect.Associations don’t have to be one-to-one. There can be many-to-one associations. Suppose somehow a person who has never seen nor heard of traffic signals and knows absolutely nothing about them is driving a car and comes to an intersection with a traffic signal. This naïve driver would likely do the same thing — keep on driving — whether the light is green, red, or yellow. So this would be a many-to-one association in this driver. He can distinguish the three colors green, red, and yellow, but he does the same thing — keeps on driving — whatever color he sees. There can be one-to-many associations also. For example a single word might have multiple meanings.The association between different arrangements of matter/energy and their effects can be a one-to-one correspondence or there can be situations where different distinguishable arrangements have the same effect as with the naïve driver. Also see the discussion below of DNA.There are situations where arrangements might be distinguishable by one actor but not by another. As mentioned above some people are color blind to various degrees. So for arrangements of matter that depended on different colors to be distinguishable, some people might see two arrangements as different whereas others see them as the same. So information depends on the user of it.If the different arrangements themselves produce specific effects, then what is the something which produces the specific effects? It is a physical system in which the different arrangements of matter/energy are possible, and in which if one of the arrangements occurred, then that arrangement’s effect would occur.What about the distinction between the agent acting on the information versus being able to act on it? Must information be acted upon to be information? It’s the ability that counts. Consider some book in a bookcase that I will most likely never read again and that most likely no one will ever read again. Would we say there is no information in that book? No, because I could read it or I might read it or someone else might read it.We might say that for a person information is a rule they apply: If arrangement 1 is present, then do action 1; if arrangement 2 is present, then do action 2; if arrangement 3 is present, then do action 3, and so on.This definition is from my book Information, Communication, Cooperation: The Force of Human Progress. It’s available on Amazon as a digital or paper book.Joe Rebholz

⭐This book is excellent at making one take a fresh look at economics, time, life and the nature of reality. The focus on complexity for me is an interesting rebuff to neoclassical models and their tendency to simplify everything into combining homogeneous factors of production.Read in combination with ‘Windows of Opportunity’ to understand how this perspective can be brought to bear on practical economic problems.

⭐This book contains interesting ideas but in the end is a bit underwhelming. He is definitely on to something in trying to bring the ideas of information science into economics. And I enjoyed his account of the general principles of information. But stops some way short of practical policy implications. I still appreciated reading it.

⭐brilliant; most thought provoking book I have read in past three years. Not just a book for economists or philosophers, but for anyone who wants to think deeply about the world and human relationships

⭐Proof that short books can distil very big ideas. A physicist on economics changes the way that information content is thought about in the pricing mechanism. The language is strange to an economist’s ear, but the ideas are dazzling and important.

⭐Fascinating, but his writing style can be sometimes impossible to read. I found myself re-reading pages over and over again because of weird phrasing and incredibly long, complex sentences.

Keywords

Free Download Why Information Grows: The Evolution of Order, from Atoms to Economies in PDF format
Why Information Grows: The Evolution of Order, from Atoms to Economies PDF Free Download
Download Why Information Grows: The Evolution of Order, from Atoms to Economies 2015 PDF Free
Why Information Grows: The Evolution of Order, from Atoms to Economies 2015 PDF Free Download
Download Why Information Grows: The Evolution of Order, from Atoms to Economies PDF
Free Download Ebook Why Information Grows: The Evolution of Order, from Atoms to Economies