2022 By The Books: Q4
The final installment for this year is in, which means it’s time for a little retrospective. I originally started this series five years ago, and even though the shape of it has undergone some minor transformations, I’ve stayed true to the original idea: Instead of offering concise summaries (which you can easily find elsewhere) or compare the books against each other on some made-up scale, I’ve tried to highlight (or criticize) the parts the resonated most with me. This naturally leads to some reviews being much longer than other, which should not be taken as a proxy for how much I enjoyed any given work. After all, forcing myself to put my thoughts to paper is something I’m doing by and large for my own benefit, so I see no reason to stretch them thin.
Looking only at the numbers, this year has been the most “productive” of the entire period, with a grand total of 50 books under my belt, or roughly one per week. The comparison isn’t entirely fair, because unlike the other years, I’ve added more and more audiobooks — 16 in total for 2022. So I’m still reading about as much as I usually do, but the gaps that I previously filled with music or podcasts — e.g. during long training runs — are now occupied by books, too.
Lastly, before we dive into the remaining reviews, time to pick a favorite for this year! The winner is James Scott’s Seeing Like A State, covered below.
Basic Economics, Thomas Sowell
A book that delivers exactly what it promises: A nontechnical introduction to basic economic principles and their applications to the real world, and a very solid one at that. It’s well written, contains engaging examples to illustrate the point and doesn’t remain within the realm of pure theory, which makes it an excellent place to start for the curious layperson.
Interestingly, the key message is delivered towards the very end, when Sowell outlines what he considers to be the most common fallacies that he thinks are chiefly responsible for the sorry state of affairs concerning economic reasoning in the public arena. These are:
- Conceiving of economics transactions as zero-sum games: What one party gains, the other must lose.
- Ignoring the role of competition in the marketplace, and how it shapes the behavior of its actors.
- Disregarding long-term or unintended consequences of policy interventions: This is of course a point made by Bastiat more than 150 years ago in What Is Seen and What Is Not Seen.
(To which I’d add “judging policies by their (stated) goals, rather than the incentives/outcomes they create” and “ignoring the fact that we live in a world of scarce resources which have alternative uses”, two points he makes frequently throughout the book.) I think all of these are excellent points, much undervalued by intellectuals and common folks alike, and Sowell deserves credit for making their relevance felt on many different occasions.
I did wonder whether I learned my marginal benefits lesson well, as in why would I go through yet another book on Econ 101? Anyway… One may quibble with the some of Sowell’s interpretations of certain facts and what they imply for economic theory — and indeed one should! — as some of them are more ideologically colored than he might want to admit. Even so, finding a counterexample to his principles, or judging their realm of applicability differently, should not lead you to reject the basic rules wholesale. Compare this with physics: We know that Newtonian mechanics is incomplete and will yield faulty predictions under certain circumstances, but no one in their right mind would infer from this that it’s a useless framework and we’d be better off without. Indeed, an engineer blissfully unaware of general relativity, equipped with the know-how to operate Newton’s toolkit, is probably a better physicists than someone ridiculing the former’s construction plans because “Einstein disproved classical mechanics”.
My main criticism would be that his own claims to the contrary, the book clearly has an ideological bent. Even though he proclaims not to pass judgment whether a free market system or its alternatives are a superior form of economics organization, the only exception he comes up with are national defense and maybe immigration. This sound exactly like the kind of thing a certain type of conservative would say, and it probably limits the likelihood to which readers of different convictions will give his work a chance. (This, by the way, seems to be a common feature of Sowell’s writings: For most of the time you wonder why he gets labeled a conservative, and then all of a sudden he makes a move that doesn’t seem to square with any of the previous “value-free” analysis.)
Brilliant Blunders, Mario Livio
On the surface, this looks like yet another variation of the tired theme “geniuses make mistakes, too”. Livio, by contrast, is up to something more fundamental, although I’m still trying to dissect the precise criterion he uses to group the blunders he presents. What, if anything, do Darwin’s reliance on blending inheritance for as a mechanism for evolution, Kelvin’s misguided attempts to determine the age of the earth, Pauling’s efforts to model DNA, Hoyle’s refusal to acknowledge the mounting evidence in favor of the Big Bang, and Einstein retraction (not introduction!) of the cosmological constant from his field equations have in common?
Note first that the “brilliant” in the title is less about the progenitors of these mistakes, but about ideas of theirs that were ingenious even when they failed. Otherwise, the book should have dwelled on Pauling’s obsession with Vitamin C or Einstein’s dead-ended quest for a unified field theory. No, they were, I suppose, brilliant in that they helped science overcome an impasse: Pauling’s version of DNA, despite its fatal flow of not actually being an acid at all, prompted Watson and Crick to model it using a double helix. Darwin could only propose his theory of evolution because he did not see through the implications of pangenesis (Mendelian genomics had not reached the scientific community). And Kelvin, despite being orders of magnitude off with his calculations, paved the way for more precise techniques of determining the age of celestial bodies. Hoyle and Einstein, however, are much harder to fit into this picture: Hoyle possibly because his steady-state universe offered an alternative to Big Bang cosmology that was consistent with the then-known facts, forcing researchers to probe deeper? And Einstein to remind ourselves that heuristics (beauty and simplicity, in that case) that worked in one case don’t always generalize?
The many vignettes about the scientific community should be of interest not just for historians of science; by contrast, the psychological speculations about what may have led these eminent scientist to commit such blunders are rather shallow and give the book a pop-sci touch.
The Emperor of All Maladies, Siddhartha Mukherjee
Mukherjee is one of the very best science writers of our generation, and the Emperor Of All Maladies is just as well-written, informative and thorough as his later The Gene, which I read a few years ago. (He also has a new book out this year, The Song of the Cell). The current volume, which marked his rise to national prominence, provides a basic understanding of cancer (or, rather, cancers), but maybe even more than that, it is an intellectual history of our understanding of cancer and our (often unsuccessful) attempts to defeat it. Other readers will find this history, at least as far as major breakthroughs are concerned, rather americentric. I simply wouldn’t know if much of the important research has indeed been carried out in the US, and therefore won’t pass judgement on whether this perspective is biased or not. Much more relevantly, it’s a chance to understand in depth why the state of the art often advances at a glacial pace (if it advances at all), why victories are often temporary, and why hopes that cancer will one day be eradicated, alas, may turn out futile. This is not to say that all is lost: indeed for some types of cancer, which used to be fatal not too long ago, we have now developed excellent treatment options. And yet, the very nature of cancer (unlike, say, infectious diseases) make it an unlikely candidate for being crushed permanently.
Essentialism, Greg McKeown
From the popular series Corporate Talking Points Masquerading As Life Advice … and yes said I would stop reading those, doubly so if the author has a side hustle as a Mormon bishop. And make no mistake: Essentialism remains firmly rooted in its native genre, complete with the usual Harvard Business touch that makes them both annoying and a natural candidate for the NYT bestseller list. But if you manage to make peace with all that, you can get something valuable out of it.
Note that focusing on what’s essential is not meant to be a guide how to put in as little effort as possible and see if you can get away with with, about cutting corners and offloading annoying tasks on your colleagues. But let’s be honest: In every organization, inertia is a most dominant force, and routine tasks stay around long past the point of positive returns. In addition to that, if you’re working in a client-driven environment, you’ll find that your customers have every incentive to add new requests without ever much considering how much value that would create. Cutting out the busywork — at work and outside of it — is easier said than done, but saying no by default, including to promising opportunities, can be liberating. The sad fact is that if you’re not deliberating deciding what matters, and pursuing a (very) limited number of things fervently, others will be eager to do the planning for you.
Of course there’s no free lunch. There are some real costs to focusing on the selected few, costs that many of us don’t wish to pay. All too often, advice about how to zero in on the essential takes the form of deceptive life hacks. And looking at nothing but the big pictures is a recipe for disaster in its own right: A company in which no one would read (or respond) to their email would soon be in serious troubles. More pointedly, does a municipal garbage collector focus on what’s essential? There is certainly a sense in which “being an essentialist” is a luxury that only certain spheres of the professional class can afford. Does that invalidate all of the above? Your call.
Experimental Economics, edited by Pablo Branas-Garza and Antonio Cabrales
A decent introduction to experimental economics, mostly focusing on game theory but with a bit of neuroeconomics added to the mix. If you have no idea what’s behind this concept and how it differs from behavioral economics, this is as good a place to start as any. Alas, the quality of the contributions varies wildly, not just in technical terms (with the exception of one chapter, the primer does almost without formulas) but also stylistically— none(?) of the authors is a native speaker, and it seems the same is true of the proofreaders. Tellingly, there are no contributions by any of the big names in the field, and the sometimes rather simplistic reasoning conveys the impression that this has been a collaboration of lower-tier researchers.
Ethical Intuitionism, Michael Huemer
Huemer should be credited for writing in a way that’s accessible for outsiders without dumbing down the content. The prose is very clear, and he makes a convincing case against standard metaethical positions before coming out in favor of intuitionism. Is it fair? I’m not a professional philosopher and can’t judge to what extent he might be strawmanning his opponents. Like everyone else arguing for a minority position, he has to explain why his view does not command much support among people who deal with the topic on a regular basis, but his assessment that intuitionism fell out of favor because of “scientism, political correctness and cynicism” strikes me as somewhat self-righteous.
He makes a lot of good points throughout the book that I won’t all list here, but I struggle two important points. The first is his contention that there are moral facts: clearly, if these exist, they can only exist relative to sentient (or intelligent) beings. Had higher forms of life never developed, would they therefore still exist? In contrast to empirical facts about the natural world such as the age of the universe — which would still be true even if no one was around to observe them — they clearly belong to another realm. He deals with the evolutionary aspect briefly, but I found his arguments therein unsatisfactory.
The second point is that he seems to place a huge premium on linguistic arguments about how people typically use ethical language, which make sense under intuitionist assumptions but much less so for e.g. non-cognitivism. While not altogether irrelevant, I was somewhat surprised that he put so much weight on “mere usage”.
In the end, of course, one may want to ask: So what? Maybe this is an unfair charge against a book on metaethics, but I do wonder what would change if more people came to accept his view that moral facts exist and are formed by intuition? Huemer claims that opposite views make us less likely to act upon our moral convictions, and that sounds plausible enough. And yet, “acting on moral convictions” itself may mean wildly different things, depending on what those convictions are. It does not, as far as I can tell, provide us with a method to decide whether a certain intuition is indeed a good one and ought to be embraced. Again, Huemer’s book is about metaethics, not ethics itself, but the question how or whether a good meta theory can help us with “the real thing” — or whether something akin to the infamous “shut up and calculate” approach in physics wouldn’t fare better — is intriguing.
The Inevitable, Kevin Kelly
An excellent book on the future of (digital) technology, in the broad sense of the word. It succeeds because it zeros in on the sweet spot between the hopelessly speculative and the uninspiring extrapolative, between the very distant future and the relatively predictable coming years. Such an approach has the advantage of being able to point to actually existing technologies and think about how their development and widespread usage will change society, contrary to, say, speculations about AI takeover scenarios, which are entertaining thought experiments but offer little else besides. Some of his points may sound rose-colored; downsides and disruptions are not what Kelly spends most of his pages on. But I for one applaud the effort to publish a techno-optimistic take when so many discussions focus on dangers, and dangers only.
Kelly has identified twelve trends that will shape the decades to come, inevitably so if we believe him. Each would merit a discussion of its own, but maybe most interesting overarching theme is about how many relationships will be reversed because the side that used to be relatively scarce no longer is: Not the production of content will be prized, but the filtering of relevant content. Not the answers will be important, but crafting good questions (ChatGPT, anyone?). Not goods but deciding what to do with them. Those of us who will understand how to capitalize on these trends, I suspect, will profit handsomely, while the rest of us will at some point be surprised how it could ever have been different.
Into Thin Air, Jon Krakauer
Much as I share Krakauer’s fascinations for mountains, I have never made any serious forays into alpinism, nor climbing in general for that matter. This no doubt deprives me of many spectacular views, as well as much of the adrenaline rush along the way, but whatever little I tried so far I got frustrated with the slowness quickly. Anyway, apart from the sluggish advances (enforced by low levels of oxygen and technical terrain) I dislike, mountaineering just seems unreasonably dangerous to me, plain and simple. Although it’s a well-known tidbit that most accident are due to human error (the very idea of launching a particular summit bid featuring prominently among them), it is illusory to think that with the right preparation or mindset, such errors could be eradicated, in part or wholly, and Into Thin Air is a first-rate account of why this is so. Amateurs and celebrated alpinists alike have succumbed to the elements, proving that neither skill nor experience are a reliable safeguard. A gripping read until the end, the book is a tantalizing mix of both WTF moments and miraculous feats of willpower. The author may not be America’s foremost stylist, but he sure knows how to keep his readers turning the pages.
Invisible Man, Ralph Ellison
Not sure what to make of this. It’s…decent? After going back and forth between the book itself, various summaries and reviews, I still can’t come up with anything of substance to say about it, except that you’re not doing yourself a favor if you’re reading it through the lens of recent protests. I’m a bit disappointed that a novel many of my favorite writers hold in such high regard didn’t click with me.
Life 3.0, Max Tegmark
There’s little in this book that is original, and if you’re familiar with the works of Hans Moravec, Ray Kurzweil or Nick Bostrom, I don’t think you’ll get much out of this. It does make a decent primer, if you don’t mind the constant name-dropping (“I once had dinner with Henry Kissinger…”), self-congratulatory episodes about his Future Of Life Institute, its conferences and petitions, and the lavish praise he heaps upon its donors (“Elon Musk was one of the smartest, most handsome people I ever met..”).
Even aside of these moment of inflated self esteem, I also think the book has some serious shortcomings that are all the more comical in a book with such limitless ambitions. To me, the first thing that any any author addressing topics as vast as “the future of humanity” should appreciate is how bloody easy it is to delude yourself about what you think you know. In one telltale instance, he speculates about the end of our universe and assigns probabilities to different Armageddon scenarios — 40% big freeze, 1% Big Crunch (I’m making the numbers up), but also 50% for “none of known theories”. I think the charge of scientism — when we put numbers on outcomes that are very hard/impossible to predict — is overused, if for no other reason that because its advocates rarely suggest a better way to deal with radical uncertainty. But to believe that the proper way to deal with it is just adding a 50% “I don’t know” to the mix is, if anything, even more astonishing.
It is probably worse than just uncertainty about outcomes. Many concepts that we use today to assess the desirability of events, say pain or consciousness, may well be a result of our parochial vision than universal features. Which means that extrapolating these naively to the future and using that as a benchmark to decide on the right course of action today — never mind the arrogant belief that our decisions would make the difference between doom and eternal bliss — can only result in confusion. It is conceivable that our “cosmic endowment” is so large as to dwarf everything we experience in the here and now. But were having hard enough a time trying to find an agreed-upon moral framework even for today’s problem — are we sure these tools are useful when considering a civilization of sentient AIs, colonizing faraway galaxies and stretching billions of year into the future?
I hasten to add that my claim is not that artificial general intelligence is impossible, nor that worrying about it pointless. However, to much of AI “safety” research seems — and yes, this is overly cynical — to be more about community rituals (discussing outlandish ideas with other smart people while convincing yourself that doing so may save humanity) than anything else.
Addendum: A friendly but ultimately damning article casting doubt on many of the core assumptions underlying the idea of superhuman AI.
Locked In, John Pfaff
You have to care an awful lot about US domestic issues to make this worthwhile (I can’t see myself finishing a book on the prison system of, say, Peru). Then again, the country is just such a remarkable outlier when it comes to incarceration rates that you can’t help but wonder what the freaking hell is going on here:
Even second-placed Thailand only stands at 520 prisoners per 100,000 inhabitants! How can we make sense of that?
Pfaff starts out by highlighting that incarceration rates weren’t always as sky-high as they are today (uncontroversial), and then goes on to dismantle popular explanations what caused them to explode. Crunching numbers, he shows that
- drug crimes, especially non-violent low-level offenses, are not a major driver of prison growth. In the overall picture, drug offenders make up too small of a fraction to affect the trajectory much.
- The same goes for parole violation and long prison sentences, both of which are often overestimated in terms of their severity.
- Nor do private prisons matter much, which again make up only a small portion of all prisons in operation (public sector unions are a different story, though).
Instead, most prisoners are actually in for fairly serious crimes (murder, manslaughter, aggravated assault), rather than because of overzealous politicians that decided to go tough on victimless crimes. And that makes it much harder to institute meaningful reforms, because reoffenders loom much large in the public consciousness than do the wrongly locked up. But of course the ideal trade-off is not to reduce type 1 errors (failing to lock up a criminal) to zero, and much work remains to be done to convince the electorate that we should be willing to err more often in this direction.
Apart from messing around with this trade-off, what can be done? Pfaff thinks it’s mostly prosecutors who are to blame, largely by excluding all alternative theories: The police catches criminals at a more or less stable rate, violence is on a downward trajectory since the 1990s, and juries’ verdicts haven’t gotten less forgiving either. Whether this means that they did their job sloppily before they got “tough on crime” is not entirely clear to me, yet there’s no doubt they enjoy a lot of discretion in how they handle their cases — necessarily so to some extent, but certainly there’s room for improvement.
I would have liked to see a more global perspective at times: Is the US just an incredibly violent place, compared to other rich countries? And if so, why?
The Lucifer Principle, Howard Bloom
“The line separating good and evil passes not through states, nor between classes, nor between political parties either — but right through every human heart — and through all human hearts. This line shifts. Inside us, it oscillates with the years. And even within hearts overwhelmed by evil, one small bridgehead of good is retained”
Following Solzhenitsyn, Bloom presents his own spin on the perennial question why we can’t all be nice to each other. His writing is lavish, polemical , entertaining and bound to miss many targets by a mile. You can almost feel his desire to polarize, and critics will no doubt be quick to point out that his background as a music publicist doesn’t exactly qualify him to weigh in on the topic. Then again, none other than David Sloan Wilson wrote the foreword, which is an endorsement few others can boast about.
Until fairly recently, the standard story of why bad things happen was that there are Bad People in the world, born this way and unchanging unless restrained by external forces. Against this, the contemporary orthodoxy tends to view crime as caused by environmental factors. Bloom likes neither theory: In his view, evil is very much situational — the Lucifer in all of us — but rather than being a victim of circumstances, he views our species as fundamentally opportunistic, exploiting opportunities that present themselves and cooking up clever rationalizations after the fact. Even though we have gotten incredibly more prosperous over the last two centuries, the hard currency is not wealth or standards of living but status, and while we may ultimately be able to lift everyone out of poverty, who gets to be at the top of the hierarchy will always be a fierce zero-sum competition. Thus, even periods of growing prosperity may be more violent than preceding eras marked by deprivation, if such growth makes the established pecking order seem vulnerable. Even non-egalitarians like me have to acknowledge that inequality poses a challenge that economic development alone can’t tackle. In that, he shares a lot of common ground with Robert Sapolsky, who in my reading does seem to hold that (extreme) inequality is worse than poverty.
Bloom raises a number of interesting points about group selection, that fiercely debated area of evolutionary theory, and how it may help to explain e.g. suicide (which, at first glance, seems like the most obvious candidate for a dispossession that should have been selected against). I don’t feel qualified to adjudicate these debates, but it offers an interesting challenge to the views of Dawkins, Pinker et al. that I’m more familiar with.
Overall, one is left with an extremely pessimistic outlook, and even the few things he’s willing to embrace as an antidote to violence (the liberal capitalistic order among them) he doesn’t expect to last, echoing Schumpeter. Some additional lowlights are the sections on homeopathy and Islam, which start out contrarian but then descend into silliness or malice.
Seeing Like A State, James C. Scott
The problem with reviewing Seeing Like A State is that there are already many excellent reviews of it out there. I’d highlight in particular this article comparing his work with Joseph Henrich’s Secret Of Our Success (another favorite of mine), a long discussion by Scott Alexander, and this Cato Unbound exchange. So what can I possibly add to this picture?
In its most condensed form, Scott describes state formation as on grand exercise in standardization. Historically, states used to know very little about the society over which they presided and hard a hard time understanding local customs, measures, property titles, relationships and so forth. All of which stood in the way of generating a steady stream of income (via taxation) and army personnel (via conscription) from their subjects. “To follow the process of state-making, then, is to follow the conquest of illegibility”, he writes, and documents a great many examples of how this unification looked in practice, from scientific forest management to urban planning. One of my favorites is the case of the so-called “window tax” levied in France, England and Scotland in the early 17th century, attempting to measure the value of a house by the number of windows it contained. You don’t have to know a great deal about European architecture to guess what unintended consequences this led to! In Scott’s own words, “[t]he window and door tax illustrates something else about “state optics”; they achieve their formidable power of resolution by a kind of tunnel vision that brings into sharp focus a single aspect of an otherwise far more complex and unwieldy reality.”
Now, Scott does not claim that standardization is always bad (no doubt much of modern life depends on it), nor that states are the only entities driving this process — he makes it very clear that he has no desire to exonerate big business from his charges. It would be naïve to assume that market forces alone ensure diversification. His point is not that cataloging things and enforcing standards is always bad, but that we are to quick to look at local customs as quaint, irrational and inefficient. They may well be inefficient from a narrow eagle-eye perspective, but they offer a way to protect us from systemic failure that threatens monolithic societies.
He has a curious chapter contrasting Lenin with Rosa Luxemburg that, in my view, paints the latter in too rosy a light. (Everyone has heard a version of her “Freedom is always the freedom of dissenters” quote, but few appreciate her role in attempts to violently overthrow the democratically elect social-democratic government of the Weimar Republic.) He makes up for this by providing the best interpretation I’ve seen so far about the importance of Jane Jacobs, whose Death and Life of Great American Cities I found rather uninspiring. Anyway, this is a must-read, whether or not you’ll agree with his general diagnosis.
Surely You’re Joking, Mr. Feynman!, Richard Feynman
Or is he? With Feynman, you never really know, but I like to think of him as a very serious person, whose curiosity and disregard of what he considered irrelevant cultural norms make him stand out from the crowd. Many fascinating episodes, such as his brief collaboration with James Watson just before the latter’s DNA breakthrough, or the fact that his first scientific presentation was attended by Pauli, Wigner, Einstein and von Neumann. I hadn’t know that Feynman was such a playboy, at one point testifying in favor of a topless bar he liked to frequent. Many of his dating adventures — in particular those involving undergraduates while he was a professor at Cornell — would certainly be judged less mildly today, and rightly so I would say. Had Feynman been born 40 or 50 years later, would he still have become the same larger-than-life figure, or would we have turned him into a pariah?
What Do You Care What Other People Think?, Richard Feynman
I don’t think there’s much marginal benefit reading What Do You Care if you’ve been through Surely You’re Joking before, but if you enjoyed the lighthearted tone of the the latter you’ll have fun reading this as well. This one is shorter and centers around two main stories (Feynman’s first marriage to Arline Greenbaum and his role on the commission to investigate the Space Shuttle disaster). The first part is strangely cold, the second shows him as an astute observer of institutional behavior — much more nuanced than the usual pronouncements about how bureaucratic organization behave.
What If?, Randall Munroe
Cute little collection of conundrums asked by people who don’t really care all that much about actual science but are really into supernovae, black holes, lightnings and earthquakes. I wouldn’t (and didn’t) read all of it, but the questions are fun, and some answers surprising. Good filler while you’re waiting for your doctor’s appointment.
Munroe rose to fame creating nerdy web comics and this review wouldn’t be complete without one:
Wolf Hall, Hillary Mantel
According to none other than The Guardian, this is the very best book (not just best novel! Book!) of the 21st century, and that’s largely why I decided to give it a try. Even by the most conservative estimates, humanity as published upward of ten million books since 2000, so this should’ve been a minor revelation (I had never heard of Mantel before). In actual fact, it wasn’t even good enough to finish more than one third or so of it. It’s not just that I’m not terribly interested in the period portrayed (the reign of Henry VIII), I never got into the storytelling either. From the little I recall, some scenes just didn’t make much any sense (such as when the protagonist’s wife dies, and he just leaves the room and continues with his life unperturbed), and the technique, if that is the proper word, of referring to everyone as “he” made it hard to understand such basic things as who was talking and who is being addressed.