Gwern's Most Important Writing (in essays, tweets, book reviews, and other forms)

Also see Most Important Slate Star Codex Posts.


This is a collection of Gwern’s most important writing that comprises

Plus,

My selection is quite opinionated and, if you publish your own, I might link to it here.


Essays

Why Correlation Usually ≠ Causation: Causal Nets Cause Common Confounding (a):

It is widely understood that statistical correlation between two variables ≠ causation. But despite this admonition, people are routinely overconfident in claiming correlations to support particular causal interpretations and are surprised by the results of randomized experiments, suggesting that they are biased & systematically underestimating the prevalence of confounds/common-causation. I speculate that in realistic causal networks or DAGs, the number of possible correlations grows faster than the number of possible causal relationships. So confounds really are that common, and since people do not think in DAGs, the imbalance also explains overconfidence. [emphasis here and below mine]

Also, Everything Is Correlated (a) to really drive this point home.

The Narrowing Circle (a):

The “expanding circle” historical thesis ignores all instances in which modern ethics narrowed the set of beings to be morally regarded, often backing its exclusion by asserting their non-existence, and thus assumes its conclusion: where the circle is expanded, it’s highlighted as moral ‘progress’, and where it is narrowed, what is outside is simply defined away. When one compares modern with ancient society, the religious differences are striking: almost every single supernatural entity (place, personage, or force) has been excluded from the circle of moral concern, where they used to be huge parts of the circle and one could almost say the entire circle. Further examples include estates, houses, fetuses, prisoners, and graves.

Littlewood’s Law and the Global Media (a):

Selection effects in media become increasingly strong as populations and media increase, meaning that rare datapoints driven by unusual processes such as the mentally ill or hoaxers are increasingly unreliable as evidence of anything at all and must be ignored. At scale, anything that can happen will happen a small but nonzero times.

Related: Scott Alexander’s Cardiologists and Chinese Robbers (a)

Tweets

“It occurs to me that of the 4 universities I have used, all 4 had open classrooms/lecture halls anyone could just walk into & learn, and all 4 had security mechanisms in the gym to stop people from using it. That says a lot about people’s priorities.” (a)

“Modest proposal: replace college w/gym. Equal signaling value Conscientiousness/conformity/discounting; cheaper; objective; health benefits; progressive not regressive; real RCT transfer to IQ, not hollow; positive externality for looks; increasingly useful in newer environments.” (a)

“One interesting thing about deep learning is that even as ever better results surface, everything we know about NNs is probably wrong. A short list (in rough chronological order)”

“Checklists are underrated: they’re like programs for life. Whenever I’ve written a checklist, whether for travel or writing an article or sending out my newsletter, it’s always been helpful. (You could try to memorize the procedure better with spaced repetition, but why?)” (a)

“self-regulation is as pervasive as water: whenever an industry or group or person does something beyond the legal minimum. Which is almost everything. Think ‘high trust societies’ vs ‘low trust’ ones.” (a)

Other things on Gwern’s site

Ordinary Life Improvements (a)

This is a list of “small ways in which my [Gwern’s] ordinary everyday daily life has been getting better since the late ’80s/early ’90s”. To pick a few:

  • the Internet/human genetics/AI/VR are now actually things
  • not being yelled at for tying up the phone line
  • electronics prices keep falling to the point where people whine endlessly online if a top-end VR headset or smartphone costs less in real terms than a Nintendo NES did in 1983
  • hearing aids are a small fraction the size, have gone digital with multiple directional microphones (higher-quality, customizable, noise-reduction), halved or more in price, become water-resistant, and even do tricks like Bluetooth
  • not getting lost while frantically driving down a freeway; or anywhere else, for that matter
  • most books and scientific papers can be downloaded conveniently and for free
  • search engines typically turn up the desired result in the first page, even if it’s a book or scientific paper; one doesn’t need to resort to ‘meta-search engines’ or enormous 20-clause Boolean queries
  • smartphones: far too much to list… (eg careless smartphone photographs are higher-quality than most film cameras from a few decades ago, particularly in niches like dark scenes)
  • we no longer need to strategize which emails to delete to save space
  • coats are thinner, more comfortable, and warmer thanks to better forms of synthetic fiber and insulation
  • LED lights are more energy-efficient, heat up rooms less & are safer, smaller, turn on faster, and are brighter than incandescents or fluorescents
  • clothing has become almost “too cheap to meter”; the idea of, say, darning socks is completely alien, clothing companies routinely burn millions of pounds of clothes because it’s cheaper than the cost of selling them, and Africa is flooded by discards

Gwern’s annual reviews

For example, here’s Gwern’s review of 2018 (a).

Annual summary of [YEAR] gwern.net newsletters, selecting my best writings, the best [YEAR] links by topic, and the best books/movies/anime I saw in [YEAR], with some general discussion of the year

Matthewsisms (a)

Those startups who least need VC investment & deliver lowest expected returns are those who receive the most VC offers of investment.

Those who most want a girlfriend are the least successful in asking girls out.

Those people with the most need for Conscientiousness & depression treatment are those least able to follow therapy like CBT.

Those schizophrenics with the most need for medication are those least likely to keep taking it.

The most ignorant who most need to follow advice & instructions are least likely to understand their need.

Countries which most need to fight corruption are least able to do so.

…Merit-based admissions aid/scholarships; being rich and getting a loan; applying for jobs…

Gwern’s collection of problems with animal studies (a):

On the general topic of animal model external validity & translation to humans, a number of opeds, reviews, and meta-analyses have been done; I would summarize them as indicating that the animal research literature in general is of considerably lower quality than human research, and that for those and intrinsic biological reasons, the probability of meaningful transfer from animal to human can be astoundingly low, far below 50% and in some categories of results, 0%. The primary reasons identified for this poor performance are generally: small samples (much smaller than the already underpowered norms in human research), lack of blinding in taking measurements, pseudo-replication due to animals being correlated by genetic relatedness/​living in same cage/​same room/​same lab, extensive non-normality in data, large differences between labs due to local differences in reagents/​procedures/​personnel illustrating the importance of “tacit knowledge”, publication bias (small cheap samples + little perceived ethical need to publish + no preregistration norms), unnatural & unnaturally easy lab environments (more naturalistic environments both offer more realistic measurements & challenge animals), large genetic differences due to inbreeding/​engineering/​drift of lab strains mean the same treatment can produce dramatically different results in different strains (or sexes) of the same species, different species can have different responses, and none of them may be like humans in the relevant biological way in the first place. So it is no wonder that “we can cure cancer in mice but not people” and almost all amazing breakthroughs in animals never make it to human practice; medicine & biology are very difficult.

There are several dozen links following that quote. One example: sex matters in experiments on party drug — in mice (a):

But although Georgiou followed his protocol exactly, she found that treated mice did not swim for any longer than mice injected with a placebo. When she and three female and four male researchers investigated this disconnect by performing the experiments, they discovered that the ketamine acted as an antidepressant only when it was administered by men.

Related: “Turns out, stress & pain interact. So if animals are stressed, they can look like they’re in pain when doing behavioral testing. So what stresses out rats? A lot of things” (a)

Book reviews

The Cultural Revolution: A People’s History, 1962-1976 (a):

The disappointment generates dissonance: many people genuinely believed that the solutions had been found and that the promises could be kept and the goals were realistic, but somehow it came out all wrong. (“We wanted the best, but it turned out like always.”) Why? It can’t be that the ideology is wrong, that is unthinkable; the ideology has been proven correct. Nor is it the great leader’s fault, of course. Nor are there any enemies close at hand: they were all killed or exiled. The cargo cult keeps implementing the revolution and waving the flags, but the cargo of First World countries stubbornly refuses to land. …

This opens up brand new avenues for elite conflict: proles can be used as cat’s-paws, there is incentive to create ever new ideological rationales to claim the moral high ground & strike first, while this same instability means no one is safe because what was once goodthink tomorrow becomes crimethink (“la révolution dévore ses enfants”) & old materials like diaries (or yearbooks) are radioactive waste, the cycle can be used as subtle loyalty tests to see who is the most subservient & eager to follow the latest fashionable nonsense, any apparent loosening where regular people genuinely speak their mind can shock the brainwashed elite & provoke a backlash of intensified ideological policing, brief admissions of fault by those in unassailable positions can be used to elicit admissions from others who can then be immediately purged having been damned out of their own mouths (a tactic also enabled by self-criticisms or attempts to pre-empt purging).

The paranoid yet logical answer is that there must be invisible enemies: saboteurs, counter-revolutionaries, and society remaining ‘structurally’ anti-ideological. No matter that victory was total, the failure of their policies proves that the enemies are still everywhere. [ah yes, China, cultural revolution] (“One man’s modus ponens…”) And the rot must go all the way to the top. (But, of course, not to the very top, as the actually powerful are too powerful to criticize; the emperor is—as always—innocent & benevolent & a benediction unto his people, and merely misled or betrayed by evil officials). In actuality, the middle’s evil incompetence & sabotage, in addition to the doubtless high levels of corruption (which may be much less than that of the top and often economically-efficient work-arounds), is merely a muddling through with a mix of ideology, pragmatism, and incompetence, and there is nothing to purge.

A History of Life-Extensionism in the Twentieth Century (a)

  • there are no simple interventions that can change average life expectancy by more than a few years or maximum life span at all
  • as a corollary, there is no single or small number of genetic or biochemical ‘master switches’ of aging, because if there, some of the thousands of interventions during the past 3 centuries of active scientific research would have flipped them directly or as a downstream effect, someone would have exceeded the Calment limit, or heritability estimates of longevity would be far higher
  • research proceeding on the basis of ‘identify a correlate of aging’ is effectively doomed: the signature feature of aging is that it is an exponential acceleration (the Gompertz curve) of mortality due to all causes ie. all organs are simultaneously becoming nonfunctional and losing homeostasis and efficacy, and these problems interact as well. Since the body is an absurdly complex dynamic system which, if drawn out as a causal network resembles the collected graphs of thousands of paranoid schizophrenics, the probability of any pair of variables being correlated is effectively 1 while the probability they are directly causally upstream/downstream of each other is close to zero. (The impressive thing is to find something which doesn’t correlate with aging, like blood magnesium levels.) It gets worse. Because the fallout from aging is destroying all bodily systems and impairing homeostasis, this implies there are hundreds or thousands of pseudo-interventions: interventions which deal with some downstream effect of aging and may help on that one thing, but nothing else. For example, if one fed amphetamines to an elderly mouse, it might act ‘young’ but it will proceed to die on schedule regardless. (This is the more abstract form of observing that curing cancer does not do much about curing aging.) This can very easily mislead one into thinking one is making progress and conducting important work: ‘I found a protein which correlates with aging and I even checked that it causally makes rats stupider by injecting it into random rats!’ These can both be true and yet I can be extremely confident that this will never lead to a useful anti-aging intervention or shed light on what aging is, and that one certainly cannot “start with an old cell, change its signaling, and make it behave like new again.” (Hence, we can predict that any exciting new discovery will turn out to experience an even more than usually severe ‘decline effect’ where the initial reports turn out to be driven by the usual methodological issues like sampling error & publication bias & non-randomized mice selection & breed-specific responses & mislabeled reagents & non-blinded evaluation & coding errors & etc or turn out to only be a pseudo-intervention on a symptom. This is because our prior for an intervention on aging is, at this point, extremely low and so all the alternative explanations are much more likely. Analogous to psychologists’ perennial quest to increase intelligence: no matter how good the study looks, it is more likely that the gains are inflated by bad methodology, the product of publication bias, not g-loaded & restricted to a few subtests, due to error or fraudulent data, or something else which in another context would look like mean-spirited raillery and desperate grabbing at straws, but when it comes to IQ gains, is, sadly, always the correct answer thus far.)
  • related: all-cause mortality is the king of endpoints; everything else can be cheated. When it comes to aging, ‘the treatment was a success but the patient died’ is unacceptable. If the treatment invigorated the patient but they died on schedule, it was not an anti-aging treatment. Use of proxies is the dark side of life-extension research: quick, easy, seductive, encouraging (one is constantly making progress), but a dead end.

McNamara’s Folly: The Use of Low-IQ Troops in the Vietnam War (a):

When forced through basic training by hook or by crook, further training generally proved pointless: there weren’t enough funds to pay for the extensive hand-holding, so the fancy education (direct instruction, apparently) McNamara put faith in either wasn’t enough or simply never happened in the first place. (Thus demonstrating the iron & brass laws of social programs - as well, program efficacy always declines as it scales up, because it must be run by exactly those people failing at the task in the first place for lack of resources/competence/incentives/meaningful-interventions.) Where education was tried, it turned out to be futile, and those who did train them found them too slow or too dangerous to trust. A man assigned to t-shirt printing shop was unable to understand alphabetization and had to pick out each letter for printing by scanning through the box one by one; a sergeant trained two men to drive military trucks somewhat successfully but they were too dangerous drivers to be used and were transferred out; another simply forgot to get back on the helicopters after a village search forcing a second retrieval mission; … another did kill his commander while on guard duty when he forgot to ask for the password before shooting; another forgot to put his rifle safety on (shooting a squad mate in the foot, who died); … (McNamara may have had good intentions, but in the social sciences, good results follow good intentions much as the rain follows the plow; which is to say, they do mostly by accident, and we find it easier to tailor our preferences to the results than vice-versa.) Only a few of the stories, like the recruit who was confused by having two left boots and two right boots but no complete pairs of boots, or the one who thought semen was urine, or the extremely-short man who received an honorable discharge & medical pension for contracting the terrible disability of ‘dwarfism’ in a war zone, or the draftee who tried to commit suicide “by drinking a bottle of Head & Shoulders shampoo” could be considered all that funny. Most are painful to read. (But educational, again, especially if you are in a high-IQ bubble and have a lack of empathy for what low intelligence means.) Once you’ve read some of these anecdotes, other anecdotes, like Scott Alexander’s experiences in Haiti no longer seem like such a stretch.

Titan: The Life of John D. Rockefeller, Sr (a)

The philanthropy transitions into an account of Rockefeller Junior, as he is entrusted with it, who emerges as diligent and effective, but not the man his father was. Senior attempted to replicate his own upbringing (without the - well-intentioned, intended to raise them properly without being corrupted by wealth - abusiveness), but as so often in dynasties, the founder’s extreme qualities do not fully carry over to his offspring, who regress to the mean.

The lesson I take away from Senior’s other, even more disappointing offspring (variously mediocre, wastrel, neurotic, or gullible) is that if you want to build a family empire, you must have a lot of offspring so the surviving maximum may be adequate, and also be willing to go outside direct descent or even adopt outsiders (eg the Romans or Japanese); this is the only way to keep a family business going for centuries. We just don’t know how to raise kids in a way which prevents them from easily turning out mediocre, dumb, insane, or unmotivated, once all the basics are provided for. …

The strategy of the rich, putting all their eggs into 1 or 2 baskets, is hopelessly fragile and a hostage to the slightest bit of bad luck. (Consider the Kennedys!) Why do so few of the rich & powerful not realize this and maximize their family size? I have to wonder. … Or perhaps it’s peer effects and nurture illusions: having more kids is what poor people do, a good rich parent has two children and makes sure they both get into Harvard by getting into elite pre-k and summer schools.

A reddit comment

On understanding correlations (a):

I can’t count how many times I have read a sociology paper which concludes ‘of course, whether X affects Y is a complex issue and correlation!=causality and more research is required’ on a known genetically or otherwise confounded trait and then turned to the presentation or media article and find the same author glibly saying ‘and this is why we need counseling or welfare programs to reduce the impact of X’. They mouth the words in the paper, but they don’t actually believe it; they don’t think correlations usually aren’t causal, they think it’s just a rare edge case which is a problem for other people who are not them, like p-hacking**, and their particular correlation surely is casual - ‘after all, it makes so much sense! How could X not cause Y? That would be absurd.’ (“I’m not saying it was aliens… but it was aliens.”) It’s your classic motte-and-bailey. This is also how we get proposals like ‘the FDA should only require trials showing safety, there’s no need to prove efficacy’. Like when you explain Turing’s halting theorem to someone and they respond, ‘yes, but couldn’t we still write a perfect virus checker if we tried really hard?’ (to give another personal example). They just don’t get it.

I’ve come to believe that correlation!=causation is one of those folk psychology situations, like Newtonian mechanics*, which everyone has been taught the right answer on some abstract level but then totally fail to internalize it at a gut level and act on it, and in every day life continue following folk theories. It’s something which is too hard for humans to get without first-hand experience, we’re too used to domains like every day life where sparse cause and effects follow each other reliably as when we wash dishes or play with Legos; whoever heard of being wrong about moving your fingers because of confounding or reverse causation? For so much of everyday life, it really is true that correlation=causation. You will your fingers to move and they move. But things like society and biology are too alien and interconnected and causal feedback so rare that our everyday intuitions (‘if you put too much stuff into your luggage, you cause it to be harder to move, so just stop putting so much stuff into it; if you get fat, it’s just because you eat too much, so just stop eating so much’) fail abysmally without our realizing; it reminds me a little bit of watching dual n-back’s effect on intelligence fall apart from 2008 to 2013 - I had read about and knew all about things like publication bias, expectancy effects, the garden of forking paths, problems with p-values, the need for meta-analysis and so on, but it took the bitter experience of seeing n-back fall apart in front of my eyes before I knew them and became properly cynical about psychology studies. And I still wasn’t cynical enough.

* there’s a depressing set of papers on Newtonian mechanics which shows you can take even grad students who know perfectly well the laws of motions, and ask them a few questions about how objects move, and they’ll still often respond in the intuitive Aristotelian manner

** “False-Positive Citations”, Simmons et al 2017:

We knew many researchers - including ourselves - who readily admitted to dropping dependent variables, conditions, or participants so as to achieve significance. Everyone knew it was wrong, but they thought it was wrong the way it’s wrong to jaywalk. We decided to write “False-Positive Psychology” when simulations revealed it was wrong the way it’s wrong to rob a bank.

PDFs on Gwern’s site

Investigating the right tail of wealth: Education, cognitive ability, giving, network power, gender, ethnicity, leadership, and other characteristics (a)

The extent to which people in the right tail of wealth are highly educated and cognitively able was examined in a sample of 18,245 ultra high net worth (UHNW) individuals with net worth’s of USD $30 million plus … Females were underrepresented, and female CEOs needed to be more select to reach the top of a company. Males and billionaires gave the most, but females and UHNW individuals gave more of what they had … The finance, banking, investment, and internet sectors dominated. Jewish individuals were overrepresented by a factor of about 234. Today, the typical UHNW individual profile includes U.S. married (Christian and Jewish) men who are largely Chairman and CEO, Republican, and earned their money in finance, banking and investments.

Also check Appendix K in the paper.

Androgens and the evolution of male-gender identity among male pseudohermaphrodites with 5alpha-reductase deficiency (a):

To determine the contribution of androgens to the formation of male-gender identity, we studied male pseudohermaphrodites who had decreased dihydrotestosterone production due to 5 alpha-reductase deficiency. These subjects were born with female-appearing external genitalia and were raised as girls. They have plasma testosterone levels in the high normal range, show an excellent response to testosterone and are unique models for evaluating the effect of testosterone, as compared with a female upbringing, in determining gender identity. Eighteen of 38 affected subjects were unambiguously raised as girls, yet during or after puberty, 17 of 18 changed to a male-gender identity and 16 of 18 to a male-gender role. Thus, exposure of the brain to normal levels of testosterone in utero, neonatally and at puberty appears to contribute substantially to the formation of male-gender identity. These subjects demonstrate that in the absence of sociocultural factors that could interrupt the natural sequence of events, the effect of testosterone predominates, over-riding the effect of rearing as girls.

Things written in response to Gwern

This reddit comment (a):

I often find myself quite certain we haven’t actually solved something that is supposed to be “solved”, or that we’re nowhere close to a solution that is nonetheless widely believed to be “just a matter of putting in the work”.

Usually this is because, when I think about the solution, it seems obvious that if we really understood x, we would also straightforwardly be able to (do, invent, understand, treat, exploit, sell) y and z. But we can’t. Ergo we clearly don’t understand x even if we have written textbooks about it.

Consider psychiatry as an example. (And I am not a Scientologist, nor do I consider it “useless”, etc.)

No one seems to emphasize the curious fact that when you go to a psychologist, their diagnosis is more or less a group of symptoms.

Usually you go to the doctor with symptoms (high fever and stiff neck), and then they determine the root cause (bacterial meningitis) and treat the disease (Rx antibiotics).

At the psychiatrist you go in with a list of symptoms (feeling depressed), and then they officially list your symptoms back to you (diagnosis: depression) and give you medicine to treat your symptoms (Prozac).

As a last side, one thing I don’t think people necessarily think about is that nearly every expert relies on the valuation of their expertise for money: therefore every expert has a strong case to oversell their expertise/the state of knowledge in their discipline.

Now, I haven’t a clue how common it is, nor how you would find it out.

But I often hear people ask, “But what motivation does a pure scientist have to lie to me?”

Well, the answer is that they make money because you see their expertise as valuable. I am not saying they are, in fact, doing so– just suggesting they certainly have reasons for wanting to do so, even if they do not.

I don’t know what the answer to that problem is, though; we have no choice but to trust experts. No one can gain expertise equivalent to an expert’s in everything they need help with; we always end up having to trust other people when they say that x or y is how things work.

“Death By 1,000 Clicks: Where Electronic Health Records Went Wrong: The U.S. government claimed that turning American medical charts into electronic records would make health care better, safer, & cheaper. 10 years & $36b later, the system is an unholy mess.” (a):

the subsidies are a bit of a pall hanging over the industry, though it’s only been about ten years. but because of government investment, this industry [EMR] mostly skipped the lean, innovative startup-esque period other software industries have. when people talk about emr problems this is what they should talk about, because these systems are unbelievably expensive — especially when you consider lost revenue from declining productivity during training. of course, with the subsidies come extreme and ever-changing regs that everyone has to adhere to. there is no free lunch.

the bottom line is that people demand health care inelastically. everyone wants perfect health care all the time no matter the cost. hit pieces will always be possible as long as even one brain tumor goes uncaught.

Appendix: important but not essential

Gwern’s site

Embryo Selection for Intelligence (a):

the best 2016 polygenic score could achieve a gain of ~3 IQ points when selecting out of 10 [embryos]

Banner Ads Considered Harmful (a) – Google AdSense reduces Gwern’s traffic by 10-15% (similar effect of ads found by Pandora, LinkedIn, & Mozilla (a)).

Nootropics (a) – important as a sheer demonstration of what’s possible and as a reminder to experiment more with ourselves:

A record of nootropics I have tried, with thoughts about which ones worked and did not work for me.

My Mistakes (a) – notably, looking at the last modified date of the page, it appears that Gwern has not made any mistakes in more than 660 days:

This list is not for specific facts of which there are too many to record, nor is it for falsified predictions … nor mistakes in my private life … nor things I never had an initial strong position on … The following are some major ideas or sets of ideas that I have changed my mind about.

Weather and My Productivity (a) – important as a reminder to think about how we personally are affected by factors with strong societal loading:

It would appear that weather does not correlate with my self-ratings to any detectable degree, much less cause it.

Related: “There is no clinically controlled evidence showing that interventional blue light, the kind that would be emitted from an LED, causes significant change in sleep quality with measures like minutes spent sleeping or an insomnia scoring system. I found this by searching “blue light” and “sleep” in clinicaltrials.gov. There are 4 or 5 studies on the matter, all negative results. The best way to interpret this evidence is that blue light will not make you sleep less or experience insomnia more.” (a)

About This Website (a):

Meta page describing gwern.net site ideals of stable long-term essays which improve over time; technical decisions using Markdown and static hosting; idea sources and writing methodology; metadata definitions; semi-annual web traffic statistics; copyright license

Links (a):

Who am I online & what have I done? Contact information; sites I use; computers and software tools; things I’ve worked on; psychological profiles

Open Questions (a)

Some anomalies/questions which are not necessarily important, but do puzzle me or where I find existing explanations to be unsatisfying.

The IQ Halo effect (a)

Comments

A comment (a) to Inverse p-zombies: the other direction in the Hard Problem of Consciousness (a)

“A further curiosity about morphine is that if it is administered before the onset of pain (for instance, as a pre-surgical medication) the subjects claim not to feel any pain subsequently (though they are not numb or anesthetized - they have sensation in the relevant parts of their bodies); while if the morphine is administered after the pain has commenced, the subjects report that the pain continues (and continues to be pain), though they no longer mind it.”

…Lobotomized subjects similarly report feeling intense pain but not minding it, and in other ways the manifestations of lobotomy and morphine are similar enough to lead some researchers to describe the action of morphine (and some barbiturates) as “reversible pharmacological leucotomy [lobotomy]“. …

Scopolamine and other amnestics are often prescribed by anesthesiologists for the purpose of creating amnesia. “Sometimes”, I was told by a prominent anesthesiologist, “when we think a patient may have been awake during surgery, we give scopolamine to get us off the hook. Sometimes it works and sometimes not.””

Book reviews

Savage Continent: Europe in the Aftermath of World War II (a):

The end of WWII left much business unfinished: Wages of Destruction covers in detail the slave labor forces drawn from conquered Europe which worked in Germany up until defeat, and the parlous food situation of Germany and Europe at large - so what happened after? With all these victorious horny occupation forces? With the slave laborers, and the Jews, and the guerrillas or partisans or thieves or black-marketeers? How were morals slowly restored after being corrupted by the exigencies of war and the struggle for survival, and what was seen as now possible after the Holocaust?

Extraordinary excerpts: Introduction-3 (a); 4,5 (a); 6-8 (a); 9-11 (a); 12-14 (a); 15-19 (a); 20-23 (a); 24-Conclusion (a). From the Introduction:

Law and order are virtually non-existent, because there is no police force and no judiciary. In some areas there no longer seems to be any clear sense of what is right and what is wrong. People help themselves to whatever they want without regard to ownership – indeed, the sense of ownership itself has largely disappeared. Goods belong only to those who are strong enough to hold on to them, and those who are willing to guard them with their lives. Men with weapons roam the streets, taking what they want and threatening anyone who gets in their way. Women of all classes and ages prostitute themselves for food and protection. There is no shame. There is no morality. There is only survival. For modern generations it is difficult to picture such a world existing outside the imaginations of Hollywood script-writers. However, there are still hundreds of thousands of people alive today who experienced exactly these conditions – not in far-flung corners of the globe, but at the heart of what has for decades been considered one of the most stable and developed regions on earth. In 1944 and 1945 large parts of Europe were left in chaos for months at a time.

From Chapter 9:

In 1988, for example, a Polish Jew named Szmulek Gontarz recorded an interview for the Imperial War Museum in London in which he admitted that he and his friends had taken revenge on Germans during the liberation, and had continued to do so for a long time afterwards. “We all participated. It was sweet. The only thing I’m sorry about is that I didn’t do more. Anything: throw them off trains. Wherever I thought I could take advantage, by beating them, we would. There was one particular instance in Austria. We stayed in stables, and there was a German officer hiding there. We found him, and we did exactly the same as they did to us: we tied him to a tree and we shot him. If you say to me now to do it, no way – but at that time it was sweet. I enjoyed it. There was no other satisfaction at that time that any of us could have had. And I’ll tell you now: I challenge any person in a similar situation who would not have enjoyed it … It was perhaps the only thing that it might have been worth to survive the war, to be able to do that. And the satisfaction was great.“50

Life in Our Phage World (a)

The world of phages is more than a little scary. They have been evolving for billions of years, their numbers are so vast every writer in this anthology resorts to scientific notation (and when they don’t, the numbers are so unfamiliar they look like typos: “By killing nonillions of Bacteria, they have major effects on global energy and nutrient cycles…”), and their generation time is as low as minutes, making for dizzying amounts of selection pressure and optimization - phages seem to have explored every possible way of attacking, subverting bacteria, replicating faster, compacting and making themselves more efficient, and won every arms-race bacteria started with them. …

why do our delicious nutritious moist mucal membranes like our noses not get eaten by bacteria? because there’s an even more incredible density of phages in mucus, 40:1, than out, 10:1 [are we all being farmed by phages??]

The Discovery of France: A Historical Geography from the Revolution to the First World War (a)

Where do I start… The extraordinary fact that until the 20th century, French was only a plurality language in France? The stiltwalking shepherds? The horrifying bits about drunken dying babies being carted to Paris by the ‘angel-makers’? The packs of smuggler dogs who smuggled goods in and out of France for their human masters? (Or the dog-powered factories?) The forgotten persecution of the cagot caste? The Parisian who sold maggots to fisherman, which he raised in his closet on a pile of cat & dog roadkill collected from the streets? The wars between rival villages? The commuting peasants who thought nothing of a 50 mile walk? The strange twists of fate that lead regions to specialize in particular wares? The villages of cretins or families who regard a cretinous child as a gift from god? The mapping of the hidden communication networks that spread rumor at the speed of a horse? The corvée system of road-building, so inefficient at points that transporting the materials to build 1 more meter of a road could destroy more than 1 meter of that same road? All of this and much more is to be found in Robb’s dizzying tour of France, past and present, a tour I found as entertaining as educational.

Excerpts: prologue, ch1 (a); ch2 (a); ch3 (a); ch4 (a); ch5 (a); ch6 (a); ch7 (a); ch8 (a); Interlude (a); ch9-10 (a); ch11-12 (a); ch13 (a); ch14 (a); ch15 (a); ch16 (a); ch17 & epilogue (a).

Bad Blood: Secrets and Lies in a Silicon Valley Startup (a)

I was perplexed at the time by the large Walgreens deal: Walgreens is a large, competent, sophisticated provider of pharmacy services, well capable of thorough testing; if Theranos was not what it was hyped up to be, how could Walgreens fail to notice? My assumption was that Theranos had done something clever to produce fake results (if not perhaps as clever as the FSB at Sochi). BB provides the answer, which is dismayingly mundane: Theranos bluntly refused to provide any kind of real validation or access to its machines, and some Walgreens execs were furious about it and correctly convinced Theranos was a fraud, but others were seduced by the vision, and the doubters signed on because they were terrified of forcing Theranos into the arms of CVS, which is a rivalry I had no idea about. (“Van den Hooff listened with a pained look on his face. ‘We can’t not pursue this,” he said. ‘We can’t risk a scenario where CVS has a deal with them in six months and it ends up being real.’ Walgreens’s rivalry with CVS, which was based in Rhode Island and one-third bigger in terms of revenues, colored virtually everything the drugstore chain did. It was a myopic view of the world that was hard to understand for an outsider like Hunter who wasn’t a Walgreens company man. Theranos had cleverly played on this insecurity. As a result, Walgreens suffered from a severe case of FOMO - the fear of missing out.” Who knew?)


You can send your comments to me by

If you enjoy reading my site, you can subscribe by email:

View previous campaigns.