Most PhDs are irrelevant

Sunday, November 19th, 2023

Elon Musk planned on studying material science at Stanford, Walter Isaacson explains (in his biography of Elon), after getting his physics degree from Penn:

Still fascinated by capacitors, he wanted to research how they might power electric cars. “The idea was to leverage advanced chip-making equipment to make a solid state ultracapacitor with enough energy density to give a car long range,” he says. But as he got closer to enrolling, he began to worry. “I figured I could spend several years at Stanford, get a PhD, and my conclusion on capacitors would be that they aren’t feasible,” he says. “Most PhDs are irrelevant. The number that actually move the needle is almost none.”


“I thought about the things that will truly affect humanity,” he says. “I came up with three: the internet, sustainable energy, and space travel.”


Musk had come up with an idea for an internet company during his final year at Penn, when an executive from NYNEX came to speak about the phone company’s plans to launch an online version of the Yellow Pages. Dubbed “Big Yellow,” it would have interactive features so that users could tailor the information to their personal needs, the executive said. Musk thought (correctly, as it turned out) that NYNEX had no clue how to make it truly interactive. “Why don’t we do it ourselves?” he suggested to Kimbal, and he began writing code that could combine business listings with map data. They dubbed it the Virtual City Navigator.


Nicholson, who had a PhD from Stanford, did not equivocate. “The internet revolution only comes once in a lifetime, so strike while the iron is hot,” he told Musk as they walked along the shore of Lake Ontario. “You will have lots of time to go to graduate school later if you’re still interested.” When Musk got back to Palo Alto, he told Ren he had made up his mind. “I need to put everything else on hold,” he said. “I need to catch the internet wave.”

He actually hedged his bets. He officially enrolled at Stanford and then immediately requested a deferral. “I’ve written some software with the first internet maps and Yellow Pages directory,” he told Bill Nix, the material science professor. “I will probably fail, and if so I would like to come back.” Nix said it would not be a problem for Musk to defer his studies, but he predicted that he would never come back.

Most kids get very little out of school

Tuesday, November 14th, 2023

Austin Scholar, a gifted teenager at the unconventional Alpha school in Austin, commented on X (Twitter) that “half of our graduating seniors score the EXACT same on standardized tests as the highest-performing 3rd graders,” suggesting that, “kids are literally wasting nine years of their lives,” accompanied by this table of math test scores for the top half of test-takers from kindergarten through 12th grade:

MAP Spring Mathematics Student Achievement Percentiles with Austin Scholar's Annotations
Many kids are getting very little out of school, but that’s not what this comparison shows. This comparison shows that gifted children can move through the curriculum far, far faster than they’re allowed to.

What shows that most kids get very little out of school is the curve of the median student’s performance, which climbs nicely for the first few grades and then almost levels off, so that a typical 12th-grader knows almost nothing more than a typical 8th-grader.

Since the typical student learns roughly nothing in high school, it’s a waste of everyone’s time to keep them there, going through the motions.

(School does provide childcare and juvenile detention, though.)

These are MAP test results, by the way. When you look at the lower and upper halves, the results are even more revealing:

MAP Spring Mathematics Student Achievement Percentiles 1-49
MAP Spring Mathematics Student Achievement Percentiles 50-99

The top 1st-graders outperform many 12th-graders.

The bottom 13 percent of 12th-graders are obviously not scholars, but one 7th-grader per couple classrooms (98th percentile) has the fundamental skills of a college-bound senior (90th percentile).

He came away with an impression that the bank was a lot dumber than in fact it was

Sunday, November 12th, 2023

Elon Musk’s superpower is chutzpah, although Walter Isaacson doesn’t put it that way (in his biography of Elon):

When Kimbal moved to Canada and joined Elon as a student at Queen’s, the brothers developed a routine. They would read the newspaper and pick out the person they found most interesting. Elon was not one of those eager-beaver types who liked to attract and charm mentors, so the more gregarious Kimbal took the lead in cold-calling the person. “If we were able to get through on the phone, they usually would have lunch with us,” he

One they picked was Peter Nicholson, the executive in charge of strategic planning at Scotiabank. Nicholson was an engineer with a master’s degree in physics and a PhD in math. When Kimbal got through to him, he agreed to have lunch with the boys. Their mother took them shopping at Eaton’s department store, where the purchase of a $99 suit got you a free shirt and tie. At lunch they discussed philosophy and physics and the nature of the universe. Nicholson offered them summer jobs, inviting Elon to work directly with him on his three-person strategic planning team.

Nicholson, then forty-nine, and Elon had fun together solving math puzzles and weird equations. “I was interested in the philosophical side of physics and how it related to reality,” Nicholson says. “I didn’t have a lot of other people to talk to about these things.”

Elon researched Latin American debt for Nicholson, but the bank wouldn’t fund his “sure thing”:

“He came away with an impression that the bank was a lot dumber than in fact it was,” Nicholson says. “But that was a good thing, because it gave him a healthy disrespect for the financial industry and the audacity to eventually start what became PayPal.”

Musk also drew another lesson from his time at Scotiabank: he did not like, nor was he good at, working for other people. It was not in his nature to be deferential or to assume that others might know more than he did.

Musk’s college-admissions test scores were not especially notable

Sunday, November 5th, 2023

Elon Musk is certainly bright, but he’s not superhuman, at least not according to his test scores and grades, as reported by Walter Isaacson (in his biography of Elon):

Musk’s college-admissions test scores were not especially notable. On his second round of the SAT tests, he got a 670 out of 800 on his verbal exam and a 730 on math.


During his first year, Musk got A’s in Business, Economics, Calculus, and Computer Programming, but he got B’s in Accounting, Spanish, and Industrial Relations. The following year, he took another course in Industrial Relations, which studies the dealings between workers and management. Again, he got a B. He later told the Queen’s alumni magazine that the most important thing he learned during his two years there was “how to work collaboratively with smart people and make use of the Socratic method to achieve commonality of purpose,” a skill, like those of industrial relations, that future colleagues would notice had been only partly honed.

He seldom finishes anything

Monday, October 23rd, 2023

Elon Musk was a good student, Walter Isaacson’s notes (in his biography of Elon), but not a superstar:

When he was nine and ten, he got A’s in English and Math. “He is quick to grasp new mathematical concepts,” his teacher noted. But there was a constant refrain in the report card comments: “He works extremely slowly, either because he dreams or is doing what he should not.” “He seldom finishes anything. Next year he must concentrate on his work and not daydream during class.” “His compositions show a lively imagination, but he doesn’t always finish in time.” His average grade before he got to high school was 83 out of 100.

After he was bullied and beaten in his public high school, his father moved him to a private academy, Pretoria Boys High School. Based on the English model, it featured strict rules, caning, compulsory chapel, and uniforms. There he got excellent grades in all but two subjects: Afrikaans (he got a 61 out of 100 his final year) and religious instruction (“ not extending himself,” the teacher noted). “I wasn’t really going to put a lot of effort into things I thought were meaningless,” he says. “I would rather be reading or playing video games.” He got an A in the physics part of his senior certificate exams, but somewhat surprisingly, only a B in the math part.

He was rejected by 16 out of the 18 colleges he applied to

Friday, October 13th, 2023

Stanley Zhong, 18, is a 2023 graduate of Gunn High School in Palo Alto:

Despite earning 3.97 unweighted and 4.42 weighted GPA, scoring 1590 out of 1600 on the SAT’s and founding his own e-signing startup RabbitSign in sophomore year, he was rejected by 16 out of the 18 colleges he applied to.


He was denied by: MIT, Carnegie Mellon, Stanford, UC Berkeley, UCLA, UCSD, UCSB, UC Davis, Cal Poly San Luis Obispo, Cornell University, University of Illinois, University of Michigan, Georgia Tech, Caltech, University of Washington and University of Wisconsin.

His only acceptances: University of Texas and University of Maryland.

He won’t be going to either of those:

Zhong just started his Google job this week,

Storytelling needs to be practiced, just like flying or marksmanship

Monday, September 11th, 2023

Ian Strebel and Matt McKenzie are intelligence officers in the United States Army and Navy, respectively, who have found that creating a compelling narrative takes practice, which traditional military training does not provide:

This can be a big problem for military intelligence professionals — they are trained to deliver intelligence, not to tell stories, so the stories that commanders tell themselves win out. Despite studies showing people are far more likely to remember stories than statistics, the military trains new intelligence professionals to brief intelligence through rote memorization and presentation of information. Neither of us ever received formal training in how to present information and intelligence as a story. This breeds uncreative military intelligence professionals concerned more with being “right” or having all the facts than whether their information is absorbed. Often, when information is presented in this manner, without context, commanders don’t remember what is important or, more importantly, why something is important.


Militaries have used wargames to train ever since Lieutenant Georg Heinrich Rudolf Johann von Reisswitz introduced the concept to an initially skeptical Prussian General Staff in the early nineteenth century. Very simply, a traditional wargame is a board game that simulates some aspects of military combat. The popular game of Risk is a very simple wargame, while chess can be considered as one of the oldest. Wargames can be successful mediums for training, in part, because the narrative holds players responsible for their actions and emotionally attaches them to the game’s results. Tabletop role-playing games are just the modern evolution of the classic wargame.


Tabletop role-playing games are unique from traditional wargames because the collaborative nature of the game means that almost anything can happen. The rules of these games only help structure the narrative and determine the consequences of actions. Players are free, even encouraged, to try anything they can imagine within the limits of that narrative. Most tabletop role-playing games have several rulebooks, but, as with military doctrine, the rules do not and cannot account for every eventuality. Instead, games such as Dungeons & Dragons rely on players’ creativity and flexibility to develop and adapt rules as they go. One of the most essential aspects of such games is the application of chance, usually employed by rolling various-sided polyhedral dice, which encourages out-of-the-box thinking for players and Dungeon Masters, especially in the face of catastrophic failure or, just as critical, catastrophic success.

These rules, when applied to wargames, can make them better — we have firsthand experience with this. Ian acted as an observer during a 2023 joint wargame using the Marine Corps’ Operational Wargame System. During the wargame, an experienced aviator wrestled with the decision of whether to use an exquisite munition to attack a threat reconnaissance drone or let the drone continue unimpeded. Recalling recent footage showing a Russian fighter jet dumping fuel on a U.S. surveillance drone, which downed the MQ-9 into the Black Sea, the aviator said he’d do the same. The wargame moderator said it was a “nice try” but that the move was outside the rules. If, instead, they’d abided by tabletop role-playing game rules, the aviator and moderator would play out the situation. Most likely, the moderator or Dungeon Master would determine, on the fly, the probability of the move’s success based on the game-defined attributes of the two aircraft and ask the aviator to roll a die. The Dungeon Master would use the die results to determine success or failure.

An experienced Dungeon Master might further adjudicate the results by applying a range of outcomes based on the die roll. For example, on a twenty-sided die, a roll of a “1” (critical failure) might result in the loss of the friendly aircraft with no damage to the drone, while a roll of “20” (critical success) might down the drone with minimal fuel loss and allow recovery of the drone sensor equipment. Rolls in between could result in varying degrees and combinations of damage and fuel loss to both the friendly aircraft and drone, as deemed reasonable by the Dungeon Master. Simultaneously, the Dungeon Master would determine the enemy’s reaction to this unanticipated event, both tactically and strategically, as well as the opposing force’s long-term adaptation to this move.

This is not so different from a military intelligence professional’s job: think like the enemy, understand their capabilities, develop possible scenarios, and then play the adversary as operators run through their plans. As previously discussed, while service intelligence schools generally teach presenting just a few courses of action, in a real conflict, there are infinite threat scenarios. Modern intelligence professionals must be flexible, responsive, and creative, in both planning and ad hoc operations. The problem is, short of an actual conflict, there are practically no opportunities for these personnel to practice working in a wide-open world — this is where tabletop role-playing games could prove valuable. As Dungeon Masters, military intelligence professionals can build worlds and scenarios and act as the enemy, or red, force. Most importantly, they will learn to respond spontaneously to unexpected player actions — regardless of whether those actions are incredibly clever or incredibly stupid.

In the Netflix series The Diplomat, Keri Russell succinctly described the problem of intelligence storytelling in three short sentences: “Intelligence is a story. A story based on incomplete facts. Life or death decisions turn on whether people buy the story.”


Dr. James Fielder explained that when games are designed correctly, a synthetic environment is created that becomes real to the players. In such an environment, the learning becomes real even if the risk is not — at least not yet. This is the challenge for both Dungeon Masters and military intelligence professionals. Telling a compelling story that enables others to envision combat environments and the threats within them accurately can be the difference between success and failure.


Storytelling needs to be practiced, just like flying or marksmanship. Pilots can safely make mistakes in simulators or with instructors in the cockpit. Shooters can miss targets on a range until they understand the weapon firing process. Similarly, Dungeons & Dragons provides intelligence personnel the opportunity to practice storytelling with the ability to make and learn from mistakes. After all, if a dragon kills a party of adventurers because the Dungeon Master wasn’t clear, they can simply try again. There are no second chances when giving an operational intelligence briefing before a strike mission.

Wargaming has seen a resurgence in professional military education, something we wholeheartedly support; games make learning fun, effective, and memorable. But integrating games into this education isn’t enough. The armed services only send a military intelligence professional to formal training a few times over a long military career. Comparatively, tabletop role-playing games can provide regular practice for the skills needed in exercises, wargaming, and the real world. After all, as James Sterrett, chief of the Simulation Education Division at the U.S. Army Command and General Staff College, said, “Experience is a great teacher and well-designed games can deliver experiences that are tailored to drive home learning.”

They make a better argument for a “free” Kriegsspiel or a Braunstein Game than for D&D, but the basic argument is sound.

He was President of the Society for the Suppression of Useless Knowledge, and for the Completer Obliteration of the Past

Sunday, August 27th, 2023

Jason Crawford recently read Samuel Butler’s Erewhon (1872), and it’s not really about Butlerian Jihad:

It is best known for its warning that machines will out-evolve humans, but rather than dystopian sci-fi, it’s actually political satire. His commentary on the universities is amazingly not dated at all, here’s a taste:

When I talked about originality and genius to some gentlemen whom I met at a supper party given by Mr. Thims in my honour, and said that original thought ought to be encouraged, I had to eat my words at once. Their view evidently was that genius was like offences—needs must that it come, but woe unto that man through whom it comes. A man’s business, they hold, is to think as his neighbours do, for Heaven help him if he thinks good what they count bad. And really it is hard to see how the Erewhonian theory differs from our own, for the word “idiot” only means a person who forms his opinions for himself.

The venerable Professor of Worldly Wisdom, a man verging on eighty but still hale, spoke to me very seriously on this subject in consequence of the few words that I had imprudently let fall in defence of genius. He was one of those who carried most weight in the university, and had the reputation of having done more perhaps than any other living man to suppress any kind of originality.

“It is not our business,” he said, “to help students to think for themselves. Surely this is the very last thing which one who wishes them well should encourage them to do. Our duty is to ensure that they shall think as we do, or at any rate, as we hold it expedient to say we do.” In some respects, however, he was thought to hold somewhat radical opinions, for he was President of the Society for the Suppression of Useless Knowledge, and for the Completer Obliteration of the Past.

Children are not relegated to a child-only world nor deemed too fragile to engage in difficult tasks

Monday, August 14th, 2023

Contemporary parenting in postindustrial societies is characterised by the idea that early childhood experiences are key to successful cognitive and emotional development:

The idea of parental influence is nothing new and, at a first glance, it seems rather banal: who wouldn’t agree, after all, that parents have some sort of influence over their children’s development? However, contemporary parenting (call it what you like: responsive parenting, natural parenting, attachment parenting) goes beyond this simple claim: it suggests that caretakers’ actions have an enormous, long-lasting influence on a child’s emotional and cognitive development. Everything you do – how much you talk to your children, how you feed them, the way you discipline them, even how you put them to bed – is said to have ramifications for their future wellbeing.

This sense of determinism feeds the idea of providing the child with a very specific type of care. As a document on childcare from the World Health Organization (WHO) puts it, parents are supposed to be attentive, proactive, positive and empathetic. Another WHO document lists specific behaviours to adopt: early physical contact between the baby and the mother, repeated eye contact, constant physical closeness, immediate responsiveness to infant’s crying, and more. As the child grows older, the practices change (think of parent-child play, stimulating language skills), yet the core idea remains the same: your child’s physical and emotional needs must be promptly and appropriately responded to, if she is to have an optimal development and a happy, successful life.

Like other such parents, in the first few postpartum months I also engaged, rather unreflectively, in this craze. However, when my son was four months old, during a period ridden with chaos, parental anxiety, sleep deprivation and mental fogginess, my husband and I made the decision to leave Europe. We packed our clothes and a few other things and hopped on a flight to Ecuador. Our final destination: a small Runa Indigenous village of about 500 people in the Ecuadorian Amazon. Our decision wasn’t as mad as it sounds. The Ecuadorian Amazon is where my husband grew up and where his family currently lives. It is also the place where I have been doing research for more than a decade. We wanted to introduce our newborn to our family and friends in the village, and we didn’t think twice before going. I could not yet imagine the repercussions this decision would have on me, both as a mother and as a scholar.

In the first weeks of our stay in my husband’s village, family and neighbours quietly observed how I took care of my son. He was never out of my sight, I was there always for him, promptly responding to (and anticipating) any of his needs. If he wanted to be held or breastfed, I would interrupt any activity to care for him. If he cried in the hammock, I quickly ran to soothe his cries. Our closeness soon became the subject of humour, and then, as the months passed, of growing concern. Nobody ever said anything explicitly to me or my husband. Most Runa Indigenous people – the community to which my husband belongs – are deeply humble and profoundly dislike to tell others how to behave. Yet it became clear that my family and neighbours found my behaviour bizarre, if not at times utterly disconcerting. I did not really understand their surprise nor did I, in the beginning, give it too much thought.

People, however, started rebelling. They did so quietly, without making a fuss, but consistently enough for me to realise that something was going on. For instance, I would leave my baby with his dad to take a short bath in the river and, upon my return, my son would no longer be there. ‘Oh, the neighbour took him for a walk,’ my husband would nonchalantly say, lying in the hammock. Trying desperately not to immediately rush to the neighbours’ house, I would spend the following hours frenetically walking up and down in our yard, pacing and turning at any sudden noise in the hope that the neighbours had finally returned with my son. I was never able to wait patiently for their return, so I often ended up engaging in frantic searches across the village to find my baby, under the perplexed stares of other neighbours. I usually came back home emptyhanded, depressed and exhausted. ‘Stop chasing people! He will be fine,’ my husband would tell me affectionately, giving me the perfect pretext to transform my anxiety into anger for his fastidiously serene and irresponsible attitude. At the end, my son always came back perfectly healthy and cheerful. He was definitely OK. I was not.

On another occasion, a close friend of ours who was about to return to her house in the provincial capital (a good seven hours from our village) came to say goodbye. She took my son in her arms. She then told me: ‘Give him to me. I will bring him to my house, and you can have a bit of rest.’ Unsure whether she was serious or not, I simply giggled in response. She smiled and left the house with my son. I watched her walking away with him and I hesitated a few minutes. I did not want to look crazy: surely she was not taking away my five-month-old son? I begged my husband to go to fetch our baby just in case she really wanted to take him away. When we finally found them, she was already sitting in the canoe, holding my son in her lap. ‘Oh, you want him back?’ she asked me with a mischievous laugh. To this day I am not sure whether she would have really taken him or whether she was just teasing me.

As an anthropologist, I admit, I should have known better. Scholars who work on parenting and childrearing have consistently shown that, outside populations defined as WEIRD (white, educated, industrialised, rich and democratic), children are taken care of by multiple people, not solely their mothers.


If the idea of an exclusive, preponderant relationship between mother and son might have seemed alien to our Runa family, equally strange, if not plain wrong, was the idea that a child’s needs should be always and promptly met by her caretakers. This is another central idea of current parenting philosophies: children’s emotions, needs and desires should be not merely accommodated, but also promptly, consistently and appropriately responded to. This translates into a form of care that is highly child-centred, whereby children are treated as equal conversational partners, praised for their achievements, encouraged to express their desires and emotions, stimulated through pedagogical play and talk, often with considerable investment of time and resources.


What these accounts, which claim roots in anthropology, fail to reflect is that, outside of postindustrial affluent societies, no matter how cherished, children are very rarely the centre of adults’ lives. For instance, Runa children, while affectionately cared for, are not the main focus of their parents’ attention. In fact, nothing is adjusted to suit a child’s needs. No canoe trip under a merciless sun is modified to meet the needs of a baby, let alone of an older child. No meal is organised around the needs of a young child. Parents do not play with their children and do not engage in dialogical, turn-taking conversations with them from an early age. They do not praise their children’s efforts, nor are they concerned with the expression of their most intimate needs. Adults certainly do not consider them as equal conversational partners. The world, in other words, does not revolve around children.

This is because children are not relegated to a child-only world nor deemed too fragile to engage in difficult tasks. From an early age, Runa children participate fully in adults’ lives, overhearing complex conversations between adults on difficult topics, helping with domestic tasks, taking care of their younger siblings. Participating in the adult world means that sometimes children can get frustrated, or denied what they want, or feel deeply dependent on others. At the same time, there is so much that they gain: they learn to pay close attention to interactions around them, to develop independence and self-reliance, and to forge relationships with their peers. Most importantly, in this adult world, they are constantly reminded that other people – their parents, their family members, their neighbours, their siblings and peers – also have desires and intentions.

Why are some periods and places so astonishingly more productive than the rest?

Sunday, August 13th, 2023

The most important question we can ask of historians, David Banks suggests, is Why are some periods and places so astonishingly more productive than the rest?

It is intellectually embarrassing that this is almost never posed squarely — I can think of only two articles (Gray, 1958 and 1961) and two books (Kroeber, 1944 and McClelland, 1961) that tackle this directly. But Gray is a lunatic, Kroeber waffles vaguely, and McClelland veers off into a fascinating but incomplete assessment. The question has never been the focus of professional attention in social history, although its answer would have thrilling implications for education, politics, science and art.


When I beard social historians at cocktail parties, they usually dismiss the problem of explaining excess genius as complex and ill-posed. But when coaxed into conversation, several ideas for facilitating factors come forward: Prosperity. They submit that a florescent culture needs the economic wherewithal to support the arts. Peace. They suggest that a climate of peace is also conducive to philosophical, artistic and (perhaps) scientific progress. (But recall Welles’ comment on Switzerland.)

Freedom. They believe that artistic freedom from state or religious control enables new growth.

Social Mobility. They think that when class distinctions are relatively permeable, then there is greater inducement for artists to excel.

The Paradigm Thing. They suppose that when a new medium or perspective arises, then art flourishes until the vein of originality is worked out.

All of these are good ideas, and superficially plausible. But most contradict the historical record.

To be specific, the prosperity suggestion fails for Athens, Florence and London. Athens spent its boom period in combat with Sparta; the income from the Delian League went to the fleet. Athenian farmers could not tend their crops (cf. The Acharnians), and such staples as grain had to be imported. Similarly, quatrocento Florence was poor compared to pre-plague Florence. The Medici bank had about half the capital of the Peruzzi bank in 1340, and Lopez (1970) documents other indications of reduced standards of living. A symptom of this desperation was the revolt of the populo minuto, which pushed the Medici into prominence. And Elizabethan London suffered “dearness without scarcity” (inflation); this fell most heavily on the aristocracy and the very poor. Then the wool trade collapsed, England entered “the worst economic depression in history” (Wilson, 1965), and Parliament anxiously debated means of averting a Bellum Rusticum.

Regarding the peace hypothesis, it clearly fails for Athens. Florence was torn by internal factions (e.g., il popolo grosso vs. il popolo minuto, the assassination of Giuliano de Medici, Savonarola). London had to contend with the Armada, the war with Spain in Holland, and internal religious dissent.

Regarding artistic freedom, the Athenian plays were written for religious festivals, and the prize was awarded according to the taste of respectable, pious and civic-minded judges (this caused Aristophanes and Euripides no end of trouble). In Florence, art was commissioned largely by the Church, sometimes by a patron, and had to voice themes prescribed in the contract. In London, note that Shakespeare’s plays avoid all mention of religion and contemporary politics; Marlowe and Jonson were similarly cautious (in literature, not in their personal lives).

Regarding social mobility, this hypothesis seems borne out by our three primary examples. Athens and Florence were both devaluing the aristocracy and promoting mercantilism. In London, the early part of the period clearly shows the rise of the middle class.

Regarding the emergence of a new paradigm, this is difficult to judge concisely. Much of the problem involves distinguishing a perturbation from an innovation. Did the introduction of a second on-stage character in Athenian plays represent a new paradigm? Was Plato’s decision to record philosophical discussion a minor influence on the content of the debate? Similarly, in Florence, painting and sculpture were well-established before the peak occurred, but the invention of perspective and the rediscovery of the classical period may have constituted a paradigm shift. Finally, in London, the key change seems to have been that small groups of strolling players discovered they could pack a hall in a city, and people would stroll to them. This enabled more elaborate props and larger companies, while pressing the need for a larger repertoire. But this kind of change is not especially Kuhnian in spirit, and the problem merits more lengthy consideration.

One could propose other factors. It seems to me that each of the three societies under consideration enjoyed a substantial military victory in the generation preceding their florescence. Athenians whose names shine today are reported to have prided themselves on being the sons of the men who fought at Marathon. Florence was not a military force (the Italian city-states relied upon mercenary condottieri in time of war) but in 1254 they conquered Pisa and Lucca. This secured an outlet to the sea, which was essential to their economic expansion. And in 1588, England conquered the Spanish Armada. This made the seas safe for colonial empire, and was a watershed for British morale.

Also, the great minds in each of these societies tended to hang out together. Socrates spoke with everyone. The playwrights talked shop, and the orators honed themselves upon each other. In Florence, artists trained under an apprentice system that pulled talents together, and Vasari describes frequent visits by the greats to each other’s studios. Leonardo and Michelangelo held a public contest over The Battle of Anghiari; meanwhile, the poets and philosophers clubbed together at Lorenzo’s mansion. In London, much of the theater circle met for drinks at the Mermaid Tavern, and one expects that their common profession ensured their lives crossed even more regularly.

Aubrey reports that Bacon visited the Mermaid Tavern too, and doubtless Bacon knew Raleigh, who was sufficiently friendly with Marlowe to rise to the Shepherd/Nymph bait. Does the social intercourse of good minds produce great minds?

A third possible factor is education. In each of the three societies, education tended to be as personal as a punch in the nose. In Athens, the upper class had tutors and the lower classes shopped for their educations among various freelance teachers. In Florence, the upper class had tutors and the masses learned as apprentices. In England, the upper class had tutors and the commoners learnt to write plays and poetry from each other, insofar as I can tell. All three of these systems emphasize individual instruction over the currently popular cattle drive approach. And there is ancillary evidence (cf. the lives of Wiener, Maxwell, Dirac, Russell, Mill, Malthus, Arnold, Feynman) that tutoring is enormously effective.

One can postulate many other factors. For example, it is suggestive that all three of Athens, Florence, and London had populations near 300,000. Also, all three had relatively democratic styles of government, and all three’s florescences were ended by right-wing revolutions (the Rule of the 400, Savonarola, and Cromwell). Finally, each of the three were in the process of reinventing their language — Periclean Athens defined the conventions of Attic Greek, Dante made Tuscan the foundation of modern Italian, and the linguistic gap from Chaucer to Shakespeare is enormously larger than the gap from Shakespeare to us (but this could be due to selection bias, since language might gel around great writings, rather than great writings arise from volatile language).

There is never any shortage of hypotheses. The useful trick is to know how to test them. In this case, one could rank a sample of cities in terms of their cultural IQ, and then decide whether the hypothesized factor obtains for each of the cities. If the factor is more common for the florescent cities than for the average or below average cities, then the hypothesis is supported (this can be made formally statistical). To an extent, this style of reasoning is what is used in this section, except that I haven’t elaborated the comparison by listing cities which have made meager cultural contributions.

(Hat tip to Byrne Hobart.)

Studying children in school is like studying orcas at Sea World

Tuesday, July 25th, 2023

Can a school make your child smarter?, Michael Strong asks:

Two hundred years later, after free public education has extended beyond the three years endorsed by Jefferson to thirteen years, we are seeing more skepticism around the benefits of education than ever before. The spirit of the times is quite different than it was when Jefferson was manifesting the Enlightenment in the U.S.

Freddie DeBoer has an essay “School Doesn’t Work” summarizing the ineffectiveness of a wide range of educational interventions


Economist Bryan Caplan’s book The Case Against Education makes the case that schooling is mostly signaling rather than adding human capital.


Between the emphasis on genetics, on the one hand, and the ineffectiveness of most education, we have reached a point at which many people believe that most K12 spending is a waste of money; Scott Alexander somewhat facetiously suggests that we give everyone the $150K (now more like $200K) we currently spend on each child so they can buy a cabin rather than waste time in school.


But clearly human performance can be improved. Anders Ericksson, the researcher most responsible for researching “deliberate practice” as a technique for improving performance, one of the world’s leading researchers of expert performance, has been arguing against the genetic determinists for decades.


From a young age, your child is biologically programmed to be a status optimizer within a given cultural milieu seeking niches for optimizing social status (including love and attention as good things!). In order to do so, they will reflexively imitate the behaviors of those who are regarded as prestigious, successful, and skillful in their social group, and whose sex/gender and ethnicity cues are something that she can emulate successfully.


These people (including Montaigne, Pascal, Mill, Bertrand Russell, Virginia Woolf, etc.) were raised by parents who were keen to immerse their children in environments in which interesting people were constantly around thinking and talking about ideas.

Today, teens with access to the internet who choose to devote themselves to achieving excellence for the sake of optimizing status among a chosen community of peers can learn extraordinary amounts without any classroom environment at all.


In the best circumstances, a school is a place where your child is exposed to a peer culture that supports learning. In the worst circumstances, the peer culture so undermines learning that all of the academic instruction becomes largely irrelevant.


Harvard’s David Perkins makes the case that much of what we regard as “intelligence” is a matter of dispositions towards thinking. Thus while it may or may not be the case that we can increase “g,” the underlying factor that is believed to result in high IQ scores, we can improve the ways in which minds think.


But as Carol Black, the screenwriter for The Wonder Years TV show notes, studying children in school is like studying orcas at Sea World. I regard almost all educational research as inconclusive garbage insofar as it is premised on schooling.

For instance, many people are excited by Bloom’s “2 sigma” finding, that students tutored one-on-one using mastery techniques performed two standard deviations better than students in a classroom environment. One reading of this is, “Wow, tutoring is powerful.” Another is, “Wow, classroom instruction is garbage.”


Schooling is a cultural monoculture that is far more damaging to human cultures than is agricultural monoculture on natural ecosystems. To shift from the Sea World metaphor, imagine studying Monsanto treated industrial wheat farms in Kansas and suffering from the illusion that one understood plant life.

The government tells children what to read, how much and when to exercise, how often to go to the bathroom

Saturday, July 15th, 2023

Alex Tabarrok is struck by how conservative and homogeneous schools are, regardless of their public or private status — which is exactly what struck me, too:

Private schools, despite having the autonomy, have not pioneered novel teaching methods. Montessori was innovative but that was a hundred years ago. A few private schools have adopted Direct Instruction, but how many offer lessons in memory palaces, mental arithmetic or increasing creativity?

I am enthusiastic about developments coming out of Elon Musk’s school and Minerva but it’s still remarkable how similar almost all private schools are to almost all public schools. The global adoption of a nearly identical education model is also disturbing, as I harbor significant skepticism that we’ve reached an optimum.

He agrees with Richard Hanania’d point that public education involves an extreme restriction of liberty beyond anything we usually accept:

The only substantial populations of individuals who have their lives structured according to time-place mandates in a free society like ours are prisoners, members of the military, and children. The mandates for children have gotten less strict over the years now that all states allow homeschooling, but opponents of school choice for all practical purposes want to do what they can to shape the incentive structures of parents so that they all use public schools (liberal reformers tend to like vouchers that can be used at charter schools, but not ESAs, which give parents complete control). Of course, children don’t have the freedom of adults, and so others are by default in control of how they spend most of their time. But it’s usually parents, not the government, that we trust in this role. Given the unusual degree to which public education infringes on individual liberty and family autonomy, the burden of proof has to be on those in favor of maintaining such an extreme institution.

This brings us back to the point of proponents of public education having to think that government is really a lot better than parents at deciding how children should spend their time. Is there a good reason to believe this is the case? Yglesias points to data showing that the evidence on whether school voucher programs achieve better educational outcomes is mixed. But there’s a lot more to childhood than maximizing test scores. In a free market system, parents would likely base their decision of where to send a child on a countless number of other factors: cost, safety, the pleasantness of the experience, the values that a school teaches, distance from home, which hours a school operates, extracurricular activities, etc. Parents who take their children out of public schools often cite a variety of reasons beyond likely impact on educational outcomes as measured by tests.

The more complicated and multi-faceted a decision is, and the more state control involves an infringement on individual liberty, the less we trust government to make it and the more we trust private parties. An American child spends almost 9,000 hours in educational establishments before graduating junior high. That’s more than what an individual would spend working at a full-time job for over four years. In the process, the government tells children what to read, how much and when to exercise, how often to go to the bathroom. This needs to be kept in mind when analyzing arguments and data.


To me, the true promise of the school choice movement isn’t that it might simply save a bit of money or avoid the worst excesses of public education. Rather, it presents an opportunity to rethink childhood. Ultimately, this can work against many of the pathologies that have emerged in American society over the last several decades, including delayed adulthood, high real estate costs, negative-sum credentialism that robs young people of their best years, and culture wars that are exacerbated by the fact that the children of people with radically different values are forced into the same institutions.

On what basis did we as a society decide that the ideal way to spend a childhood was to attend government institutions 5 days a week, 7 hours a day, 9 months a year, for 12 years? That most of that time should be spent sitting at a desk, with say one hour for lunch and one for recess?


I’m convinced the main reason we accept public education is the status quo bias. If someone proposed that any other population be placed in government buildings at set times organized by neighborhood and told what to do and think, people would recognize this as totalitarian. If told this was for their own good, citizens would demand extremely strong evidence for this claim and still likely oppose the program even if they found any evidence provided convincing.

They were told to exaggerate as much as possible the main error that the instructor identified

Thursday, June 29th, 2023

A persistent error is a sign that learning has occurred, but the student has inadvertently learned the “wrong” way really, really well and needs to unlearn it:

Researchers at the University of Verona (Milanese et al., 2008) conducted a study of thirty 13-yr olds learning how to perform the standing long jump in three sessions spread out over a three week period.


One group received instruction using an experimental teaching method called “Method of Amplification of Error” (MAE group). More details on this in a moment, but the tl;dr version is that this method involves doing things wrong on purpose, not just doing things correctly.

Another group received the traditional instructional method of verbal instruction (direct instruction group).

And the third group received no instructions at all, and just practiced on their own (control group).


The kids who received no instructions at all performed pretty much the same at both tests. They jumped 158.9cm on the first day of training and 160.6cm on the final day of training. A difference which isn’t statistically significant, and is pretty much what you’d expect.

On the other hand, the students who received instruction and feedback during training did improve over the course of three weeks. They started out at 159.4cm, and improved to 162.3cm by the final test – a gain of 2.9cm. And though an improvement of just over an inch may not sound like much, this would have been the difference between medaling and not medaling in the men’s long jump at the 2020 Tokyo Olympics.

And how did the amplification of error group do?

Well, the kids who were coached using the Method of Amplification of Error improved by an average of 20.4cm, going from 159.5cm on the first day of training to 179.9cm three weeks later. This is almost 7 times the improvement of the regular instruction group, and would have been the difference between Gold and Bronze at the same 2020 Olympics. In the same exact amount of training time!


On the surface, the Method of Amplification of Error training was not hugely different. The only difference was that instead of being instructed to jump with the correct technique, they were told to exaggerate as much as possible the main error that the instructor identified.


It seems pretty counterintuitive to practice doing something the exact wrong way, but the researchers explain that this actually deepens our understanding of what not to do and initiates an internal search for a better way to perform the skill.

Purposefully doing things extremely wrong provides us with a lot more information

Tuesday, June 27th, 2023

In a recent study, Arizona State University professor Rob Gray trained casual baseball player to hit a baseball the right way or the wrong way:

One group — the “right way” group — practiced hitting the ball the correct way. Their instructions were to “hit a hard line drive into fair play.”

And during their training sessions, a coach would observe and give them feedback on their technique and mechanics. And provide suggestions on how to improve their performance.

The “wrong way” group on the other hand, received no technical instructions and no corrective feedback during their training sessions.

And their hitting instructions changed from one training session to the next.

One week they were asked to “hit the ball as far to the right as possible.” Another week they were asked to “hit the ball as far to the left as possible.” Then they were asked to “try to pop the ball up in the air.” And then to “try to drive the ball into the ground.”

They were also asked to “hit a hard line drive into fair play” in one of their practice sessions, just like the right way group. But in five of their six practice sessions, they were asked to practice hitting the ball all the wrong ways.


Well, as you would expect, the right way group that got coaching and practiced hitting into fair play improved their hitting in several key areas.

Their batting average improved, they struck out less often, and they hit more doubles/triples/home runs than they did in their initial test too.

But the wrong way group, which spent 5/6th of their time practicing hitting balls into foul territory, and other undesirable hits also improved their batting average, strikeout percentage, and slugging percentage.

And they not only improved in these areas, but improved by a lot more than the right way group did!


The value lies in learning how to achieve specific undesirable outcomes, on purpose, with some consistency. Because it seems that purposefully doing things extremely wrong provides us with a lot more information about how to do things correctly, than trying to do things correctly and accidentally getting it slightly wrong.

The goal of productive failure is not to get the correct answer faster and more easily via shallower learning

Saturday, June 24th, 2023

Early floundering can lead to better learning:

Generally, teaching looks something like:

  1. Explain how to do something (lecture)
  2. Show students what it looks like in action (demonstration)
  3. Fix their off-target attempts, to help them minimize “failure,” and reward them for their successes (feedback)

This sequence tends to emphasize getting to the correct answer as expeditiously as possible.


A pair of researchers (Kapur & Bielaczyc, 2011) conducted a study of “productive failure” to see if early floundering would lead to better learning than the traditional teaching approach (“direct instruction”).


The direct instruction class began learning about average speed with a lecture.

The teacher explained the concepts, worked through some examples, encouraged questions, and had students solve practice problems.

Then they reviewed the problems and discussed the solutions.

For homework, they were assigned similar problems in their workbook.


The productive failure class was split up into small groups, and each was tasked with solving two complex problems…

They were given these problems with no teacher support or guidance, but simply allowed two class periods to try to solve each problem (4 classes total).

There was also no homework, though they did receive extra problems to work on individually when the group problems were complete (2 class periods).

After 6 sessions of working on their own, the class spent their final class session sharing their solutions and strategies with the teacher and each other.

Only then did the teacher finally explain how to solve these problems the “correct” way, and help the students go through their previous work, fix their mistakes, and ensure they could arrive at the correct answer.

Ultimately, the productive failure group spent 7 class sessions working on calculating average speed, just like the direct instruction group. But they spent most of these classes floundering on their own, and doing many things wrong. It was only during the 7th and final class that they learned the correct way to approach these problems.


As you can probably imagine, the direct instruction group did waaaay better than the productive failure group in the early stages of learning.

The direct instruction group averaged a score of 91.4% on their homework.

Meanwhile, the productive failure group performed miserably on their unguided attempts to solve the complex problems. Only 2 out of the 12 groups (16%) arrived at the correct solutions. And when they had to work on the problems individually, their average score of 11.5% was even worse.

But a very different picture emerges when you look at the groups’ performance on the post-test.

On the final test, the performance between the two groups flipped, and the productive failure group outscored the direct instruction group by a significant margin.

On the simple problems, the productive failure group earned an average score of 84.8% (vs. 75.3% for the direct instruction group).

And on the complex problem, the productive failure group earned an average score of 59.7% (vs. 42.4% for the direct instruction group).


However, in much the way that spaced, random, and variable practice lead to worse performance in the short term, but better performance in the long term, it seems that the goal of productive failure is not to get the correct answer faster and more easily via shallower learning (“unproductive success”), but instead, to cultivate a deeper understanding of the fundamental principles and various ways of arriving at a solution even at the expense of short-term performance.