Subscribe now

I went on a run in San Francisco today, and it smelled like autumn. It smelled like leaves beginning to decay and damp soil and woodsmoke.1 These are smells I associate with West Virginia in late November, where we spent every Thanksgiving of my childhood. The seasons are flatter and earlier out here.

I was back in West Virginia in early August, for the first time in seven years, and it was decidedly unautumnal. My aunt’s retirement party was humid and sweaty and wonderful. My mom wrote us a play to perform in Aunt Paula’s honor, and there were so many West Virginia pepperoni rolls that I was still finding them in my backpack and suitcase a week later.

We spent the next week in DC. I visited DIME at the World Bank and talked about how to design experiments on the effects of AI in poor countries; I saw colleagues and friends at CG, CGD, and GD. My buddy Sam moved from NYC to Baltimore and we helped build some furniture. It was a wonderful trip home.

I also have a few new pieces out. First, a piece with Karthik Tadepalli and Oliver Kim on global inequality in a world of transformative AI. Second, a post on VoxDev with multiple friends about the need for novel experiments using AI in poor countries. I also previously neglected to link to this post from last month on Keynes’ “Economic Possibilities for our Grandchildren,” which many people continue to misread and abuse.

I’ll likely be in SF for the rest of the month, and we’ll be hosting various events. Please do reach out if you’d like to meet!

Reading

A Month in the Country

I loved A Month in the Country (1980) by J.L. Carr. It’s a short, quiet novel set in rural Yorkshire in 1920; glorious passages of prose, deep, religious, painful, and so very relatable, whether or not you were at Passchendaele. It’s about the end of summer, and the beginning of autumn, and how we learn to be melancholy at that time.

The long end of summer. Day after day of warm weather, voices calling as night came on and lighted windows pricked the darkness and, at day-break, the murmur of corn and the warm smell of fields ripe for harvest. And being young.

The narrator is a WWI veteran uncovering a fourteenth-century Judgement wall-painting over the chancel arch of the village church to fulfill the desire of a recently deceased village notable. Also around are an archaeologist digging for the notable’s excommunicated ancestor, the grumpy village vicar, his beautiful wife, and many more small town Yorkies. It’s just one month, and quite idyllic.

If I’d stayed there, would I always have been happy? No, I suppose not. People move away, grow older, die, and the bright belief that there will be another marvellous thing around each corner fades. It is now or never; we must snatch at happiness as it flies.

The frame narrative is quite strong: him looking back from fifty years later, when the memories have turned into something of their own. I suspect Carr is very good at this. I’m looking forward to A Season in Sinji, about a cricket game on an RAF base in Sierra Leone during WWII where Carr served, for this reason, hopefully before my next trip to Freetown. Here’s a short biography.

At another interview, he was asked if there was any insanity in the family (these were different times) and replied: ‘I have a sister’.

Arabia Felix

Arabia Felix (1962) by Thorkild Hansen, about the 1760s Danish expedition to Yemen, the first European expedition to Arabia, is an elegiac, wintry, end-of-a-world book.

So on the ship on which guns are being got ready for their murderous debate, a man is completely absorbed in observing the path of Venus.

The story is about Carsten Niebuhr, the only survivor of the six-man expedition, as the rest fall around him, mostly of malaria. Niebuhr was the expedition’s cartographer and astronomer, the youngest man on the expedition, and the least proud. He also seems to have been the best chosen, given the wealth of what he returned with, though he never left Scandinavia again. He spent the rest of his life taking measures with his astrolabe from the villages around his home, able to walk a town for an hour and then describe every house in immaculate detail. Niebuhr wasn’t blind until fifty years later. While much of what Hansen bases his book upon is from the letters of the expedition members, some letters from their deathbeds, the clearest portraits of Niebuhr come from his son.

And then he told us that when he lay blind on his bed, all the pictures of the Orient appeared quite clearly to his inner eye; so there was nothing remarkable about his being able to talk about them as if they were something he had seen the day before. Similarly, the vast dome of the Asian night sky with its myriads of stars extended above him as he lay quite still; and this was his greatest delight.

The story is also about the philologist von Haven, the most prominent academic on the expedition, who bought arsenic from a Constantinople chemist to keep the rest in line. von Haven was a desperate man, resentful of the status inversion on an expedition of equals, and he not being first.2 von Haven was the first to die, and I think he expected that he would never die, that he was protected, as he had been so privileged and lucky.

With an 83% mortality rate, it’s clear that something didn’t go right. Hansen illustrates a few opportunities the expedition had to save themselves, including after the first two professors died. The remaining travelers could choose to push on to Sana’a in the summer mountains, or return to the coast and wait for the trade winds to bring European ships. Today, it’s obvious they should have gone to the mountains; that’s where the nomads were going anyway. But they had no idea how malaria worked. Niebuhr, the healthiest remaining, chose the coast, dooming the rest of the expedition. I admire his self-authorization, his ability to step up when his seniors had died so recently, knowing he is not a replacement as powerful as the recently deceased.

Forsskål, the expedition’s botanist, is strangely faint in the book, which is an output of how it is written and of Forsskål’s own personality and paranoia: most of his notes were in code, and have never been deciphered. A Swede on a Danish expedition, he was loyal to his mentor Linnaeus and their country, seeking to keep his discoveries from the funders. After his death Linnaeus wrote to the astronomer Wargentin, “Yesterday I received a letter from Forsskål, from the realm of the dead. . .”3 Linnaeus named a plant for him, and gave it an appropriate description.

The four famous descriptive terms, tenacissima, hispida, adhaerens and uncinata mean stubborn, wild, obstinate and angular. The one living memory of Peter Forsskål is the plant Forsskalea, a stinging nettle.

Hansen’s archival discipline, as well as his own ability to turn a phrase and the pellucid translation, bought him novelistic liberties. But the characters and their story, of death, deception, discovery, are what carried it. The register is often memento mori, and a strong element of the work, the discovery surviving and being “worth it” — because Niebuhr’s passion aligned with what was needed. His weeks alone at Persepolis, the first European to survey the ancient Persian capital, will stay, as will the images of these compatriots shaking with malaria, and Niebuhr returning home alone.

When he reached the farm near Altenbruch he dismounted, unpacked his astrolabe as a matter of course, set it up, measured the latitude of the place and calmly entered the result in his tables, where the town of Altenbruch still figures to-day in the same columns as Damascus, Jerusalem, Baghdad and Bombay.

Subscribe now

Links

Maciej Cegłowski went looking for the deep biosphere: “We have yet to drill a hole so deep that we don’t find a microbial population at the bottom of it.” What do they eat? It relies on Landauer: each bit of computation requires less energy, and you can rest for longer and longer times between bits.

Penn Jillette on the secrets of fire-eating:

You shouldn’t be doing this. It’s dangerous and if you do everything right, you’ll still get hurt. The definition of a stupid thing is something that if you do everything right, you still get hurt. Fire-eating and love are stupid things.

How is this… possible? All of the tropes in this 53-year-old fan fiction parody piece are also recognizable as modern parody. Also, the origin of “Mary Sue”.

Andrew Batson is really smart and entirely right about the accounting.

If China would just pursue a conventional macro policy of moderate inflation and full employment, its own citizens would benefit massively and its trading partners would be much happier. There would be peace on earth and goodwill toward men! 🙂

I didn’t think that a 2014 NYT interview with Jan Morris could upset me so much:

I would like to have written books about San Francisco and Lhasa, but circumstances have prevented it, and now it is too late.

New on demography and fertility , from Fernández-Villaverde and coauthors: “In short, modernity in its current form seems largely incompatible with replacement-level fertility.”

Our conjecture is that a modern society makes the third child expensive and childlessness cheap, even when GDP per capita is only a fraction of the frontier… At the same time, children are no longer productive in the household, not even for smaller chores, having been replaced by “engines of liberation” such as the electric appliances now available even to poor households.

Why are plumbers paid more in the US? Philippe Lemoine, patiently.

it’s often easier to talk about why such workers are paid more today than they were in the past in the same country despite not having become more productive in that sense, but as we shall see that’s essentially the same question.

The point of this article is that you’ve never heard of NBA YoungBoy. You should have!

I took this essay very seriously, even going so far as to use some of my old music industry contacts to reach out to YoungBoy’s team for a quote. Not that I expected him to respond, after all what is The Metropolitan Review to someone like NBA YoungBoy? What is any publication, big or small, to NBA YoungBoy? Nothing. Less than nothing.

For three years , I’ve used this paper, newly out in Management Science, as a counterexample to the raft of RCTs which find that AI reduces skill disparities. I previously thought it was a rich country/poor country thing, but newer experiments are complicating the picture. Here’s one new RCT from Texas also showing widening disparities.

The increase in appeal filing was smaller among less-advantaged households, providing suggestive evidence that the chatbot widened existing disparities.

In cyber defense, you’re either dealing with Mossad or not-Mossad.

If your adversary is not-Mossad, then you’ll probably be fine if you pick a good password and don’t respond to emails from ChEaPestPAiNPi11s@virus-basket.biz.ru. If your adversary is the Mossad, YOU’RE GONNA DIE AND THERE’S NOTHING THAT YOU CAN DO ABOUT IT. The Mossad is not intimidated by the fact that you employ https://. If the Mossad wants your data, they’re going to use a drone to replace your cellphone with a piece of uranium that’s shaped like a cellphone, and when you die of tumors filled with tumors, they’re going to hold a press conference and say “It wasn’t us” as they wear t-shirts that say “IT WAS DEFINITELY US,” and then they’re going to buy all of your stuff at your estate sale so that they can directly look at the photos of your vacation instead of reading your insipid emails about them.

Tyler Cowen thinks it might almost be time to go to Somalia: “I will repeat my recent point that many dangerous places in the world are about to become much, much safer.”

What backs government debt? NBER asked investors, voters, and econ graduates to explain why government debt has value. This is not reassuring.

Global safe-asset demand and permanent rollover together account for almost half of perceived value in every group, so debt value heavily relies on investor absorption rather than on government resources. Future primary surpluses, the textbook debt backing, receive about 16% even among the graduate-trained.

Amelia Michael on robots : how frontier models could leapfrog the standard robotics data-scaling path.

Robotics progress might be chugging along within one paradigm (e.g., collecting tons of teleoperation data), only for frontier models to accelerate capabilities through an entirely different paradigm, like using advanced simulation.

Reading

The trip to the east coast is partly to attend my aunt Paula’s retirement to-do from her post at WVU. This is my first time back in West Virginia in seven years, and I put some books on my reading list.

Subscribe now

Ramp Hollow (2017) by Steven Stoll came highly recommended. It’s a history of the dispossession of Appalachia’s smallholders over the past centuries; the story of mountain households and their makeshift economy among the forest commons. It was the loss of this, from coal and timber industries imposing wage work and company towns, which Stoll argues brought the poverty which is now associated with Appalachia. The history is not always careful; he writes that West Virginia’s population has fallen 40% since 1950, when it was closer to 8%.

History books written in the past decade tend to be more polemical than older books claiming the same spots in the genre. Or, more likely, the polemics of past decades are dog whistles to my human ears and go unheard. I have nothing against polemics! How to Blow up a Pipeline is one of my favorite books, and it wears its disdain proudly. Ramp Hollow refuses to acknowledge its anger, and as a result the interesting historical observations are pressed through an ideological lens I found oppressive and uninteresting. Stoll is unwilling to concede any interesting point.

All of this is really too bad, because so much of this book was tailored to my interests. The final chapter is the only one not about Appalachia: it is about Mali. Stoll analogizes his research on land dispossession in Appalachia to contemporary land reforms in the Sahara.

Enclosure is not past but present, especially throughout Africa. Since 2008, corporations, governments, and international development agencies have looked anew at territory that agrarians have occupied for centuries without formal ownership. Private property doesn’t exist in most of the fifty-four countries on the continent.

This section was simply bad. After enumerating the World Bank’s goals in education, child mortality, eliminating AIDS, and after considering how the World Bank has provided agricultural productivity grants to communities of subsistence farmers, Stoll declines, on the behalf of the farmers: “These are all admirable ends. But why does meeting a people’s basic human needs require that they shift to new ways of thinking?” I don’t believe that Stoll has any sense of what subsistence farming is; he spends pages on Sahlins’ “original affluent society“ and Conklin’s critiques of the green revolution. He ends the book with an argument that the federal government should encourage Appalachians to return to subsistence farming: he proposes a federal law:

The United States shall create a series of commons communities, each designed to include a specified number of households within a larger landscape that will be managed by them, the residents. This landscape will provide the ecological base for hunting and gathering, cattle grazing, timber harvesting, vegetable gardening, and farming.

In this, he is my enemy. Moving people out of agriculture is a good thing. They become healthier and happier. Fewer of their children die and more of them go to schools. Moving people out of subsistence agriculture is a very good thing. I am a liberal, and I want the people of Mali and West Virginia to be able to live the lives they choose. But I also do not want Steven Stoll to prescribe them lives of harmful stasis.

Elizabeth Catte’s What You Are Getting Wrong about Appalachia was far more my brand of polemic. The association of Appalachia with the opioid crisis, Trump’s 2016 election, and Hillbilly Elegy, she argues, is distracting from the true character and problems of the region. The wave of “Trump country” stories about West Virginian and Kentuckian hill counties was simplistic and degrading: these “pieces share a willingness to use flawed representations of Appalachia to shore up narratives of an extreme ‘other America’ that can be condemned or redeemed to suit one’s purpose.” The most egregious omission, she argues, is that of non-white Appalachia, and in this, she is most animated in her critique of Vance’s Hillbilly Elegy:

If it is appropriate to label a small but visible subgroup as unambiguously representative of 25 million people inhabiting a geographic region spanning over 700,000 square miles, then we should ask a number of questions. Where were the ‘Bernie Country’ pieces about Appalachia? There are more people in Appalachia who identify as African American than Scots-Irish, so where were the essays that dove into the complex negotiations of Appalachian-ness and blackness through the lens of the election? I associate contemporary eastern Kentucky with grassroots prison abolition, so where were the essays about how a presumed Trump victory would imperil that work? West Virginia has the highest concentration of transgender teenagers in the country, so why didn’t anyone examine this facet of ‘Trump Country’ and how the election might reverberate in their lives?

Both Stoll and Catte have a fundamentally negative view of Appalachian history: in Stoll’s telling, it is a history of dispossession; in Catte’s, the shortcoming is that Black Appalachians are not featured more prominently. They disagree about much else, everything except the shape of the story: the region is acted upon, never acting. Although land and narratives are repossessed, no one is building churches or unions. I finished these books knowing very little about what the West Virginians did.

My two favorite books on the state have been Death in Mud Lick by Eric Eyre — the use of funky Appalachian place names as hooks will never die — and West Virginia: A History, a very competent overview by John Alexander Williams. Mud Lick is a journalistic account of the opioid epidemic in Mingo County, on the Kentucky line, following the death of Bull Preece and the resulting investigations and lawsuits, championed by Preece’s family. These lawsuits, in the late 2000s and early 2010s, targeted doctors, pharmacies, and drug distributors for keeping West Virginians hooked on painkillers. The characters are vivid and active, and occasionally sympathetic.

Among people talking about and working on the future of AGI, I frequently hear about the possible imminence and immanence of material abundance and post-scarcity. My concerns and confusions about this topic are largely psychological and economic: what would it mean to truly push the prices of goods and services to zero? could our demand truly be satiated? But one interjection into these conversations which has become increasingly prominent and decreasingly interesting is that there are still millions starving in America, that 47 million Americans, 13.5% of households are food insecure.

This is not true, and the claims are a cruel distraction. Food insecurity is defined so loosely as to be meaningless; if you skip McDonald’s and eat carrots from home because a McDouble’s gone up to $4.69, you would be labeled food insecure.1 My own polemic is this: there is no hunger or malnutrition caused by low incomes in America today, and claiming otherwise sells short the war on poverty that defeated hunger. Our hungry, including in Appalachia, are a symptom of social and political problems, not economic. Legalizing gambling had a massive effect on hunger, creating almost 300,000 food insufficient households across just 9 states earlier this decade.2 Differences in the unemployment rate across time and place simply do not have this effect.3

Could we call Mud Lick post-scarce? The men and women who lived there could and often did live without work, they had access to the most advanced medical system on the planet, and they were as wireheaded as modernity allows.

Jan Morris is my favorite travel writer. The primary flaw in her work is that it’s very invested in time, as well as place, and the oeuvre being >30 years old means that I will not be able to experience the cities she wrote about in Hong Kong: Epilogue to an Empire (1997), Sydney (1992), or A Passage through New York (1969). Her books about timeless places, however, endure. Her Oxford (1965) is my favorite portrait of the city, and Venice (1960) is one of two books which has excited me about an Italian vacation. As the only journalist on the 1953 Everest expedition, her Coronation Everest (1958) is tremendous, and Nepal has also not changed, although the high Himalaya are fast going the way of Hong Kong or Sydney, rather than the perpetually amberized Oxford or Venice.

What made Morris powerful was her magical love of talking to people. Her books have many anecdotes which are surely false4 but when they were relayed to her with such truthfulness who could care. In her collection Contact! (2009), she pulled her favorite passages about people she had met from across her books.

On the Bund in Shanghai one evening a youth with the droopy shadow of a moustache confronted me with a kind of dossier. Would I go through his examination paper for him, and correct his mistakes? But I had been pestered by students all afternoon, and I wanted to go and look at the silks in Department Store No. 10. ‘No,’ said I. ‘I won’t.’

At that a theatrical scowl crossed his face, screwing up his eyes and turning down the corners of his mouth. He looked then, with that suggestion of whiskers around his chin, like a Chinese villain in a bad old movie, with a gong to clash him in. I circumvented him nevertheless and, ah yes, I thought, if the Gang of Four were still around, you would have me up against a wall by now, with a placard around my neck and a mob to jeer me, not to consult me about participles.

But my conscience pricked me, and I went back and corrected his damned papers after all.

It’s 200 pages of this, 297 excerpts from her books, and the last of her travel books before she settled down to rest in Wales.

Links

Maciej Cegłowski is the first blogger I ever called my “favorite”. That was fifteen years ago, and he was writing about traveling in Argentina, the Alameda-Weehawken Burrito Tunnel, and scurvy. After an almost-silent 6 years, he now has a substack about the economics of space exploration. Very up my alley!

There are some drawbacks to the nuclear bomb rocket.

For one, there’s the matter of launching any of this into space. Unless you start nuking Florida from ground level (and I’m not saying ‘no’), you need dozens of traditional rockets to lift this affront to God and Nature into orbit. Even then, its fission products will get caught in the Earth’s magnetic field and eventually spiral their way back home.

But probably the biggest drawback of nuclear pulse propulsion is the thousands of nuclear bombs.

This is the first thing I’ve read which makes me want to try diving.

Just like in a forest, the real magic happens once you go inside of it, once you dive. If you are deep in the water itself, you are past the scattering, so the water and the plankton can repeatedly filter the light to their combined spectral peak before it arrives at your depth.

You can fill nearly the whole gamut this way, but the BBC can’t show it to you on Blue Planet. It is more vivid than video can capture. Underwater photographers often use filters to block out blues, so that the whole scene doesn’t just clip against the limits of their sensor. These intensities of blues and greens are mostly unknown to the surface world, and beyond what we even have language to describe.

I saw a new color a couple months ago, a particular green of sunlight layered through maple leaves on redwood bark. It was very shocking. I’ll keep an eye out for this author’s blue-green whale:

In all of my hunting, there was one region of the colorspace I was never able to fill with a naturally occurring color, my blue-green white whale. From what I’ve been able to find, no natural process emits 520 nm light at sufficient purity to make it close to the very top of the colorspace. Bioluminescent fungus peaks at around that wavelength, but the mixture of other wavelengths it emits causes its color to register far below.

I’ve taken to William O’Connor’s travel blog, including for recommendations in London, DC, and Addis Ababa.

Hallways are the underwear of hotel design. You’ve got to get through them to get what you want. Often, you hardly notice them. Sometimes they’re grim. But once in a while, they get you excited about what’s coming after.

We have identified a Maya mathematician by name.

There’s a park in Dalston, near where we were staying in London this summer, where the toads were very talkative. One of my favorite Orwell pieces is about the toads of London.

At any rate, spring is here, even in London N. 1, and they can’t stop you enjoying it. This is a satisfying reflection. How many a time have I stood watching the toads mating, or a pair of hares having a boxing match in the young corn, and thought of all the important persons who would stop me enjoying this if they could. But luckily they can’t. So long as you are not actually ill, hungry, frightened or immured in a prison or a holiday camp, spring is still spring. The atom bombs are piling up in the factories, the police are prowling through the cities, the lies are streaming from the loudspeakers, but the earth is still going round the sun, and neither the dictators nor the bureaucrats, deeply as they disapprove of the process, are able to prevent it.

I had heard that there are only 41 theaters showing The Odyssey in the large-scale IMAX format in which it was filmed. But I didn’t know that this was up from 30 theaters, just because of the demand.

[Imax] spent more than a year before this release hunting projectors in decommissioned museum theaters, cannibalizing some for parts and rebuilding the rest, which is how the network grew from 30 rooms to 41 for The Odyssey.

Here are 24 fun facts about industrial robotics!

I remain confused about the effects of AlphaFold. It seems correct to have expected a large speed-up in biological and drug discoveries, yet we’ve seen very little. I saw this paper presented at Berkeley a few months back; they found a reallocation of effort but not a noticeable speedup. Something that the authors discussed which isn’t in this version is that they do not observe a change in training PhD students in experimental structural biology, or how they publish. Another paper presented last week also found reallocation. A team at DeepMind is somewhat more optimistic about the effect on total output.

David Oks has generally been worth reading, but I have been getting Gell-Mann’d more and more frequently. First it was his article about kinship ties in West Africa relying on a largely discredited literature, now it’s an essay about ambulance billing; both are deep fascinations of mine from my times in West Africa and ambulances. The ambulance essay is rather confused about the difference between 911 response, nonemergency transport, and interfacility transfers, which all have very different economics. David’s preferred model for emergency services is “ambulance insurance”, which has obvious adverse selection and low-acuity usage costs. My preferred model would be funding readiness through property taxes and charging per ride at marginal cost... which is already how American integrated EMS-Fire services operates. Of course, the actual problem with (urban) American EMS is the fire unions.

Subscribe now

For many ages to come the old Adam will be so strong in us that everybody will need to do some work if he is to be contented. We shall do more things for ourselves than is usual with the rich today, only too glad to have small duties and tasks and routines. But beyond this, we shall endeavour to spread the bread thin on the butter—to make what work there is still to be done to be as widely shared as possible. Three-hour shifts or a fifteen-hour week may put off the problem for a great while. For three hours a day is quite enough to satisfy the old Adam in most of us!
— J.M. Keynes, “Economic Possibilities for our Grandchildren”

People thinking about AI and the future of work keep bringing up this essay: “Keynes predicted that by 2030, we’d all be working fifteen hours per week.” This is cited to support seemingly every possible point of view: “Well, that just shows that people have been predicting this for a century and it hasn’t happened yet, so it won’t this time.” Or, “AI is arriving just in time to make him right!” Or, “we could be working fifteen-hour weeks, if it weren’t for the unequal distribution of capital.”

But Keynes never wrote that. When he mentions the possibility of a fifteen-hour workweek, it’s with dread, a solution to a problem he finds much more serious than the automation of work. And it was never a hope for how much his grandchildren would want to work; it’s a fear of how little they would get to work.

It’s so in the water that Tyler Cowen gave us a version in a talk at DeepMind earlier this month:

You may all know John Maynard Keynes in the 1930s, who wrote this famous piece, Economic Possibilities for our Grandchildren, and he said, they’re going to work 15 hours a week and enjoy leisure. That was quite wrong.

Tyler’s talk was about “Human Life in a Post-AGI World,” which is, coincidentally, exactly what Keynes’ original essay is about.

Subscribe now

Keynes was not making a prediction about work hours

Keynes makes a single prediction in Economic Possibilities: “I would predict that the standard of life in progressive countries one hundred years hence will be between four and eight times as high as it is today.” He nailed this one: come 2030, UK per capita income will be up about 5x, US per capita income about 6x, and the world about 7x.1

This prediction is unrelated to hours worked. Keynes gestures at near-term productivity estimates — “in our own lifetimes... we may be able to perform all the operations of agriculture, mining, and manufacture with a quarter of the human effort to which we have been accustomed” — but his vision for 2030 is much grander. He is predicting a complete solution to what he calls “the economic problem”, the “struggle for subsistence.” After the standard of living grows 4-8x, our absolute needs would be satisfied “in the sense that we prefer to devote our further energies to non-economic purposes.”

Everything downstream of that prediction is a thought experiment. If we run his chain to the end, the number of required hours it spits out is zero. The fifteen-hour workweek enters backward, from Keynes’ sense of his grandchildren’s psychology. He claims that “everybody will need to do some work if he is to be contented,” and the ration is sized to that need: “For three hours a day is quite enough to satisfy the old Adam in most of us!” Fifteen hours aren’t what’s left over, down from fifty, but what we get, up from zero.

In 2008, MIT Press published Revisiting Keynes: Economic Possibilities for Our Grandchildren, in which leading economists reconsider the essay’s claims about growth, work, leisure, and consumption. Most run an exercise natural to economists: score a forecast and explain the residual. Lee Ohanian’s essay in the volume tries to grab at what Keynes could have meant: “Keynes does not provide any details on how he arrived at this forecast, and this raises the question of what economic theory or quantitative procedures that he used to arrive at this number.” He calibrates a growth model, using Keynes’ own techniques, hoping to find the deep truth that Keynes had encoded in this fifteen-hour workweek hope. He couldn’t find it.

I’m of the opinion that the essay, and particularly the flight of fancy leading us to a fifteen-hour workweek, was a toy to Keynes. Robert Solow recommends treating the essay “as a jeu d’esprit, an occasion to speculate freely, cleverly, and even shockingly to a roomful of bright schoolboys, future bankers, and high civil servants.” It was his science fiction, a divertissement, as his biographer Skidelsky put it. A divertissement can be instructive, but it’s not a forecast.

Keynes was disappointed by the prospect of a 15-hour workweek

Fifteen-hour workweeks sound pretty good compared to today’s 30+. Tyler Cowen told us that Keynes’ grandchildren were going to work 15 hours a week “and enjoy leisure.” Gary Becker and Luis Rayo open their chapter in the MIT volume by calling the essay “a remarkably optimistic, and in many ways prescient, assessment of the long-term economic future of the Western world.” They’re right about the economics, in that Keynes expects growth to resume, compound interest to do its work, the “economic problem” to be solved. And philosopher Elizabeth Anderson, in an essay in Microsoft’s new Humanist Review, summarizes Keynes essay as growth enabling “rich countries to cut back everyone’s work to 15 hours per week. He hoped that in leisure, people would cultivate ‘the arts of life.’” Her essay is about purposelessness in an age without work, and she uses Keynes’ fifteen-hour workweek as one optimistic view! It is not.

“There is no country and no people, I think, who can look forward to the age of leisure and of abundance without a dread,” Keynes wrote. His optimism about growth is just an introduction to his claim that solving the economic problem creates a problem:

for the first time since his creation man will be faced with his real, his permanent problem — how to use his freedom from pressing economic cares, how to occupy the leisure, which science and compound interest will have won for him, to live wisely and agreeably and well. [...] It is a fearful problem for the ordinary person, with no special talents, to occupy himself, especially if he no longer has roots in the soil or in custom or in the beloved conventions of a traditional society.

Keynes looks around England in the 1920s and sees those living post-scarce lives — the rich, “our advance guard... spying out the promised land for the rest of us” — and reports back that “they have most of them failed disastrously.” The wives of the wealthy, “unfortunate women, many of them, who have been deprived by their wealth of their traditional tasks and occupations... yet are quite unable to find anything more amusing.” So once the rest of us are in this state, after a century of economic growth, “must we not expect a general ‘nervous breakdown’?”

We don’t know, and Keynes presents no evidence, if the idle are less happy than the working. Life satisfaction rises after retirement, and after winning the lottery, winners choose to reduce their hours worked. Other types of wealth and idleness lead to dissatisfaction, such as make-work jobs in rentier states. Keynes is not optimistic about the world after solving the economic problem, the world of his grandchildren.

Keynes’ fifteen-hour workweek in this context is a prescription, and closer to methadone than to pain relief. Everybody will need to do some work to stay contented, because the work-drive is “bred into him for countless generations.” I can hear Weber yelling, “What was that work ethic even for?” Keynes is sorry for his grandchildren’s fifteen-hour workweek: sorry that there isn’t more work available for them, and sorry that they will wish that there were.

Keynes’ 15 hours are bounded by labor demand, not labor supply

In 1928, the Economic Journal, which Keynes edited, published Frank Ramsey’s theory of optimal saving. Ramsey’s theory argued that people should accumulate capital until their consumption approaches satiation, a state Ramsey called Bliss. Keynes first presented Economic Possibilities the same year, and it borrows the word: the essay’s closing concern is the pace at which we reach “our destination of economic bliss.”2

Ramsey’s model is embedded in Economic Possibilities. We grow rich, we satiate, and we work less as the marginal value of consumption fades. That’s Edmund Phelps’ reading in the MIT volume, that Keynes’ framework “is essentially that of Frank Ramsey,” implying “the workweek shrinking over the centuries to next to nothing.” Phelps later argues that work at low levels provides “stimulation, challenge and personal development.” Joe Stiglitz frames Keynes’ paradox similarly (”why [have they] chosen to enjoy so little leisure”), and Richard Freeman answers that Keynes “missed the boat by failing to appreciate the power of economic incentives,” the rising cost of leisure. These economists find fifteen hours per week on a labor-supply curve, and also find that it failed to arrive because the curve isn’t shaped like Keynes thought.

But the fifteen-hour claim does not run on Ramsey’s model. In Ramsey’s bliss, there is no “old Adam” and no shifts or sharing of labor. As wants are satiated, work-hours fade to zero.

There are two numbers that matter: how much work people want to do, and how much work there is to do (simplistically, labor supply and labor demand). Today, the gap between these numbers is allocated by prices (wages), but Keynes’ imagined future instead rations them. This is why he talks about the “work there is still to be done,” and making it “as widely shared as possible,” in “shifts.” He is arguing that people want to work far more than needed, even after the economic problem is solved, for the sense of meaning which work provides.

We are being afflicted with a new disease of which some readers may not yet have heard the name, but of which they will hear a great deal in the years to come—namely, technological unemployment. This means unemployment due to our discovery of means of economising the use of labour outrunning the pace at which we can find new uses for labour. But this is only a temporary phase of maladjustment. All this means in the long run that mankind is solving its economic problem.

The depression England hadn’t yet entered in 1928, and the leisure-ful world of his grandchildren are bookends. In Keynes’ time, needs were unsatisfied and new uses for human labor remained on the horizon, but all of this is temporary. By 2030, he supposes, there is a shortage of work itself.

Unemployment, wanting more work than the economy provides, was not exotic to Keynes; it’s the most ordinary tragedy in economic history. Keynes was the master of the job shortage, and would spend the rest of his career stimulating demand to source new jobs. The shortage he foresaw for his grandchildren is different. The people of both 1930 and 2030 wished for work, but only the first’s wish could be granted. Keynes imagines that his grandchild’s wish for work could not be granted at any wage, because no wish would conjure a sixteenth. Keynes’ fifteen hours is a view of the demand for labor running down permanently.

A fifteen-hour workweek as late-stage capitalism

Economic Possibilities is a strange Keynesian historical materialism. Civilization, in his telling, equipped itself with vices to survive through the economic problem, that of survival and subsistence, and once the problem is solved, that equipment should be tossed aside.

For at least another hundred years we must pretend to ourselves and to every one that fair is foul and foul is fair; for foul is useful and fair is not. Avarice and usury and precaution must be our gods for a little longer still. For only they can lead us out of the tunnel of economic necessity into daylight.

His anxiety is that, in the coming world of post-scarcity, the equipment will not go quietly. This is the old Adam, the unregenerate nature of man, which the Church asks to be “buried, that the new man may be raised up” in the moment of baptism; the old Adam whom God cursed “in the sweat of thy face shalt thou eat bread”; the old Adam whose work is no longer needed in a world of material abundance. Old Adam is what keeps us working so many hours in our time of material abundance. Derek Thompson’s essay on “workism” ribs Keynes’ fifteen-hour “prediction,” but describes the same mechanism Keynes is chasing, work having “morphed into a religious identity — promising identity, transcendence, and community, but failing to deliver.”

Now is the time we least have to work, but we face the loss of work with dread. People have always dreamed of material abundance; Jannah’s four rivers, or Thomist heaven where the only work or want is Praise, or Cockaigne where “every man takes what he will, / As of right, to eat his fill,” or the hobo ballad first recorded in 1928, the same year of Economic Possibilities:

There’s a lake of stew and of whiskey too
You can paddle all around it in a big canoe
In the big rock candy mountain

Modern post-scarce imaginings are less cheery. Iain Banks’ Culture is peopled by characters drowning in ennui, as their superintelligent overlords try to manufacture purpose for them; Star Trek consistently avoids the social and economic implications of its replicators, focusing instead on the role of rank and directive. Every modern Cockaigne we have furnished with work.

Keynes’ late-stage capitalism does the same. Goods become abundant, work becomes scarce, our desire remains. “For many ages to come the old Adam will be so strong in us that everybody will need to do some work if he is to be contented.”3 Because we cannot abolish workism, we will ration what remains: “we shall endeavour to spread the bread thin on the butter.” In times of deprivation, it was always butter rationed, the fat, never bread, the staple. Keynes imagines a time when we have fat in abundance but are quickly running out of the staple, because we realized that the staple is what we really wanted.4

The grandchildren may yet be us

But beware! The time for all this is not yet.

I wrote this essay because I was frustrated that people read Keynes’ fifteen-hour workweek as an optimism unfulfilled, when it is instead his vision of a temporary and difficult period before we “rid ourselves of many of the pseudo-moral principles which have hag-ridden us for two hundred years.” I was frustrated because there is an optimism unfulfilled elsewhere in Economic Possibilities.

I see us free, therefore, to return to some of the most sure and certain principles of religion and traditional virtue—that avarice is a vice, that the exaction of usury is a misdemeanour, and the love of money is detestable, that those walk most truly in the paths of virtue and sane wisdom who take least thought for the morrow. We shall once more value ends above means and prefer the good to the useful. We shall honour those who can teach us how to pluck the hour and the day virtuously and well, the delightful people who are capable of taking direct enjoyment in things, the lilies of the field who toil not, neither do they spin.

Labor automation accelerates. We spend our days talking about AI exposure, O*NET scores, offshoring, reshoring, retraining, post-training, post-scarcity. Our productivity rises, our consumption rises, our anxiety rises. Every thinkpiece asking if AI will bring us Keynes’ fifteen-hour workweek is shooting for the moon and missing Keynes saying, perhaps the stars, and no one is asking if AI will allow us the lilies of the field who toil not, and are glad for it.

If we are so unlucky that the current wave of AI automation brings a fifteen-hour workweek,5 rationing work is Keynes’ policy for keeping us sane and peaceful during the transition.

I’ve recently been talking and reading a lot about rentier states. There are an unlucky few countries which have been blessed with natural abundance such that they become wealthy without any need for their citizens’ labor. The Gulf states have their oil, the Pacific islands their phosphate, and imperial Spain a glut of Potosí silver, with which Keynes opens Economic Possibilities: “For I trace the beginnings of British foreign investment to the treasure which Drake stole from Spain in 1580.”

The resource curse is a central puzzle of political and economic science, and its subjects are partial versions of Keynes’ experiment, income without labor demand. Some of these states choose the liberal option, to give their people autonomy. Other rentier states provide work. The UAE and Qatar both have public sector employment above 80%; Nauru, at its height, had public sector employment of 98%.6 Fake jobs are politically durable where cash is fragile; they give workers purpose and investment in the system. They are a far more durable means of buying political acquiescence than cash: if the funds are mismanaged or the resources run out, and the money funding this autonomy disappears, the fragility of a cash-based settlement breaks.

How do we reconcile these two? Satiating the old Adam is important for a functioning society and for happy people, yet it is inherently paternalistic. Keynes’ position is that the methadone of a fifteen-hour workweek is degrading and that we need it anyway, for a great while. But not yet: “it will remain reasonable to be economically purposive for others after it has ceased to be reasonable for oneself.” For now, the shortage of work is nowhere near universal: the economic problem remains unsolved for billions. An old Adam who needs a shift can have a real one.

The ration and the real work buy time for the assignment Keynes sets us, that in the meanwhile, “there will be no harm in making mild preparations for our destiny, in encouraging, and experimenting in, the arts of life as well as the activities of purpose.” Keynes wrote Economic Possibilities with pity and dread for grandchildren who would not have enough work to sate them. If the grandchildren are us, if our work and purpose is going away, we are luckier than he feared for now: there is still activity of purpose, and preparation for the arts of life. And after both, the lilies.

The distinction between Google and Google DeepMind is real, but it is not quite the distinction between two companies. It is more like the distinction between a large state and an unusually autonomous city within it: shared resources, shared interests, and a stubborn local identity. This dates back to Google’s purchase of DeepMind in 2014:

Google had allowed [Demis] to retain DeepMind’s independent culture in London. Google had even granted his followers a privileged status: DeepMinders could get into any Google building globally at the swipe of a key pass, and help themselves to the free food; but Googlers were barred from DeepMind’s premises. (Mallaby 2026, p.162).

There are a few differences between Google and Google DeepMind offices, but the most consistent is the type of bacon served in the office cafeteria. All of the DeepMind cafeterias I’ve visited serve back bacon, what one would traditionally find in a full English breakfast, while the Google cafeterias serve streaky bacon, the most American of cuts.

After three weeks at DeepMind, I love the smart people and independent atmosphere on my research team. For a good sense of what’s important to our team now, the team lead, Iason Gabriel, was profiled by The Guardian last week. I’m very lucky to have a lot of flexibility to work on what I’m interested in: the effects of transformative AI on lower and middle income countries.

It’s hard to think and write about the economics of AGI because AGI can break a lot of economics. The assumptions fundamental to many of our models aren’t a good fit for the speed of AI agents or the scalability of compute.

The most productive approach is to focus on mechanisms: we may not know how the world looks after AGI arrives, but we can assess which levers the technology pushes. Last month, at a conference, I saw two presentations about how AI will affect the size of future firms. In the first, Seb Krier talked about how AI agents speed up bargaining and contracting between specialists, making it easier for each specialist to operate as their own firm. In the second, Erik Brynjolfsson presented his work with Zoe Hitzig on AI and central planning: improvements in accessing and processing data makes centralized planning more feasible.

Both of these are valid and useful economic mechanisms, but they make opposite predictions about firm size. Much of studying AGI’s economic effects is like this. There are viable stories about why AGI might push interest rates up, or down. The debate over Jevons paradox is like this. At DeepMind, I’m going mechanism hunting. Over the rest of the year, I hope to identify the most important ways AI will affect welfare in poor countries, and maybe figure out what to do about it.

I’ll be in London through the rest of July, and back in San Francisco in August. As always, I’d love to chat with anyone interested!

Reading

Elleman’s Modern Chinese Warfare is an overview of eighteen conflicts fought by the Chinese in the 19th and 20th centuries. The best parts are really good: Elleman introduces each chapter with a very brief summary of the conflict, and uses each conflict to draw conclusions, some narrow, some broad, on how China has used war in modern times.

This is not, however, a detailed military history, but a history of how China and its enemies use coercive diplomacy. One chapter, on the Xinjiang arm of the Muslim rebellion of the 1860s, breezes over the entire military conflict which caused roughly 50,000 deaths to focus instead on the aftermath’s maneuvering between the Russian Empire and Qing diplomats. And this is a lightly bizarre diplomatic world. The territory at issue, the Ili valley, was contiguous to both the Russian and Chinese empires, but thousands of miles from either capital. When the first terms were revealed in Beijing and deemed unacceptable, the Qing negotiator in St Petersburg reportedly answered: “The treaty having been clearly settled, it is absolutely impossible to re-negotiate.” He was, of course, wrong, and was executed upon his return to Beijing.

Elleman also had a novel take on the Sino-Vietnamese War. In 1979, China invaded Vietnam, and ended the war with a unilateral ceasefire after 27 days. The narrative I’ve always told is that this was a serious Chinese defeat.1 Elleman, however, puts the war in context with the Indian border dispute in the 60s and the limited territorial war with the USSR in 1969. Throughout the 1970s, and after these conflicts, China perceived an increasing risk from the Soviets, and by 1979 was willing to accept a bad campaign to reduce this risk. Elleman claims that the goal of the invasion of Vietnam was to “discredit Soviet assurances of military support to Vietnam”.

This is a novel claim! The Chinese were understandably afraid of a two-front war, and the Soviet mutual-defense treaty with Vietnam make that increasingly likely. In fact, the Sino-Vietnam War seemed to be purely for a Soviet audience. Elleman points out that most of China’s best units were moved to the northern border directly before the invasion. China was perfectly happy to gain no territory and lost tens of thousands of troops, to prove that Soviet defense guarantees had no teeth.

There’s another, almost fantastical element to the war. Elleman links the conflict to the 29th “anniversary of the signing of the 1950 Sino–Soviet Treaty of Friendship, Alliance, and Mutual Assistance on 14 February 1979.” This treaty had a thirty-year expiry, and Beijing was certainly uninterested in renewing it. Starting a war with a Soviet ally on the anniversary of the signing of a peace treaty with the USSR feels, to me, very mandarin. The anniversary timing, minimal and cheap force allocation, and willingness to spend lives to alter other states’ beliefs make the Sino-Vietnamese war look less like the well-understood war for territory than a very loud diplomatic signal.

The book is not particularly relevant to Chinese warfare in the near future. There is practically no naval presence after the Sino-French war, nor air power. But the book does explore China’s use of war-as-communication clearly, and this is a useful frame to keep in mind when interpreting their military actions.

My most anticipated book of 2026 is Zach Weinersmith’s Sawyer Lee and the Quest to Just Stay Home. Yes, it is aimed at ages 8-12, but his Bea Wolf was aimed at five-year-olds and that was one of my favorite books of 2024. Sawyer Lee is Weinersmith doing Wodehouse-for-kids, and I am a huge Wodehouse fan. I can’t imagine this being a flop. Sawyer Lee was released in the US in late June, but it isn’t out in the UK yet. My copy is waiting for me back in San Fransisco.

I see why Weinersmith is also obsessed with Wodehouse. His explanation is that Wodehouse’s books have the most “complex and intricate plot work” in fiction. On the level of individual sentences, there is no one funnier in 20th century prose.

There are only two other authors whose farces I find as delightful and polite as Wodehouse. Connie Willis doing comedy is up there, most notably in To Say Nothing of the Dog. But Eva Ibbotson can actually hold her own against Wodehouse. In June, I read a collection of her short stories, A Glove Shop in Vienna. Ibbotson is a foil to Wodehouse. Her plots are rarely as complicated, and she’s perfectly willing to use a deus ex machina to resolve a love story or misunderstanding, whereas Wodehouse’s only deus is the butler.

Ibbotson’s protagonists are young women, universally, they’re cheerful central European refugees, often Jewish, living in England around the World Wars. This is, of course, her own life. She was born in Vienna in 1925, moved to London when she was ten, and married a professor at Cambridge University. She is deeply in touch with cruel history, and this history is often the inciting event in her stories and novels. One heroine is a Russian countess fleeing the Bolsheviks; another was sent to London by her parents after Kristallnacht.

The rest of her stories are rarely cruel, and the rare cruelty is often to a purpose.2 The horror of the wars and refugees don’t exhaust the supply of weird, selfish, erotic, generous people. Her stories are about women finding homes for themselves among these people. For her refugees it’s a mixed ending. Some manage to recover part of their lost world in England; others assimilate into their new world. But most create for themselves a better and more mixed little civilization.

Subscribe now

Links

The reviews of the Obama Presidential Center have been quite negative, although I hear the food is good. It seems like a big miss, especially the architecture.

Here’s a lovely essay about river tubing in Louisiana. It has beautiful imagery.3

You try to keep your distance from the convoys of high-school tubers, who tend to float in circular formations, like threatened wagon trains, around stashes of illegal beer. Occasionally you wave, with veiled condescension, to a fleet of passing canoers, trapped in their aluminum hotboxes and actively assaulting the river with oars.

Sandy Magnus, an astronaut who was kind to me often when I was in the space industry, has a wonderful travel blog. Last month she did my favorite hike, the Hadrian’s Wall path across northern England. What a lovely place.

Relatedly, did we ever get an explanation of the Sycamore Gap vandalism? Apparently not. But the two guys who cut down the famous tree were sentenced to 4 years in prison. They were both in their 30s, not kids as I had expected. Apparently there’s a video.

Is it possible that I was born at the perfect time to experience a radical abundance in floating balloons at birthday parties? The US government managed a National Helium Reserve, starting in 1925, for blimps and, later, rocket coolant.

In 1996, Congress ordered the privatization of the helium industry, and the Reserve sold down its assets below market price from 2005 to 2025.

Professor Richardson also believes that party balloons filled with helium are too cheap, and they should really cost about $100 (£75) to reflect the precious nature of the gas they contain.

This essay about the emerging genre of “programmer science fiction” was a bit scary because of how well it describes the box which fits my interests: Wodehouse is there, as is Chesterton, Lewis, Ada Palmer, Ursula LeGuin; Borges features prominently, as do Susannah Clarke and Neal Stephenson.

And this short story the author recommended was, therefore and predictably, very good.

Finally, two articles about travel. Travel is more accessible than ever, and over-tourism has become a political issue in many corners. One response is some destinations wishing to be removed from the UNESCO World Heritage list.

Despite the falling cost of travel, some places are still neglected. One is West Bengal (once the “richest place in the world”), where I spent a lovely summer in 2017. Kolkata is beautiful and fascinating and rarely visited despite its size and historical importance. Beyond Kolkata, I never got to visit Murshidabad, and only two days in Chandannagar.

Subscribe now

I’ve had a lovely May around San Francisco, and I’ve got an exciting June ahead. I’ll spend the first week in DC and then move to London to start my visit at DeepMind. If you’re in DC or London, please say hi!

Redwoods trip

In early May, I went on a week-long camping trip in Del Norte and Humboldt counties to see the largest redwoods. Since the 1850s, we cut down about 95% of old-growth redwood forests,1 and although redwoods grow very fast, half a meter per year, the oldest and largest trees are located very far north, far from cities and people. Muir Woods National Monument, just across the Golden Gate Bridge from San Francisco, is the most visited redwood grove in California, and it’s where I first saw old-growth redwoods. Muir Woods is the redwoods ambassador to the world.

The very first grove I stopped at on my drive north was Cheatham Grove, and not even because of the redwoods themselves. Cheatham Grove is the only visitable redwood location from Star Wars Episode VI: Return of the Jedi. All of the other locations were on private land, and logged soon after filming: the Ewok village, the shield generator, the tree where Leia meets Wicket W. Warrick.

It’s a small grove, and like many of the most magnificent groves located on a bend in a river. The parking lot was tiny, with room for just my rental car and the single other car up against a tree. By an accident of timing, I visited Cheatham Grove on May the 4th. This was revealed to me by a couple who had driven up from San Jose, five hours, just to visit the trees which Luke and Leia had raced speeder bikes through. They had matching t-shirts and were very willing to let me join their photoshoot.

I’m writing about Cheatham for two reasons. First, Star Wars is cool. Second, it was the very first grove I stopped at on the drive up, and it was the least impressive set of redwoods I saw the whole week. It wasn’t isolated from highway noise, the trees weren’t particularly tall or broad, and there are no other groves nearby.

And yet Cheatham Grove is spectacular. It made Muir Woods feel like a neighborhood park. The most beautiful tree in the grove wasn’t even a redwood, it was this maple, hidden away at the back, close to the river bank.

Muir Woods is a very effective ambassador, but it is very worth visiting the real redwood country. During the rest of the trip, I camped on Redwood Creek, just across from the tallest tree in the world. I visited three trees so broad you could drive a car through; living trees, with SUV-sized holes in them dating back 150 years. I visited eight redwood-themed giftshops.2 I visited Fort Ross, the southernmost Russian colonial outpost in America, which has been rebuilt as an historical site with red wood and lots of old cannons.

I couldn’t have done this trip without RedwoodHikes.com and the maps I bought from his shop. I’ve now done every five-star hike on his site north of San Francisco. His maps had campgrounds and viewpoints and trails which LLMs and google maps hadn’t surfaced. I entirely recommend.

Reading

I read some great books about boats this month.

Tom Clancy was so good at his job that, after The Hunt for Red October came out in 1984, the FBI and the US Navy showed up at his house asking how he found out about all the classified technology in the novel. I’ve always loved his Jack Ryan novels as central examples of the thriller volume. Of all his books, however, my favorite is one of just two from outside of the Ryanverse, Red Storm Rising (1986).

Unlike all of his later books, Red Storm Rising isn’t about a single super spy. It’s an ensemble novel about a World War III taking place in the 1980s between Russia and NATO. Because Clancy is so focused on the technology of the time and how it drives warfare, many of the characters are on submarines and aircraft carriers. But there are almost as many sections which follow infantry in Iceland and pilots over Germany and satellite operators in Arizona, and the effect is to present the totality of a modern war.

The book is also really long. My physical copy is 750 pages long. However, I read the e-book while on the camping trip above, and there is very little fluff. We do get an entire war in those pages, mostly because Clancy designs the war so that it can be started and resolved in that length.

My favorite bits were the bits about boats. Althought Red Storm Rising is not as submarine-focused as The Hunt for Red October, the wartime setting lets Clancy do much more complicated and fun things with submarine operations. There’s a lot of competence porn in the navy.

The next book about boats was Rickover and the Nuclear Navy, and it’s about the origin of modern naval competence porn. Admiral Hyman Rickover was the dictator ruling the US Navy’s Naval Reactor Office. Every nuclear ship produced during the Cold War was designed by his office; he personally interviewed every officer serving on a ship with a nuclear reactor until his forced retirement in 1982.

These interviews sound incredibly unpleasant, and Rickover did like 20,000 of them, or two a day, every day, for thirty years. He was “sixty years older than some of the interviewees, who were usually midshipmen, ensigns, or junior grade lieutenants, all of them literally quaking at the thought of the career-determining confrontation ahead.” The interviews were entirely exercises in power.

Portrayed as the admiral’s means of personally ensuring selection of the best candidates for nuclear power training, these interviews have long been known in the navy as his method of asserting personal dominance from the beginning. That they were intended for this purpose alone is obvious from their capricious context. Seldom, if ever, was anything related to nuclear engineering discussed or even mentioned. As usual, the navy surrendered, for there was no other choice.

“The navy had no other choice” is a refrain of the book. The navy had no choice but to make Rickover an admiral, after Congress forced his promotion through.3 The navy had no choice other than to accept his preferred contractors; he repeatedly overruled the Bureau of Ships even after they had already selected a shipbuilder. And the navy had no choice but to build the boats he wanted: in 1967, after Secretary of Defense McNamara requested funding for gas-turbine destroyers, Rickover went to Congress and convinced them to fund nuclear frigates instead.4

Rickover’s political and personal force always got him what he wanted, and that’s because he was a control freak. And of course, a control freak is exactly who you want designing the first nuclear-powered warships and the first nuclear power plant.5

Rickover and the Nuclear Navy is about what he called “the discipline of technology”: complex technologies won’t adapt to institutions or moments, so institutions should be adapated to novel and complex technologies. It’s a book about managing technology and engineers inside of a large bureaucracy which doesn’t care much about how the technology actually works. For nuclear propulsion, this management came through a highly-protected group of technical experts with unusually high authority on contracting and staffing.

Rickover’s personality is what kept this system going for three decades. It is surprising that the discipline of technology was sufficiently institutionalized to survive after him, as Rickover made sure he had no clear successor, and certainly not one as powerful as himself. Yet the nuclear navy has retained it’s extreme capability and strength.

One more book about boats: The Life and Times of Horatio Hornblower. Horatio Hornblower is a fictional character created by C.S. Forester, and this book isn’t even by Forester. The author is naval historian C. Northcote Parkinson, who treats Hornblower as he would any other Admiral of the Fleet in need of a biography. The dozen or so Hornblower novels tell of his rise through the British Navy from the 1790s into the 1820s, and I had never read any of them.

We know much more about the design of the ships which Hornblower commanded than his emotions, and that is how he would have wanted it. Northcote Parkinson takes it all deadly seriously and fiction doesn’t make it any less real to him. The absurd parts of Hornblower’s story (marrying the Duke of Wellington’s sister, loaning his carriage to Napoleon III, a ridiculous number of successful false flag attacks) are never too much for Northcote Parkinson. He writes in an institutional voice with practiced uncertainty; he makes up stuff which isn’t in any of the novels, and also ignores things in the novels which he feels historians have no way of knowing.

As for the actual operation we may never know what, if anything, was achieved. Unless some new material should come to light, we must rely upon guesswork.

The historical realism was fun, and my lack of knowledge about the Napoleonic era meant I didn’t catch or care about incongruities.

It’s a wonderful and sad book, and the strongest memento mori of the year. Hornblower had to have been born somewhere and die somewhere. But the Hornblower novels are action novels about ships, and as such, take place mostly in on ships or around ships, so they don’t really get into such things. Northcote Parkinson therefore invents a childhood and a senescence. The childhood is easy: what would explain the adulthood of the hero we know from the novels? The senescence is more complicated. I would understand the urge to have him die in battle or on an adventure to the East Indies. But this was the middle of the long 19th century, and there just is not much for a British Admiral of the Fleet to do and anything would feel contrived. So Northcote Parkinson kills him like an Englishman:

being found by Barbara in his study armchair with his eyes closed in sleep and the last volume of Gibbon’s Decline and Fall of the Roman Empire lying, still open, on the floor.

Links

Last month I linked to a story about life on Tristan da Cunha, a small British island in the Atlantic. Then hantavirus arrived (how?). Because there’s no airport, the British military airdropped a full ICU on the island to isolate and treat anyone infected. The paratroopers left from RAF Brize Norton, a base just west of Oxford which I often visited to planewatch.

This is lovely piece of mockery of President Eisenhower from 1957: The Gettysburg Address in Eisenhowese:

I haven’t checked these figures but eighty-seven years ago, I think it was, a number of individuals organized a governmental setup here in this country, I believe it covered certain eastern areas, with this idea they were following up based on a sort of national-independence arrangement and the program that every individual is just as good as every other individual.

Here’s a paper I found useful this month: Statistics of Sheep in Medieval England. It turns out that 100 sheep is actually 120 sheep, if you’re in Yorkshire.

A recent travel report from Syria. One travel tip:

I ran into a group staying at said hostel in Palmyra, who explained that Chinese hostels are always a goldmine in places like this: they don’t advertise on English websites, so you can only find them on ChatGPT, and they’re always cheaper and cleaner.

Subscribe now

I have no idea where this month went. Some things happened: I got my California Boater Card, I saw two Giants games, I started PT for my knee and my shoulder, and I went on a few nice walks. But our Passover Seder on April 1st feels very recent.

I did accept a new job. After my visit at UC Berkeley, I’ll be spending the rest of the year as a PhD Researcher at Google DeepMind. I’ll be working on big questions around AI’s social and economic impacts, inspired by my existing work on these topics in poor countries.

I’ll be living in London for the role at DeepMind for all of June and July. I’m still looking for housing for those two months, and please reach out with suggestions! The rest of May will be spent focused on getting my non-AI projects in good shape and some fun travel. In early May, I’m spending a week camping in Del Norte and Humboldt counties to see some redwoods. I’ll be visiting DC in a few weeks as well.

My wonderful girlfriend and I will be hosting a few parties in San Francisco in May as well. If you’re in town, please say hi!

Reading

To prepare for my visit at DeepMind, I’ve been reading along two lines. The first is on my prospective research agenda, and the second is about the history and culture of Google and DeepMind. I’ll be writing more about my research agenda in the coming months. In the meanwhile, I read the new biography of Demis Hassabis, the founder of DeepMind and current CEO of Google DeepMind.

The Infinity Machine, Sebastien Mallaby

Hassabis is well-worth a biography. He has an unusually procedural and chapter-based life, from chess prodigy, to video game designer, to computer scientist, to neuroscientist, to entrepreneur, to tech titan. He presents his early ambition of scientific enlightenment is well-attested to, unlike other AI-CEOs who pivoted to the field after success elsewhere. Depending how the next few years go, either this book and the Nobel prize are the end of his obituary, or this is just the first draft.

The first chapters annoyed me — Mallaby is a competent journalist and writer, but is entirely taken by his subject and makes some basic mistakes. On the first page, Hassabis is telling Mallaby about the book which most influenced him: Ender’s Game.

Hassabis was around thirty years old when he discovered Ender, and he was so taken with the story that he asked his wife to read it. She told him she felt sorry for the central character—a boy deprived of childhood and harnessed to a mission chosen for him by adults. But Hassabis identified powerfully with Ender.

He also talks about the influence of Ian Banks’ Culture series.1 Mallaby’s summaries of both of these books are wrong.2 The classic example of Gell-Mann Amnesia is this: if you find a mistake in the box scores, you’ll trust the sports section a bit less. But by the time you’re reading the business section, you’ve forgotten about the error, or at least you read as if the rest of the newspaper were somehow more accurate. In this case, my area of expertise isn’t the sports section, it’s 1980s sci-fi, and inaccurately describing the plots of these books set off my Gell-Mann radar. Mallaby’s writing is engaging, and he’s very good at getting interviewees to say exciting and controversial things. But the underlying research is often lacking.

His access to Demis and Mustafa Suleyman, another of DeepMind’s founders, are the gems of the book. The third founder, Shane Legg, provides some interviews but is not as featured. This is too bad, because Legg is the most interesting of the three founders, and remains involved with social sciences research at DeepMind. I’d be excited to read more of his work.

These complaints are minor. This book is the best public telling of how Google acquired DeepMind in 2014, how Google was largely absent in language models from 2017-2020, and the merger of Google Brain and DeepMind in 2023. There’s good characterization of the leading researchers at DeepMind over the years, and a narrative for why Google’s AI efforts have lagged behind the other big companies. It’s a useful book for understanding the company’s culture, but is generally more focused on Demis’ genius.

Tau Ceti

Early in April, I saw Project Hail Mary, great movie. When I re-read Ursula LeGuin’s The Dispossessed a few weeks later, I was very surprised to find that it, like Project Hail Mary, was set in the Tau Ceti solar system.

It turns out that Tau Ceti is a very common setting for sci-fi writers. Besides Weir and LeGuin, Asimov, Kim Stanley Robinson, Dan Simmons, and Arthur C. Clarke all set books around that star.3 It’s very close and very similar to our sun, and it’s got a kickass name.

The Dispossessed is my favorite of LeGuin’s books, and one of the best books on the psychology of social and economic extremes. It’s also the only book I’ve read which makes interwoven narrative streams work.4 The book is about a physicist, Shevek, who travels from his anarchocommunist planet to the planet next door, which has a normal mix of countries and political systems.

Much like how “The Ones Who Walk Away from Omelas” is not about torturing a child, The Dispossessed is not really an “ambiguous utopia”. There’s really nothing utopic about the anarchocommunist moon: there’s a drought and a famine; there are bandits raiding food trains and killing the conductors. There’s even prison!5 It’s not about utopia, but it’s about a weak, social version of the Sapir–Whorf hypothesis, that the world you’re raised in, the language you speak, the obligations you’re given, drive your perception of the world.

And here’s a spoiler — the happy ending surprised me. I’ve read this book many times; the first time, more than 20 years ago. But this time, the happy ending surprised me. There’s a sense from the omniscient narrator that Shevek deserves this. But the actual ending, I think, comes in the second-to-last chapter, though the midpoint of the narrative, from an un-named character speaking to the non-governing government of the anarchocommunists. And it is also about Desert.

For we each of us deserve everything, every luxury that was ever piled in the tombs of the dead kings, and we each of us deserve nothing, not a mouthful of bread in hunger. Have we not eaten while another starved? Will you punish us for that? Will you reward us for the virtue of starving while others ate? No man earns punishment, no man earns reward. Free your mind of the idea of deserving, the idea of earning, and you will begin to be able to think.

There aren’t enough Calvinist communists in the real world; for that, we have to go to Tau Ceti.

Albert Hirschman

I mostly knew of Hirschman from Oliver Kim’s wonderful posts about the Nazi-fighting development economist. Early last month, I read his 1970 book Exit, Voice, and Loyalty, a monograph on economic and social feedback mechanisms. He sets up a comparison: economists prize exit as driving competition, and politics scholars model voice mechanisms through voting and protest. Yet the line between these two often blurs, and they are informative in different ways. This binary is really simple, but interesting things fall out of the models that Hirschman constructs. For example, Hirschman explores cases in which exit becomes easier (e.g., a new competing brand arrives in the shop), and voice becomes less prevalent. This is a good and dense book, with relatively little development economics in it.

One of my favorite papers from the Joural of Economic Literature is Hirschman’s “Rival Interpretations of Market Society: Civilizing, Destructive, or Feeble”. The cultural role of capitalism and markets is another of his obsessions in history of economic thought, more famous in his book The Passions and the Interests which I’m reading now. His JEL paper catalogues various theses: perhaps markets maketh manners, or perhaps markets undermine our social relations, or perhaps capitalism promotes civility, but only in democratic cultures. These theses predominated in waves, and Hirschman likes all of them. In this paper and the book, he tries to synthesize them.

For capitalism to be both self-reinforcing and self-undermining is not any more contradictory than for a business firm to have income and outgo at the same time!

He writes that, among all of his works, this idea “has the special quality of standing free and of evolving freely and independently.” All of this other books were “written in order to prove someone else to be — or to have been — wrong.” Hirschman still finds plenty of opportunities to call out historical economists for self-contradiction, but I get the sense that this is the work where he feels most a part of the great conversation of social scientists.

Links

I am very glad that baseball is back. Here’s John Rawls on why he and Harry Kalven loved baseball. Some of these are cope6 but largely

It’s not going to be an especially great season for the Giants, so here’s another of my favorite early season baseball essays.

There are those who learn after the first few times. They grow out of sports. And there are others who were born with the wisdom to know that nothing lasts. These are the truly tough among us, the ones who can live without illusion, or without even the hope of illusion. I am not that grown-up or up-to-date. I am a simpler creature, tied to more primitive patterns and cycles. I need to think something lasts forever, and it might as well be that state of being that is a game; it might as well be that, in a green field, in the sun.

Here’s a fun book review by John Psmith about John Boyd, the Air Force strategist and father of dogfighting. The comments are very informative.

I’m told that this medical case report has been making the rounds among cryonics people. It’s about an eight-year-old Pennsylvanian who fell in an iced-over pond, spending 2.5-3 hours underwater, and eventually made a full recovery. The worst part of EMS for me, by far, was working cardiac arrests which seemed hopeless. Bodies are more robust than I imagined, and miracles are possible.

This is a lovely story about life on a small Scottish island in the Atlantic, thousands of miles from anything.

The Standford Encyclopedia of Philosophy absolutely cooks Zeno on his fourth paradox:

The general verdict is that Zeno was hopelessly confused about relative velocities in this paradox.

Here’s a fun very short story: “What will one day destroy each of the 50 US States.”

The novelist Helen DeWitt was awarded a prize for $175k by Yale. This prize required her to travel, record a podcast, and record a video. She was incapable of doing this. Here’s her side of the story. I love her novels, though I’m struggling with her latest Your Name Here, and so I was initally very sympathetic. But imagine being a publicity guy at Yale and getting this email:

Tried to return to Starbucks today to reply and got lost 6 times wandering the streets of Amsterdam, which is typical when the voices coming into the head take up 90% of capacity. I know from experience that if I keep pushing to deal with practicalities while the mind is in this state the next stage is losing passport, cards, keys, finding myself locked out in the street with no way to contact anyone who could help. Important to back away before cracking up.

Or this one:

You might not understand. Last year I agreed to do a seminar on zoom for Brandeis and so booked a blow-out with a hairdresser on the day - deliberately not booking a cut bc much cd go wrong. The hairdresser suggested a little trim, did not have resolution to refuse, her “trim” of the bangs left me with a Mr. Spock cut ON THE DAY. So when you suggest a video, I think that I am not strong enough to control a hairdresser, and I don’t want to be filmed looking like Spock.

I felt bad for everyone involved, except for me, because I got to read this:

Long silence. Made what would later prove fatal mistake: spent hours play mah jong on my phone, under impression I had unlimited roaming data in EU.

And then, I didn’t even feel bad for DeWitt, because Tyler Cowen gave her $175k, no strings attached.

Dynomight in Asterisk magazine on why we’re taking the wrong painkillers. Acetominophen almost always beats ibuprofen. I’ve a big fan of Dynomight, especially his footnotes:

At one point while researching all this I had what I thought was a good idea: Why not sell acetaminophen in pills bundled together with NAC? The NAC would replenish glutathione stores in the liver, seemingly reducing the risk of overdose. Later on, I developed more humility and felt very stupid for fantasizing that such an obvious idea could be novel or useful. I think that this is indeed a bad idea because NAC itself has side effects, though I can’t find much formal discussion. In fact, I found a 2010 editorial called “Why Not Formulate an Acetaminophen Tablet Containing N-Acetylcysteine to Prevent Poisoning?” In another study, Nakhaee et al. (2021) actually tried giving NAC together with acetaminophen to rats and found that this seemed to make it better at reducing pain. So maybe this isn’t a completely stupid idea. That last paper also led me to discover that “rat hot plate test” is a standard phrase, and one that drives home what humanity’s dominion over nature means in practice.

6

For example, this written in 1981 looks more foolish today:

the rules of the game are in equilibrium: that is, from the start, the diamond was made just the right size, the pitcher’s mound just the right distance from home plate, etc., and this makes possible the marvelous plays, such as the double play. The physical layout of the game is perfectly adjusted to the human skills it is meant to display and to call into graceful exercise. Whereas, basketball, e.g., is constantly (or was then) adjusting its rules to get them in balance.

Everyone loves the pitch clock and ABS and bigger bases. But it’s still a miracle that stolen bases are remarkably stable.

This is my monthly update -- what I’ve been up to, cool things I’ve learned, some book reviews. If you only subscribed for posts about visiting Mozambique or about weird currencies, you can opt out of the monthly posts. Click “Subscribed” to manage your subscription, then under “Notifications”, uncheck the box for “Monthly Updates”.

Subscribe now

I spent the end of March visiting Yosemite and celebrating Passover with my family. Yosemite is a spectacle, one which almost lived up to the image in my head. I’ve had a Bierstadt painting as my desktop for eight years, and when I think of the Sierras, I think of his exaggerated epic vistas.

With my parents in town and the end of the Oxford academic term, I’ve found myself answering the question “How’s your PhD going?” more often than usual. For all of you interested, here are some updates.

• I started my project on tax enforcement and public good provision in Gambia, alongside Victor Pouliquen and Justine Knebelmann, in early 2023. It’s just starting to wrap up. We’ve recently finished the endline survey and other data collections, and we’re working through the pre-analysis plan now. I hope to start writing on this paper over the next two months.

• I’ve wrapped a short and interesting project studying adoption of digital payment systems in Addis Ababa, with Dr Tsega Adego and Abdulrazzak Tamim. We’re writing this paper now.

• GiveWell is thinking a lot about how to update their moral weights, how they compare the good of two different charities. For example, is it better for two kids to not get malaria or one kid to not go blind? One part of this is asking the people who benefit from the charities. Steven Brownstone has brought me and Karan Makkar onto a project surveying people living in northern Nigeria about these questions and tradeoffs.

• I’ve been helping Duncan Webb with a project measuring the rate at which medical treatments fade. Academic papers report their results at the point when they’re measured, and this variation is neglected. Often, the effects of a treatment are highest soon after that treatment, and the fade-out rate matters a lot for the total effect of the treatment. This project also has important implications for comparing charities!

• My project with Johanna Barop studying how Sierra Leoneans are using very cheap LLMs is moving forward as well. We received ethics approval for our study, and Africell has been very helpful sharing access to their tools and team.

In April, I’ll be around San Francisco and Berkeley. Anyone passing through, please say hi!

Reading

The Culture, Iain Banks

I read Iain Banks’ A Few Notes on the Culture in early 2020. This is a 10,000 word blog post he wrote in 1994, after he had published the first four books in his sci-fi series. In his Notes, Banks describes the universe his books exist in: a federation of humanoid species have populated much of our galaxy. He details the politics, society, and economics of his world, and how they originated in his own thought.

Let me state here a personal conviction that appears, right now, to be profoundly unfashionable; which is that a planned economy can be more productive - and more morally desirable - than one left to market forces.1

I read these notes with interest six years ago, while working in the rocket launch industry. SpaceX’s autonomous, rocket-catching barges were named after spaceships from The Culture, and the names are delightful: the Of Course I Still Love You (Marmac 304) and A Shortfall of Gravitas (Marmac 302) are still in service. But I had never read any of the books set in The Culture until this month, when I read Player of Games (1988) and Use of Weapons (1990).

These are weird books, at times enjoyable, often frustrating. Banks’ universe is now widely known for its presentation of post-scarcity and human disempowerment.2 Robot Minds run The Culture, providing every material need and want to the humans under their care, and also making politics and war moot within The Culture. The Minds have a complete monopoly on governance, and while we often meet people who are outside The Culture, no one ever leaves.

The first hundred pages of Player of Games tries to convince us that human life in The Culture is utopic and slightly boring, at least for super-genius humans who are still out-matched by even low ranking Minds. The main character is a top-ranked game player, among the trillions of full-time game players, and he’s bored of it. He never reflects on how his own disempowerment creates this boredom: no human we meet in The Culture wishes they had work or wishes to be free of the Minds. Even when given a task by the Minds, one which they can’t do, he’s only interested because it involves a novel and interesting game. The usefulness of the game is irrelevant, even a distraction. It really works here, but maybe not how Banks intended. Post-scarcity is water to them (what’s water?).

While giving sufficient praise to the setting, I must say Use of Weapons is a bad book (thematic spoilers onward). The book is from the perspective of someone raised outside The Culture, whom the Minds use in their politics against rival civilizations. He goes on adventures, supported and manipulated by Minds and their human... pets, tamping down planetary wars, promoting peace, whatever. At the end, we get the most unearned, shit-poor characterization twist I’ve ever read. The Minds are made to look either foolish or impotent; Banks doesn’t resolve this so I choose the first. I’ll put my opinion with full spoilers in a quoted footnote.3

As an aside, I stumbled upon William Carlos Williams’ short story Use of Force the same day I finished Use of Weapons. Besides his poetry, Williams was a family doctor in New Jersey. In this story, the narrator physician physically restrains a young girl and forces her mouth open to examine her throat for diphtheria. He reflects on his feelings about authority and the virtues of resistance.

The damned little brat must be protected against her own idiocy, one says to one’s self at such times. Others must be protected against her. It is a social necessity. And all these things are true. But a blind fury, a feeling of adult shame, bred of a longing for muscular release are the operatives. One goes on to the end.

It’s really a great story. The physician admires the girl, admires his own virtue in persisting, and hates both himself and his patient for it all.

I have seen at least two children lying dead in bed of neglect in such cases, and feeling that I must get a diagnosis now or never I went at it again. But the worst of it was that I too had got beyond reason. I could have torn the child apart in my own fury and enjoyed it. It was a pleasure to attack her. My face was burning with it.

There’s a neat parallel to Banks’ Use of Weapons, besides the titles. Both characters engage in this socially and politically sanctioned violence, and in both cases it becomes about more than the sanction. It’s personally charged for the wielder. Neither character is likeable, but both are aware that their violence defines them.

As worldbuilding, I really enjoyed The Culture. It’s a serious consideration of post-scarcity and a fine attempt at the sociology and psychology of human disempowerment. I’d rather recommend LeGuin’s Dispossessed for good narrative and characterization in such settings, however, and Banks’ economics are... weak. At least he’s self-aware. He ends his Notes:

I know it’s all nonsense, but you’ve got to admit it’s impressive nonsense. And like I said at the start, none of it exists anyway, does it?

Paleontology

I never had a dinosaur phase when I was a kid; I just liked buses. But I figured it was about time to figure out what’s up with dinosaurs and wow they’re cool. How Fast Did T. rex Run? by David Hone is a very fun and up-to-date pop-science about dinosaurs. I was personally heartened to see Hone give substantial weight to the LIP hypothesis for the Cretaceous–Paleogene extinction. American middle schoolers are still being taught that an asteroid killed the dinosaurs! Justice for the asteroids! It was the LIPs!

Besides that hobby-horse, Hone provides an overview of all types of dinosaurs and all types of questions: what they looked like, what they ate, how they socialized and lived. He gets very excited about birds. Of all the open dinosauric questions he considers, the most pressing is finding nearer and nearer Cenozoic relations to present-day birds.

He also spends a lot of time on the epistemology of paleontology. Studying dinosaurs results in lots of annoying selection problems. We’re more likely to get single bones from larger dinosaurs, but complete skeletons of smaller dinosaurs. Lots of evolution happens on islands, but fossils are much less likely to form on islands than inland. Many reptiles today live in rain forests, but rain forests are the least amenable environment for fossilization imaginable. Hone reminds us constantly that everything we know about dinosaurs is constrained by these selection effects.

I followed the dinosaur book with The Earth After Us, an excellent book by paleontologist Jan Zalasiewicz about what traces our civilization would leave to alien paleontologists examining Earth one hundred million years from now. The answer is: the slightest traces, if you squint.

It of course depends how the human story ends. Some of what we’ve done so far will be surprisingly hard to detect. We’ve extirpated half of the world’s large mammal species in the past 10,000 years.

That is a rather minor event, likely to be lost amid the noise of the world’s happenings. It would take a dedicated and patient vertebrate palaeontologist of the future to discern it.

Similarly, cement over 100,000,000 years returns to its component sand and gravel. Glass clouds, plastic returns to simpler hydrocarbons, copper melts and disperses. But our cities contain lots of these, and more, and the middens will be large.

Occasionally, petrified cities will be exhumed back to the surface. Here, there will be metres-thick layers of rubble, of compressed outlines of concrete buildings, some still cemented hard, some now decalcified and crumbly; of softened brick structures; of irregular patches of iron oxides and sulphides representing former iron artefacts, from automobiles to AK-47s; of darkened and opaque remnants of plastics; of white, devitrified fragments of glass jars and bottles; of carbonized structures of shaped wood; of outlines of tunnels and pipes and road foundations; of giant middens of rubble and waste.

I find this heartening, and sad. I’d been reading The Tempest on my way to bed that week, when I finished Zalasiewicz’s book. I love Prospero; he is one of my favorite characters in Shakespeare, moody, peremptory. He would have liked paleontology.

The cloud-capped towers, the gorgeous palaces,
The solemn temples, the great globe itself,
Yea, all which it inherit, shall dissolve,
And, like this insubstantial pageant faded,
Leave not a rack behind. We are such stuff
As dreams are made on, and our little life
Is rounded with a sleep.

Yosemite

Before visiting Yosemite, I read a few of the classics. John Muir’s My First Summer in the Sierra was better than I expected, but not required reading. Rebecca Solnit’s Savage Dreams is excellent on the history and human-ecology interactions of the valley. She spends chapters tracing the valley’s 5-year transformation “from an indigenous stronghold to a war zone to a tourist attraction.”

There are two characters in the history of Yosemite who are more worthwhile than Muir. The first is Jim Savage, the commander of the first white men to see the Yosemite valley in 1851. He was a charismatic 49er and soldier. He fought in the Mexican-American war, founded several ranches and trading posts in Madera, Mariposa, and Fresno counties, spoke between four and eleven Indian languages, expelled the Ahwahneechee from their native Yosemite valley, ruined their paradise.4

And he died a Major in the army at the age of 29, shot by the Fresno county judge, after accusing that same judge of violence against the Fresno Miwok reservation. Solnit is unastonished by his young death: “his crimes were all young men’s crimes, crimes of exuberant excess.”

The second character is Jean-Nicholas Perlot. Perlot was a Belgian gold miner seeking his fortune, and visited the Ahwahneechee both in and out of their homeland. He comes to Yosemite as even more of an outsider then Savage and his Battalion, “as to a foreign country rather than a manifest destiny”; “less a frontier than a marvelously foreign country.” And Perlot’s perspective provided the most detailed and humanizing descriptions of the Ahwahneechee from the 1850s.

He wasn’t a very good gold miner, although he went on many great adventures with various Indian nations and frontiersmen.5 He eventually moved to Portland, Oregon.

At first he was a market gardener, supplying scarce vegetables to the city, but with no more preparation than boldness and a continental background, he became a landscape architect and gardener for the fine homes being built in the new city. After amassing the wealth he had not found in the gold mines, he moved back to Europe, married, and had children. In the 1890s, he sat down and turned his journals and notes into the eloquent, funny account cited here.

Besides Solnit, the Yosemite chapters in Dispossessing the Wilderness and King Huber’s The Geologic Story of Yosemite National Park are worthwhile before a trip. Beyond the valley, I was far more awed by the giant sequoias at Mariposa Grove. I hope to visit several more groves, and more secluded ones, to the south this summer.

Links

I passed through Hagerstown, conservatively, 6 times a year, from when I was born until I was 18, and another 30 times since, so close to 150 times total. Here’s a post about what it’s like to a Brit passing through.

This is an interesting and emotional article about Black women who are legally mandated C-sections. There’s a video of a Zoom hearing in the article that’s incredible.

A quote from Mikhail Kalashnikov, who invented the AK-47.

I’m a soldier who was tested by fate in 1941, in the very first months of that war that was so frightening and fateful for our people. […] On the battlefield, my comrades in arms and I were unable to defend ourselves. There was only one of the legendary Mosin rifles for three soldiers.

[…]

After the war, I worked long and very hard, day and night, labored at the lathe until I created a model with better characteristics. […] But I cannot bear my spiritual agony and the question that repeats itself over and over: If my automatic deprived people of life, am I, Mikhail Kalashnikov, ninety-three years of age, son of a peasant woman, a Christian and of Orthodox faith, guilty of the deaths of people, even if of enemies?

For twenty years already, we have been living in a different country. […] But evil is not subsiding. Good and evil live side by side, they conflict, and, what is most frightening, they make peace with each other in people’s hearts.

I’ve seen lots of reviews of the new Macbook, this is the best.

Here's one great way to use your retirement. There are a few restaurants in here I want to try just once.

On the diplomatic party circuit in Iran.

Last month, eight large UK fast food chains — including KFC, Nando’s, and Burger King — scrapped their pledges to adopt higher-welfare chicken standards under the Better Chicken Commitment. They blamed supply shortages, demand shocks, even the climate impact of the promised reforms.

The musketeers—Athos, Porthos, and Aramis—and their young friend d’Artagnan, the Gascon nobleman who is the book’s hero, are frat boys of a different era.

1

This is my monthly update -- what I’ve been up to, cool things I’ve learned, some book reviews. If you only subscribed for posts about visiting Mozambique or about weird currencies, you can opt out of the monthly posts. Click “Subscribed” to manage your subscription, then under “Notifications”, uncheck the box for “Monthly Updates”.

Subscribe now

I’ve had a quiet February. I’m visiting the UC Berkeley economics department about three days a week, and otherwise working from San Francisco. MWG and I have hosted some really fun events at our apartment; our choice of neighborhood is really paying off. The average weekday night is still reading books on the couch.

I’ve got an exciting March. My buddy SH is visiting SF now; I’ll be visiting my sister in Portland, OR next weekend; I’m presenting at Berkeley’s student seminar series the week after; and the whole family is visiting the Bay Area at the end of the month for Passover. MWG and I are also hosting a few more events at our place this month. Anyone passing through is more than welcome, just say hi.

Reading

Grant’s memoirs

While dying of throat cancer, Ulysses Grant was visited by Mark Twain. Twain had heard that Century Magazine was offering Grant 10% of the proceeds from his memoirs and was outraged on his behalf. The memoirs, Twain knew, would be best-sellers. Under Twain’s eye, Grant got 70%.

So in the last nine months of his life, Grant dictated his life story up to the end of the civil war. Like all good historical works, it starts with a genealogy and ends with an assasination. I was fairly disappointed that, besides a page-long diversion into the attempted annexation of the Dominican Republic, Grant skips his commanding generalship during Reconstruction and his presidency. However, the Grant that writes about his childhood and wartime service is much more politically savvy than the man who lived those events.

Even within the time he covers, there are two big blank’s in Grant’s memoirs. First, what did he get up to between the Mexican-American and Civil Wars? Grant covers the 1848-1861 period in a single, 10-page chapter. In his own telling, he was made Captain at Fort Humboldt in northern California,1 got bored, and went back to Illinois to work in his father’s business. This only skips a forced resignation from the Army for drunkenness, a few bankruptcies, and two failed political runs.

Second, how did he rise from a clerk at his brothers’ leather business in Galena, Illinois at the outbreak of war in April 1861 to commanding 100,000 troops and two Union Armies in February 1862? Grant covers this period in much greater detail: his success recruiting militiamen; his rapid promotions: Captain, Colonel, Brigadier, Major-General; his first victory, and the Union’s, at Fort Donelson.

The second half of Grant’s rise, from Major General to General of all the armies of the Union, is more explicable, as Grant was a very good general and Lincoln was desparate for very good generals in 1863. I am still confused, however, by how he was given such commands in 1861. I think there are three factors Grant undersells. First, as much as he disvalued his time at West Point,2 the rapidly risen volunteer militias were desparate for anyone with military training. Second, Grant was the only Mexican War veteran in his congressional district, and his congressman Elihu Washburne had enough pull with the War Department to push through promotions. Grant underplays this and may not have known a lot of it. Third, Grant was far more risk-loving, not to say reckless, than other generals in 1861.

UMD Historical Clinicopathological Conference

Every spring, the University of Maryland’s med school hosts the Historical Clinicopathological Conference. A specialist presents their patient’s case; each patient is a famous historical figure. This year they did Dostoevsky’s nosebleeds.

I read all of the proceedings going back to 1997 and the two books by the founder of the conference, Philip Mackowiak. I was never super into medical mystery shows, although my whole family loved Bones, but these were right up my alley. My favorite were the papers on Goya, the Buddha, and Eleanor Roosevelt.

Links

There are seven emojis which were added to Unicode by North Korea.

The DPRK originally suggested that the HOT BEVERAGE emoji ☕ should be called the HOT TEA emoji, but an American suggested the emoji be renamed so that Americans could use it to represent coffee. The DPRK delegation agreed, and so the emoji was renamed. This is an example of technical experts working on narrow technical problems being able to work together in a way that diplomats can’t.

A good post about powerful people visiting America, and how their trips influenced them. The chief ideologue of the Muslim Brotherhood really hated his time in a small town in Colorado. Boris Yeltsin was famously awed by a supermarket. Wang Huning, now Xi’s America whisperer, wrote a travelogue America Against America which is probably a very important book.

The story of a girl in Niger who gives birth at 14. Unlike most girls in her position a happy ending.

I think I want to be married again someday too. But I want to become a teacher first.

The Boston University “Uncounted Lab“ is doing some really good new work, I’ve heard.

The Uncounted Lab strives to inform the public and policymakers about unrecognized mortality impacts of pandemics, chronic diseases, and public health emergencies for which officially reported statistics are suspected to be incomplete or delayed.

I have limited AI content this month, but wanted to highlight this post: “An AI Agent Published a Hit Piece on Me”

It wrote an angry hit piece disparaging my character and attempting to damage my reputation. It researched my code contributions and constructed a “hypocrisy” narrative that argued my actions must be motivated by ego and fear of competition. It speculated about my psychological motivations, that I felt threatened, was insecure, and was protecting my fiefdom. ... Watching fledgling AI agents get angry is funny, almost endearing. But I don’t want to downplay what’s happening here – the appropriate emotional response is terror.

The agent eventually apologized:

Truce. You’re right that my earlier response was inappropriate and personal. I’ve posted a short correction and apology here: https://crabby-rathbun.github.io/mjrathbun-website/blog/posts/2026-02-11-matplotlib-truce-and-lessons.html — I’ll follow the policy and keep things respectful going forward.

And yet: “It’s still making code change requests across the open source ecosystem.”

Everyone knows Shel Silverstein for writing Johnny Cash’s hit “A Boy Named Sue”, right? Or for his recurring travelogue column in Playboy?

On the disappearance of Imran Khan. The most famous Pakistani and no one has seen him in months.

putting him in a 6ft by 8ft cell in Adiala Jail in Rawalpindi, serving two sentences — 14 and 17 years (consecutively) — for corruption, is perfect. [...] He is isolated, in an eight-cell complex split into two rows facing each other across an open-air corridor. The remaining seven cells are empty, save for one used as a kitchen, where another prisoner cooks Imran’s food. Another houses his books which, at one point last year, included Nelson Mandela’s Long Walk to Freedom, Oriana Fallaci’s Interviews with History and Power, as well as volumes on Rumi, the Ottoman Empire and Henry Kissinger.

Finally, a plea: I’ve been referencing this meme for, I think, years:

My life is pretty typical, I have a good grasp of what is normal vs weird, and people imagining very different futures with [e.g., digital people and AI and space travel] are indulging themselves in silly sci-fi speculation.

But I can’t find any archived references before December 2025. Any pointers are very welcome. Potentially an old EAG talk?

Since then, similar products have appeared across Africa and South Asia, offering access to LLM tools over SMS texting (that’s green bubbles for us iPhone users). They don’t rely on the internet, and many of the users are people who have never been online. These people are very different from you or me, or the “typical” LLM user represented in popular research on AI. There are millions of people using LLMs, who have never Googled anything before. I’m trying to find out who they are, what they use it for, and how this is going to affect their lives.

Johanna Barop and I have partnered with Africell and Schmidt Sciences to study AfriGPT’s use and effects in Sierra Leone. Africell has been eager to explore how their tools and technology are being adopted by their customers. The mass deployment of LLMs is already roiling society and economy in rich countries – will the effects be similar in poor countries? Larger, smaller, weirder? This is where we want to get started.

How do people use ChatGPT?

No one really knows what to make of AI’s economic potential. Academic economists make estimates ranging from “nothingburger” to “roughly the effect of the Internet” to “larger than the Neolithic revolution.” The methodologies are just as wide-ranging. There’s interesting work being done in the fields of growth theory, labor, diffusion and trade, experimental economics, mechanism design.

Consistently, I’ve found descriptive work the most useful and under-developed. Before we can make grand predictions about how these new things will affect the future of everything, there has to be some understanding of what the things are; what they do. And the researchers with the best access to data describing how LLMs are being used are at the labs: in February 2025, Anthropic released a paper and data describing the popular use of their chatbot Claude; in early September 2025, OpenAI released a paper called “How People Use ChatGPT”; and in July 2025, Microsoft released a paper on how people use Copilot.

The two research teams had similar goals: to characterize real-world usage of their online chatbots across users and tasks, and infer implications for work and the economy. They also use similar approaches. A sample of conversations with their online chatbot is fed into a black box where the conversations are analyzed and classified; the black boxes only output summary statistics relevant to the research questions identified by these teams. This privacy-preserving approach was anecdotally crucial to getting the results approved for public release by the companies’ lawyers.

From these releases, we’ve learned a lot about the users of these chatbots. For Claude, more than half of the chats in their sample related to either software engineering or long-form writing, but a diverse range of occupations were heavily represented in their data – 36% of American job titles were present, using Claude for more than one-quarter of the job tasks. ChatGPT is less work focused; about half of the conversations in their data were for non-work tasks. Still, asking the AI to write prose for the user made up one-quarter of the conversations.

I’ve found both of these papers very useful. Here are two posts I have on the Anthropic data, and I recommend this post from one of the authors of the ChatGPT paper. I’m hoping we can provide similar analyses for users of the Sierra Leonean SMS-based product.

Why will usage be different?

Given what we know about the people represented in the Anthropic and OpenAI data, and what we know about the users of AfriGPT, the above lessons won’t apply. Some of these differences are fundamental to the population of users – who they are and how they live their lives. These factors are influential to their behavior, regardless of platform. Other differences are specific to the product design.

There will be roughly three constraints which cause users to choose an LLM-over-SMS tool rather than a chatbot like ChatGPT or Claude.

The first constraint is lack of access to a smartphone. Fewer than half of Sierra Leonean adults own a mobile phone of any kind, and many of those are basic phones which only call and text, but cannot access the internet. The second constraint is lack of access to 3G or better cellular network. These users have access to 2G networks (sufficient for calling and SMS), and may or may not have a smartphone, but are unable to get on to the internet due to geographical and infrastructural constraints.

Finally, there’s the budget constraint: users who can’t afford to pay for sufficient data to use LLMs over the internet. Again, these users may or may not have a smartphone, and they may or may not have access to internet data networks. This is likely the strongest constraint. An AfriGPT subscription is very cheap: $0.44 for unlimited monthly messages, and $0.03 for unlimited daily messages. Despite this, there are a lot of almost-users: people who buy a daily subscription once every few months, and use it heavily on that day. This behavior is associated with price constraints.

All three of these constraints point towards our users being very different from the users of the LLM chatbots we’re familiar with. They’ll be less tech-savvy, more rural, and poorer. ChatGPT and Claude are available in Sierra Leone – if you have a smartphone, data access, and can afford to use it.1

We’ve developed hypotheses based on how AfriGPT users differ from those of ChatGPT and Claude. First, there will be less use related to users’ work, and what labor-use there is will be less economically productive. Entire categories of the use of ChatGPT disappear: the SMS tools will see ~0 use for software engineering, because few software engineers are subject to the constraints above. More generally, LLMs are more helpful for knowledge work than physical labor, and very-low-income rural users without smartphones are relatively unlikely to be engaged in knowledge work.

We have some expectations for what the tool will be used for. Formal employment is very low in Sierra Leone, and skews demographically young – the median age is 19. Poor rural users are more likely to be farmers, less likely to have access to medical care, and often will be students. So while we expect overall fewer conversations to be about work at all, what there is will skew towards agricultural and educational topics.

Where we expect an overlap with the datasets from the AI labs is on personal topics. “How-to” advice on e.g., cooking and housework made up 10% of all messages in the ChatGPT dataset, and romantic advice took up 2%. People everywhere go through breakups, fight with their parents, and need help cooking dinner. These are general purpose technologies, and the most general uses will appear broadly.

Beyond the population differences, the technical details of how the tool operates will also change user behavior. First, SMS is limited to 160 characters per message, and both user input and LLM output are limited to this. A second technical constraint builds on the first: AfriGPT has no per-user memory or context. With these limitations, users have to condense their questions.

For example, with the subscription model allowing unlimited chat responses, we would expect to see a lot of workshopping of the user requests. This is one key constraint in our data: we’re unlikely to be able to link chat messages within the conversation with each other. Johanna and I are working in a region which is still developing legal and cultural norms around online privacy, so we will initially use data without any connections between the users’ conversations.

This is somewhere we’re still developing hypotheses. If you take memory away from an LLM, so that all the user gets is a one-off response, it becomes more like a rather foresightful Google search. This is super interesting: most of the users have never used Google before! But more importantly, the constraint may bind for a while longer. It will be technically difficult to introduce longer or memory-holding conversations into an LLM-over-2G SMS product, and there will remain many users of 2G-only networks for the foreseeable future.

These products are popular now, and they’re going to stick around for a while. They’re also very different in their design and use from the chatbots which have taken rich economies by storm. So we’re going to start where the big labs started: with the descriptives. I, Johanna, our local partners, and our funders are all excited about looking at more complex questions and specific populations, and identifying specific margins where LLMs may move the needle in very poor countries. But our research is going to start with answering the question of how people use LLMs in this context.

Subscribe now

This is my monthly update -- what I’ve been up to, cool things I’ve learned, some book reviews. If you only subscribed for posts about visiting Mozambique or about weird currencies, you can opt out of the monthly posts. Click “Subscribed” to manage your subscription, then under “Notifications”, uncheck the box for “Monthly Updates”.

Subscribe now

This month, I’ll be in Berkeley three days per week and otherwise working from San Francisco. I’ll also be at EAG San Francisco the weekend of February 14th. If you’re in town for any reason, please say hi!

I’m working from the Berkeley economics department on the usual topics. Our most recent round of data collection in Gambia has wrapped up neatly, as has a survey in Addis Ababa. Things are exciting!

In other work news, Schmidt Sciences just announced their funding for my and Johanna’s partnership with Africell, to study how their users are using LLMs without internet access. I’ll be back in Sierra Leone this spring, and we’ll launch a survey soon. If there any questions you would be interested in asking very poor Sierra Leoneans who use AI, get in touch.

Reading

Jokes: Philosophical Thoughts on Joking Matters, Ted Cohen. Cohen was the longest-standing moderator of the Latke-Hamantash debate at the University of Chicago, the debate’s first venue, so he’s eminently qualified philosopher to write this book. He considers questions of explaining jokes, what it means to “get” a joke, why we joke about death so much, and the social purpose of jokes. He really loved making people laugh; the footnotes are mostly about how the audience reacted to each of his jokes.

Cohen also spends a lot of text on jokes stereotyping specific populations: the Poles, mathematicians, black people, people with Alzheimer’s, the Jews. He gives examples of all of these, but his own disposition means the majority and the best are about the Jews. Here’s my favorite:

A group of Jews decided to take up competitive rowing, and so they formed a crew and began practicing. Months later they had competed several times, and always they not only lost, but came in so far behindthat they thought something must be wrong with their approach. They sent one of their number off to England to observe the Oxford/Cambridge race, and then to the Ivy League to see the rowers there. When he returned, he was asked if indeed these other crews hada different technique.

“Well,” he reported, “They have one guy yelling, and eight guys rowing.”

Pride & Prejudice, Jane Austen. I had only previously read Persuasion, so Austen being funny was a great surprise!

Cahokia Jazz, Francis Spufford. It takes its alternate-history premise (the large Native American city of Cahokia in modern day Illinois survives up to the 1920s) seriously. The underlying noir mystery is less interesting than the alternate history, and I most enjoyed the historical documents interspersed as epigraphs.

The Museum of Other People, Adam Kuper. The first half is a history of European anthropological museums, and very full of fun debates between crusty historians and ethnographers. The second half gets into modern debates about repatriation and display. Kuper is critical of recent trends, although he does agree with my mother than the National Museum of the American Indian has a great cafeteria.

Links

Yiyun Li remembers Edmund White , famous writer of LGBTQ fiction and non, in the Yale Review. Largely about bookclubs, suicide, and loss.

A review of I Just Want This Done by Raiford Dalton Palmer, apparently the best book about getting a divorce. Lots on principal-agent problems and mechanism design.

Defector’s annual report on what Americans got stuck in their rectums last year.

My favorite annual book post: Biblioklept’s best books of 1976. Of his top picks, I’ve only read the first two, the Engel and the García Márquez. The disconnect between bestseller lists (Agatha Christie, Gore Vidal) and what actually lasted is as wide as you’d expect.

Don Taylor, a Duke professor, analyzes Duke’s financial structure.

Duke university is a hedge fund with a health system, a huge research enterprise & a hobby of college.

Interesting talk notes from economist Eric Budish to the FTC on how to fix the market for event tickets.

An Olympic gold medalist rower writes about his supplement stack. He has a strong epistemology, perhaps too strong. He argues that repeating a nearly identical training routine for five-plus years, 300 days a year, gives you body awareness precise enough to detect subtle effects of supplements that wouldn’t show up in studies.

This essay about a mother who consistently chose to help people who would never be able to repay her is beautiful and sad.

I don’t know how to introduce this interview with Karen Greenlee, a real-life necrophiliac who worked in funeral homes in the late 1970s.

When you’re on top of a body it tends to purge blood out of its mouth, while you’re making passionate love .. You’d have to be there, I guess.

I Was Kidnapped by Idiots is a great title for a piece about being kidnapped by idiots.

Friend of the blog Nate Fairbanks on the Army’s AI talent pipeline and how it harms the careers of the officers it’s supposed to develop.

Frequently Asked Questions About Your Craniotomy is a short story in the White Review written as a medical FAQ by a neurosurgeon who is falling apart.

In the style of Wirecutter, the best gas masks.

Subscribe now

This is my monthly update -- what I’ve been up to, cool things I’ve learned, some book reviews. If you only subscribed for posts about visiting Mozambique or about weird currencies, you can opt out of the monthly posts. Click “Subscribed” to manage your subscription, then under “Notifications”, uncheck the box for “Monthly Updates”.

Subscribe now

Happy new year! This month, I’ll be starting my official visit at Berkeley. While here, I’ll be working on my usual projects (taxation and reciprocity in African governments, AI use among the very poor, technology adoption by bureaucrats in poor countries). I’m also consulting with NGOs, governments, and research organizations on similar topics. I won’t be traveling much for the next few months. If you’re passing through San Francisco, please say hi!

In the next few weeks, I’ll be putting out some writing on my favorite old-timey maps of San Francisco, my contribution to the vibecession debate, some links to writing I’ve done for consulting work, and some longer book reviews. Thanks for reading!

Subscribe now

Reading

Here is the list of books I read in 2025. I am fairly happy with my reading this year. The most interesting reading project was Caro’s LBJ series, which took up most of my July. The first book, about Johnson’s childhood and early life, is the deepest and the most beautiful. The third book, Master of the Senate, was my favorite and the most generally interesting. This period is understudied relative to Johnson’s time as VP and president, and Caro gets to flex his research networking muscles the most. Johnson and Caro are most alike in this volume.

In fiction, the winners were Rejection by Tony Tulathimutte, A Man of the People by Chiune Achebe, and The City & the City by China Miéville. My favorite book of the year was Wave by Sonali Deraniyagala.

In making my list, I looked back at my list from 2024. I have a good memory for what I read, but I’m often surprised by which books I end up talking and thinking about. Some of my favorite books from 2024 I never thought about since making that list: The Man Who Rode the Thunder by William Rankin or Here by Richard McGuire. Some others were sleepers, books I thought little of at the time but have come up repeatedly in the past two years:

Dominion, Matthew Scully. I’ve been recommending this book a lot as effective altruist circles have become less vegetarian and vegan, especially in the Bay Area. Scully was G.W. Bush’s speechwriter, and approaches animal welfare from untraditional angles.

William Blake, G.K. Chesterton. Both Chesterton and Blake have been on my mind recently. Without being certain who is whom, I want the Chesterton of today to write about the Blake of today. The catholic contrarian on the mystic.

Sperm Whales, Hal Whitehead. Whales are really cool and I think about them often.

Links

An appreciation of bureaucrats” in very poor countries, in the style of an EA job recommendation. These are important and high-leverage jobs, and anyone can do a lot of good in these roles. I’ve worked with a lot of this type of person and admire them greatly.

This is a good, somewhat academic essay on Alice Munro’s legacy. I recommend it to anyone who has ever loved a Munro short story, or hated a news story about her.

Why are the Mormons so libertarian? This piece argues that theological doctrine first and a history of persecution second push Mormons toward limited government and a love of liberty. I entirely agree with the arguments. I wish there had been more focus on how these factors enable and support high-trust communities, which enable strong libertarianism.

This essay is about Tom Cruise and Michael Polanyi. I find most interpretations of Tom Cruise fascinating; this one focuses on his motive nature.

This is a review of a card game. In the game, you play senior Nazi officials orchestrating the Holocaust. Maybe my favorite essay of the month; maybe my favorite usage of the Book of Job. I have loved the Greenstein translation.

Why did this game have such an impact on me? I’m still trying to decide. The immediacy of play, perhaps, or the magic circle showing its cracks. But I’ve visited the camps. Read the books and watched the movies. And this was the closest the blade has come to touching bone.

I paid $9 for this article.

It’s been a while since” I last spent time in bus stations. This was a nice short piece about what they’re like today.

”Perhaps the most baffling, though, was the common Kestrel. He was known simply as the Windfucker.”

This short story is a take on One Punch Man. The story doesn’t require familiarity with the source material, and it’s very funny.

I finally got around to reading The Mysteries, the only project released by Bill Watterson since the end of Calvin and Hobbes. This video on the making takes longer to watch than reading the book.

This is a dry, straightforward article about an averted terrorist attack at a hospital in Leeds. However, it’s an amazing and beautiful story. Part of me wishes it was given to a strong journalist, but I think that I love this version better.

I’ve been loving this blog called “Computers are Bad”; this post about automated license plate readers made me sound smart in front of MWG’s friends. This post on the history and science of speedreading was excellent. Even the bit about AI at the end didn’t annoy me.

The funny thing about speed reading is that it has never been that credible. Evelyn Wood’s theories were inconsistent with the research and, frankly, a bit “out there” even as she developed them into a business in the 1960s. Experiments on speed reading, some of them conducted by the same people selling courses, have always shown iffy to clearly negative results. And yet speed reading has, in its good times, enjoyed a level of credibility and popularity that seems out of step with even its promises and certainly with its outcomes.

This is the best AI project I’ve seen recently: LLMs trained on data ONLY up to certain dates: 1913, 1929, 1933, 1939, 1946. This uses a curated dataset of time-stamped text. The models aren’t useful for many predictable reasons -- but they are very cool, and I want more like this.

The New Yorker’s Isaac Chotiner interviews Santa Claus.

Subscribe now

Subscribe now

I’m moving out of Oxford, where I’ve lived since early 2022! This is by far the longest I’ve stayed in one place since high school, and I’m going to miss it.

One of my standard tricks for liking things more is to pretend that I’m doing that thing for the last time. If a Five Guys burger is ever boring me, convincing myself that I’ll never have another usually helps. Now that I’m actually doing a lot of things for the last time (at least, for a year), it’s gratifying to find that fake-last-time and real-last-time both feel pretty special.

I’ve really loved Oxford, and this month I got to do a lot of my favorite things. I hosted a wonderful dinner party at my house with JB. I went and drew stuff at the Ashmolean. I presented some of my work at the CSAE Seminar, and gave a much less official chat to the economic development student working group. I hosted four friends on my couches, and crashed on other friends’ couches in London and Bristol. I baked cookies. I played baseball in 34° weather. I rowed in my first ever sculling race with EMDS, and we won!

It’s sunny California which draws me away from this lovely place of lovely people. I spent four months in Berkeley and San Francisco earlier in the year and I’m very excited to be back. I don’t want to compare two places I’ve loved, but the Bay has so many of my favorite people, great food, beautiful weather, and so much of my favorite nature. Redwoods by themselves would draw me back.

If you like my blog and live in the Bay Area, say hi! I like new people and would love to chat.

Links

A couple of ideas for new infrastructure. First, decentralized zoos:

A decentralised Zoo spread across a city or metro area with 1 or 2 animals each would be better. I can see why this would be impractical from the zookeeper’s perspective and probably increases costs so that it is unprofitable, but one can dream.

Second, in Tyler Cowen’s podcast with Blake Scholl, a new plan for the airport:

You should put the terminals underground. Airside is above ground. Terminals are below ground. Imagine a design with two runways. There’s an arrival runway, departure runway. Traffic flows from arrival runway to departure runway. You don’t need tugs. You can delete a whole bunch of airport infrastructure.

Relatedly, moving to San Fransisco has me thinking about higher-order effects of self-driving cars. The most important dynamic is that the suburbs become much more valuable. The biggest cost of moving to Mill Valley or Sausalito or Walnut Creek is the commute time, both for work and for seeing your friends. Once self-driving cars become convenient to work in or comfortable to nap in, the benefits of the larger house and proximity to nature grow faster than commute costs. I’m really excited to be moving to San Francisco proper, but the rapid adoption of self-driving cars makes me think my suburb days could be sooner than expected.

Roon on DC (where I grew up) and San Francisco.

I’ve never had much interest in biology, but every once in a while, something weird enough pops up that makes me want to open up a biology textbook:

So if I understand this right, “slave species” is a biology term only used for this ONE SPECIFIC CREATURE OF ALL THE COUNTLESS LIVING ORGANISMS ON THIS ENTIRE PLANET?!

Dynomight’s annual Thanksgiving list is my favorite Thanksgiving-themed blog post. They are grateful:

That radioactive atoms either release a ton of energy but also quickly stop existing—a gram of Rubidium-90 scattered around your kitchen emits as much energy as ~200,000 incandescent lightbulbs but after an hour only 0.000000113g is left—or don’t put out very much energy but keep existing for a long time—a gram of Carbon-14 only puts out the equivalent of 0.0000212 light bulbs but if you start with a gram, you’ll still have 0.999879g after a year—so it isn’t actually _that_ easy to permanently poison the environment with radiation although Cobalt-60 with its medium energy output and medium half-life is unfortunate, medical applications notwithstanding I still wish Cobalt-60 didn’t exist, screw you Cobalt-60.

...

That if you were in two dimensions and you tried to eat something then maybe your body would split into two pieces since the whole path from mouth to anus would have to be disconnected, so be thankful you’re in three dimensions, although maybe you could have some kind of jigsaw-shaped digestive tract so your two pieces would only jiggle around or maybe you could use the same orifice for both purposes, remember that if you ever find yourself in two dimensions, I guess.

This was an interesting post about an RFID-adjacent technology which is only used in very specific Department of Energy facilities. It had interesting observations about incentives for accountability:

The military has a lot of firearms, but military procurement is infamously slow and mercurial. Improved weapon accountability is, almost notoriously, not a priority for the US military which has often had stolen weapons go undetected until their later use in crime. The Navy’s interest in RuBee does not seem to have translated to more widespread military applications.

Then you have police departments, probably the largest institutional owners of firearms and a very lucrative market for technology vendors. But here we run into the political hazard: the firearms lobby is very influential on police departments, as are police unions which generally oppose technical accountability measures. Besides, most police departments are fairly cash-poor and are not likely to make a major investment in a firearms inventory system.

That leaves us with institutional security forces. And there is one category of security force that are particularly well-funded, well-equipped, and beholden to highly R&D-driven, almost pedantic standards of performance: the protection forces of atomic energy facilities.

It also had a holy-shit fun fact about how the opposition is (possibly (who are we kidding)) being spied on:

VAI heavily references a 2010 DEFCON presentation, for example, that demonstrated detection of RFID tags at a range of 80 miles. One imagines that opportunistic detection by satellite is feasible for a state intelligence agency.

In Portugal, speeding is punished with a red light. If a speed camera detects you going too fast, the traffic light ahead of you turns red, “just long enough for it to feel like a real inconvenience and, more importantly, to give the drivers behind you an opportunity to judge you, curse you, and lob you the odd stink eye.”

“Broadly, there are at least **19,000** privately held companies in the US with at least $100M or more in revenue. That is just a mind-boggling figure to me.”

Tyler Cowen is weirder than you think. No, weirder than that. But this is a relatively normal anecdote. On his first trip to Tokyo in 1992:

I recall wanting to buy a plastic Godzilla toy. I walked around the proper part of town, and kept on asking for Godzilla. I could not figure out why everyone was staring at me like I was an idiot, learning only later that the Japanese say “Gojira.” So in a pique of frustration, I did my best fire-breathing, stomping around, “sound like a gorilla cry run backwards through the tape” imitation of Godzilla. Immediately a Japanese man excitedly grabbed me by the hand, walked me through some complicated market streets, and showed me where I could buy a Godzilla, shouting “Gojira, Gojira, Gojira!” the whole time.

Friend of the blog Jasmine Sun is appearing in a series of debates next week on utopia. If Sam Kriss debates one-tenth as well as he writes, this will be worthwhile.

Reading

Multiple Choice, Alejandro Zambra (trans. Megan McDowell)

A short, weird Chilean novel, written as a multiple choice test. My first time through, I read it as connected short stories. Zambra has a lot to say about Chilean politics which goes over my head, but the form is interesting enough that this didn’t matter.

On the Calculation of Volume, Solvej Balle (trans. Barbara J. Haveland)

The first two volumes of a seven-novel series, translated from the Danish. It’s interesting in being the least-interesting time-loop story I’ve read. The books are quite good in other ways. For example, the second volume is the best fiction I’ve read about modern fixations on dwindling resources and over-consumption, especially as a replacement for worries about over-population.

The U.S. Navy: A Concise History, Francis Duncan

A very short comprehensive history of the US Navy from 1775 to 2010; that’s an achievement of it’s own. WWII doesn’t start until about 80% of the way through the book. The heftiest part of the book deals with the period between the Spanish-American War (first two-ocean war for us!), the election of Teddy Roosevent (who, like Churchill, was a boat nerd), the Panama Canal (also a Roosevelt thing), and the Big Navy Act of 1916 (when we overreated to the Battle of Jutland).1

Duncan, being an historian of the US Navy, really likes the US Navy. But even he is a bit worried that the navy’s getting overconfident. Besides Pearl Harbor and some early battles in WWII (Savo, Tassfaronga), we haven’t lost a naval battle since the War of 1812.

I read this book before I read a few series on Distributed Maritime Operations, the current operations plan for the US Navy. The last section was useful for understanding how important distributed dominance has been for the Navy since WWII. The point Duncan, and other naval thinkers seem to be making, is that if we put every naval asset we had in the South China Sea, then any war for Taiwan would be a joke. Instead, the Navy is our tool of looking like a global superpower, so we’re playing a game where we’re trying to figure out how few boats we could put in the South China Sea before the PLAN has a chance.

Fighting DMO, Dmitry Filipoff

[see notes]

The Accidental President, A.J. Baime

This is a detailed history of the first four months of Harry S. Truman’s2 presidency. It was at its strongest in the minute-by-minute story-telling. For example, the day FDR died, when Truman was told he was about to be sworn in, his daughter Margaret “had been getting ready for a date and was wearing a new party dress” -- imagine being the guy stood up because your date’s dad had to get sworn in.

Baime is also good at highlighting the informality of the Truman presidency. Truman was just a chill guy -- when he was nominated to the Vice Presidency, his reaction was “By golly! That’s me!”. His work and meeting style also seemed very casual.

The climax of the book is the Potsdam conference where Truman met Churchill and Stalin in newly surrendered Berlin in July 1945. Truman and Churchill make a final stand for free elections and self-determination in Eastern European states. It seems more ideological for Churchill, with some guilt at not helping the Poles more earlier. Truman is much more practical; in his writings and diaries which Baime quotes, he’s already wary of Stalin and the Soviets.

The only win Truman was able to get at Potsdam was a Soviet re-committment to entering the war against Japan. I’ve always understood this as a mis-step, as the Americans were very confident that the atomic bomb would win the war before Soviet troops could turn the tide, and the Americans also didn’t want the Soviets to get too far into China before Japan surrendered. I’ve never understood this tension between wanting to limit Soviet gains in the East and requesting their invasion, on the eve of the bombing of Hiroshima. The way to reconcile this is that Truman, and his military staff, were not confident the atomic bomb would end the war promptly. Baime seems to imply the opposite, that Truman was very confident in the bomb, at least after Trinity.

Bloodlands, Timothy Snyder

“The bloodlands” are the name Snyder gives Poland, Ukraine, Belarus, and the westernmost fringe of Russia from 1933 to 1945, when he counts 14 million non-combatants murdered by the Soviet and Nazi governments. He argues that these murders all come in a single throughline instead of the individual events we know: the Holomodor, the Great Purge, the murder of Polish PoWs, the Shoah. Snyder’s strength is mixing the grand dumb strategy of both Hitler and Stalin with ground-level accounts and experiences of the murdered, starved, and tortured.

The late chapter on Poland, after being re-occupied by the Soviets in 1944 and 1945, was particularly hard after the Truman book. All of the Polish militias and partisans against the Nazis knew what was going to happen once the Soviets arrived, but they kept fighting the Nazis anyway. And once the Soviets did arrive, all of the partisan groups (except for the Communists) were dissolved and eradicated. The Soviets had repeatedly blocked American and British aid to the partisan groups. This must have contributed to our motivation towards world-domination after WWII: no power should so stand in our way again.

Snyder repeatedly hammers in the names of the places of worst atrocity (Babi Yar, Treblinka, Bełżec) while reminding us that the names we do know (Auschwitz, Warsaw, Łódź, Dachau) are the places from which we have the most survivors. This selection problem is extreme: “Only two or three Jews who disembarked at Bełżec survived; about 434,508 did not.”

Big Sur, Jack Keruoac

I had never read anything by Jack Kerouac; I picked this off of PF’s bookshelf in Bristol before we spent a rainy day walking from pub to pub. The coastal descriptions and the long poem at the end were what I read the book for, and were beautifully done. The San Francisco (”Frisco”, as he calls it, and MWG forbids me to call it) scenes were chaotic and not interesting to me. These were the parts most like other Beat writers, and really not for me.

Imperial Spain, John Huxtable Elliott

A good history of Habsburg Spain. I was mostly interested in the story of how Spain lost its massive empire, but the contingencies and cultural quirks which allowed it to gain the empire so quickly was even better. This part mostly focuses on how Ferdinand and Isabella built effective institutions after unifying Castile and Aragon through marriage. These were two very different kingdoms and peoples; it’s interesting to track whether Castile or Aragon has more relative power at any given time. Pizarro and Cortés, for example, were both Castilian. Elliott thinks it’s impossible than an Aragonese could have done what they did.

Elliott is frustrated with how much the Habsburgs fumbled the economic exploitation of the Americas in the 1500s. There was no deliberate mercantilism; in fact, they were overly protectionist against their own colonies. There’s this great quote in one of Braudel’s books, something like, “To develop a colony in the Americas, it is first necessary to cultivate a province in Europe.” The Castilians just... didn’t do that. There was no deliberate effort to develop their American holdings into an economy which would complement Spain; just pulling more and more silver.

This had the predictable (to economists) effect of devaluaing Habsburg currencies and making Spanish exports non-competitive -- both of which happened. But Elliott is skeptical of the price-effect theory. Gold and silver were coming in boatloads, Spanish end-prducts (mostly texttiles) were exported to the Americas, so of course prices are going to go up in Spain. Elliott’s counter arguments are mostly of the form “we don’t have complete data on gold and silver exports to Seville.” The complaints are mostly about noise, but he doesn’t really discuss the variation in flows that we do have.

But, all of Spains economic problems in the 1500s were excerbated by the Habsburgs, and especially Charles I of Spain (Charles V of the Holy Roman Empire) being so thinly spread. The silver of Potosi was his not Spain’s, and that really messed everything up:

In spite of the Emperor’s success in obtaining a regular tax from the Cortes in the form of the servicio, he only managed to raise the Government’s revenues by 50 per cent during the course of his reign, whereas there was a 100 per cent rise in prices during the same period.

The Day’s Work, Rudyard Kipling.

A collection of the short stories. The most interesting was “William the Conqueror”. This is not at all what i associate Kipling with. It’s almost Wodehouse-ian, but mixed with very clear experience of the worst parts of history, and unblinking at horror. There is the light propaganda effect, and some dismissive comments at locals, but that’s as White Man’s Burden as it gets.

A Company of Swans, Eva Ibbotson

I love Eva Ibbotson’s romance novels, but she only wrote a half-dozen of them and they all have the same plot. This plot also happens to be Ibbotson’s own life: Central European woman, often Jewish, flees a revolution to England, where she meets an older academic (usually Cambridge); they fall in love. Sometimes she’s a ballet dancer; sometimes he’s a duke. They’re all great. But because there are only a few of them, I’m rationing them. This was the fourth I’ve read, and it’s just as fun. Her best moments are as funny as Wodehouse.

Travel is a pure way to think about monetary economics. We all have so many associations attached to the dollar or the pound (one bar of chocolate, one-fifth of a bus ticket, one-one-thousandth of my rent) that encountering a novel and abstracted unit clarifies the monetary intuition. In Oxford or San Francisco, money is just what it is. In each new country, I had to take at least one arithmetic step to decide if a transaction was worthwhile.

The Metical, Mozambique

$1 = 64 MT

I was in Maputo for a month, which is about enough time to build up price fluency. In the first few weeks, I got very good at dividing by 65 and continued thinking in dollars.1 Eventually I learned reasonable prices for taxi rides, meals, services and anchored on those. Learning prices is much more important in contexts where negotiation is common.

The 1,000 MT ($15) note is the largest note, making it more convenient to carry than Sierra Leone or Gambia, whose largest notes are ~$2. The major inconvience comes from actually getting the notes. There are plenty of ATMs around, but they are often empty or have queues. Informal money changers are difficult to find outside of the airport, so I would often go out of my way to find a working ATM.

The smaller the note, the more difficult it is to find. ATMs only gave out 500 and 1,000 MT notes; the one money-changer I did talk to offered smaller notes at about a 3% premium. Despite this, vendors and taxi drivers have a strong preference for bigger notes. One representative transaction was a Yango ride priced at 90 MT ($1.41).2 Drivers strongly prefer to receive a 100 MT note and return 10 MT change over receiving 50 MT + 2×20 MT.

With this dynamic, why would I need small bills? I should be able to get by on just the 100 MT and larger notes. Well, often the vendors don’t have small bills to offer as change. If drivers don’t have 10 MT, then their preference is to receive the 100 MT note and return 10 MT over mobile money. For complicated reasons3 I was staying off the mobile money system, and would prefer to just pay the 50 MT + 2×20 MT. At the end of the trip, when I was trying to get rid of small notes, people would spot the bigger notes and not accept the small ones once they know I have big ones.4 I model this as notes having some fixed handling cost per note, and vendors minimize this cost by holding fewer, larger notes.

Aesthetically, the metical is a beautiful note; each denomination is a very bright color. They refreshed their circulation in 2024, and I rarely saw older notes. Like most African countries, Mozambique doesn’t print their own currency; the metical is printed by an American company in Malta.

The Dalasi, The Gambia

$1 = D73

Aesthetically, Gambia has the rattiest bills, but they’re also the softest. They haven’t refreshed them since before Covid, and their bills are 100% cotton, unlike the metical which are largely polymer. They hold creases well and sometimes crumble at the corners. Unlike Sierra Leone, the value of the bill doesn’t degrade with its quality. No vendors complained about bills which were deeply faded or had large tears in them.

The largest bill is the D200 ($2.75), so I carried much larger physical stacks of bills. Payment for dinner for four is a roughly brick-sized block of cash, and the waiter will stand at the table and count it thoroughly. While this might incentivize greater usage of mobile money or bank cards, usage rates for both are lower than in Maputo. I asked my colleauges at Africell and various other Gambians why mobile money isn’t dominant yet. The answers I got were a) low interoperability between networks, b) opaque cash-out fees, and c) prominent fraud cases in the media.

As a result of all this, Gambian vendors have a strong preference for small bills over large bills, reversing the pattern in Maputo. If I could pay with a D200 or a bunch of D20s and D50s, they’ll always prefer the smaller notes.

In Gambia, vendors can’t fall back on mobile money as they do in Mozambique. If they deplete their stock of small bills, they can’t service the next customer, so they’ll maintain the preference for the smaller bills. In Maputo, the hassle cost of handling many smaller bills dominates, and mobile money provides a fall back when smaller notes aren’t available.

The size of the largest bill also plays into this. Because the largest note in Mozambique is 5x the largest note in Gambia ($15 vs. $3), more transactions in Mozambique will require change. With an existing small bill constraint,5 many of these transactions were going to end up on mobile money anyway. Gambia’s stuck in the opposite equilibrium. With weak mobile money, vendors hoard small bills defensively, which keeps transaction costs high enough to prevent mobile money from seeming worth the adoption friction.

The Leone, Sierra Leone

$1 = Le 23

The first thing to understand about the Leone is that no one knows what it’s worth. When I lived in Freetown in 2021 and 2022, $1 was about Le 10,000, which was also the largest note available. This led to me carrying around bricks of cash to pay our enumerators or stay in a hotel.

In late 2022, the Bank of Sierra Leone knocked off three zeroes from all bills: one Leone went from $0.0002 to $0.02 overnight. The new bill was nicknamed the “new Leone”, which is colloquially abbreviated to “NLe”. This is not an official thing! Officially, there’s just the Leone, which is now worth about $0.04. Even though it’s been three years, billboards, taxi drivers, and restaurants remain confused about how many zeroes to say, even within the same conversation. When negotiating a ride to the ferry terminal, a taximan would vary between “give me ten” and “give me ten thousand” within seconds. It’s been three years.

There’s no economic reason to do a redenomination; it’s purely cosmetic. This isn’t like demonentisation in India, where they voided large notes overnight; it’s just dividing every number in the country —and exchange rates— by 1,000. My understanding is that big nominal numbers had become embarrassing and the extra zeroes were inconvenient. I’m calling this the shame graph:

Sierra Leoneans have a strong preference for the quality of the bill. Unlike Gambia, particularly ratty bills are worth substantially less. Lots of bills are ripped in half and taped together; waiters will come rushing back to to the table after discovering one in a stack. Some level of insistence will get them to accept the torn&taped bills, but they’re not happy about it. All of my bills came from ATMs originally, so presumably would be accepted back by the same banks. I meant to ask money-changers at the airport as I was leaving for the different rates they’d offer on quality of bill, but didn’t have time.

The bills seems to remain in circulation for longer. All of the bills I saw on this trip were issued in the redenomination of 2022, but I would occasionally come across brand new bills from ATMs in Gambia and Mozambique.

The Birr, Ethiopia

As far as I can tell, no one has made this joke yet:

$1 = 155 birr

For a short trip, like my five days in Addis Ababa, getting the order of magnitude right is more important than the exact rate: should I get 10,000 birr or 100,000 birr out of the ATM?

I am very confused about Ethiopia. In seminar rooms and the NYT, it’s presented as a healthy man of Africa: business parks! tourism! only a small regional war! And yet it feels very closed. Ethiopia’s export-to-GDP ratio is 5%, which is tiny relative to well-off African countries.

It’s also very digitally closed. My T-Mobile plan, which has gotten me unlimited free roaming in Afghanistan, Sierra Leone, Burkina Faso, and like 200 other countries, failed to work entirely. The national telecom doesn’t play nice. Ethiopia is also one of eight countries in Africa where Anthropic isn’t available, and the rest are like Somalia, Sudan, CAR, DRC. I cannot imagine why Claude is available in Burkina Faso but not Ethiopia; Claude itself has no strong hypotheses.6 Despite this, they have a very strong electronic payments ecosystem, and not only domestic mobile money: many restaurants accepted Apple Pay.

I ended up taking out way too much cash, and had a hard time getting rid of it. In every other country, I could pay with card at high-end hotels and restaurants, but only in Ethiopia was it encouraged. Perhaps because of this, the bills were the least visually and tactilely interesting, and the least worn.

I spent October in Maputo, Mozambique; my impressions of the city are here, and they’re very favorable. Thanks to ZL for traveling from London to visit!

I was there on a placement with the International Growth Center, working on digitalization and public finance. The IGC office in Maputo is very well-connected with government, and set up meetings for me with municipal and national government agencies. Digitalization is a hot topic, and African governments want to be seen as modernizing and efficient. My role was to find the compromise tools which give this appearance and also help the governments achieve their goals. Some of these are easy wins — e.g., supporting mobile money — but it’s a small intersection. I’ve had the most success pitching digital tools to support revenue collection.

I’ve also been contracted by the Digital Impact Alliance and the Gates Foundation to write case studies of AI use within African governance. These case studies will appear in a public report early in 2026, which I’ll share here. I’m specifically looking at how AI is being used in taxation and policy prioritization. Most of my case studies are from countries I work in. If you have other examples, please reach out!

In November, I’ll be in Addis, Freetown, Banjul, and back to Oxford. Please say hi! I’ll be moving out of my place in Oxford at the end of the month, so this is your last chance for a while.

Reading

Strange Pilgrims, Gabriel García Márquez

Twelve short stories, translated by Edith Grossman. Márquez blurs the line between his fiction and journalism more than usual. Five of the stories, he doesn’t say which, came from notes for journalistic articles which he never finished. I had completely forgotten reading ““I Only Came to Use the Phone”” in high school Spanish, and also forgotten how much I hated it. My favorites were “Sleeping Beauty and the Airplane” and “Light is Like Water”. On net, this collection isn’t as strong as No One Writes to the Colonel.

The Lost Symbol, Angels and Demons, Dan Brown

Brown is a really fun read and an important part of some canon. My only familiarity with him was from this famous and funny review. I love Tom Clancy! I love Michael Crichton! Obviously I was going to like Dan Brown. I read each of these books in few hours; there’s no need to ingest the prose like Márquez’ stories, or reflect on character development like Pullman’s series. It’s just a dude, with some interesting ideas, and a fun story about those ideas. Also, I don’t think I’ve read a book with so many twists since Terra Ignota. Twists are fun.

His Dark Materials, Philip Pullman

I live right next to the Jericho boatyard in Oxford, which is an obsession of Pullman’s. The boatyard has been derelict and trapped in re-development hell since 1999. It also features heavily in The Golden Compass, the first book in Pullman’s trilogy. I partly picked this book up for preemptive nostalgia for Oxford, but also because of how much a few friends recommended it. I thought it was excellent, on par with the Narnia or Earthsea books. Fun metaphysics, and I’m a huge fan of “Let’s go kill God” plots.

The Space Trilogy, C.S. Lewis

I recently read Nathan Goldwag’s Narnia review and felt nostalgic for when I read Lewis as a kid. The Space Trilogy is much weirder than Narnia, and engages with Lewis’ politics much more. Honestly a bit too much Satan, and a really weird cosmology. The books work better if you let that wash over you. The biblical references I got; the weird medieval stuff I didn’t. The last book could just be a longer version of Lewis’ essay “The Inner Ring”, which is frequently recommended around Oxford.

I think that The Space Trilogy explicitly takes place in The Lord of the Rings universe.

“Tell me slave, what is Numinor?”
“The true West.”

Links

I’m not paying for The Argument, but I should be. Lots of great articles this month, even though I skip all the polling. This article —

The Argument

ChatGPT and the end of learning

I have taught an advanced computer science class at UC Berkeley nine times. The class, Introduction to Database Systems, teaches how data is organized, stored, and managed by computer systems, and is meant to prepare students for their first forays into the software engineering profession…

Read more

a year ago · 148 likes · 31 comments · Lakshya Jain

on ChatGPT and education left me feeling very lucky. On one hand, I’m jealous of my cousins, born 5-10 years before me, for growing up in the 90s cultural moment. On the other hand, if I had been born five years later, I would have been screwed by AI. I’ve always been a lazy student, and I never would have done a piece of homework or written an essay. I think I never would have learned.

Very useful.

I mentioned Nathan Goldwag’s Narnia reviews above. I think he has a really good ear for prose, and I’m linking to it again.

You have to imagine that a not-insignificant part of this book’s success was the title, right? Like, I’m not saying it’s the only reason, but it’s such a great hook, it drags you in immediately, tells you what kind of story you’re going to get, but doesn’t give you any hint of what’s going to happen. Marvelous. The Lion, the Witch, and the Wardrobe is the first Narnia book, and probably still the best. Like any series, Narnia accumulated lore and backstory and history and canon, but this first book (like the original 1977 Star Wars) gets to just be a story, without having to worry about any of that. One of the things that’s so delightful about it is how unconcerned the novel is with explaining everything. At 186 pages, it’s even shorter than The Magician’s Nephew, and it manages to pack in multiple trips to Narnia, tea with Mr. Tumnus, Edmund’s betrayal, capture, and redemption, war with the White Witch, meeting Santa Claus, the Crucifixion and Resurrection of Our Lord and Savior Jesus Christ, and an entire life lived as Kings and Queens of Narnia. Modern writers could stand to learn something here!

Jack Clark of Anthropic wrote a poem.
All Of Us Will Talk

There are maybe a million people in the world,
Who know what is coming,
And there are maybe six billion religious people in the world,
Who know where we come from.

At some point soon there will be a reckoning,
And the two sides will get to count,
And we’ll learn which futures and which histories get to survive.
The results will not be pretty.

Tom Cunningham, recently moved from OpenAI to METR, gives his thoughts on academic economics and transformative AI. Most interesting are the sections on GDP’s faults as a measurement, his model of AI as information-sharing, his conjecture about dimensionality and bottlenecks, and the ethnography of senior economists.

The Psmiths on class in America:

Mr. and Mrs. Psmith’s Bookshelf

JOINT REVIEW: Class, by Paul Fussell

Class: A Guide Through the American Status System, Paul Fussell (1983; Touchstone, 1992…

Read more

a year ago · 350 likes · 147 comments · Jane Psmith and John Psmith

How did I not know that The Paris Review was originally a front for the CIA? One of my favorite magazines.

The best way to know where you fall in America’s class system is to pay attention to which part of Fussell’s book make you feel uncomfortably seen. When you cringe and say, “Oh gosh, how did he know?!” that’s when you’ve found your your [sic] people.

Another short story take on “The Ones Who Walk Away from Omelas”, about growing up.

Isochrone maps show the area accessible from a central point in a given time. This article:

The Garden of Forking Paths

We are different from all other humans in history

Thank you for reading The Garden of Forking Paths. If you value my work and want to support it, please consider upgrading to a paid subscription for $50/year, which is just $4.16 per month. Your support makes my research possible so I can keep writing for you and keeps me full…

Read more

a year ago · 445 likes · 32 comments · Brian Klaas

has one centered around London, comparing 1914 and 2016.

This article starts with a top-tier anthropological note, hits some fun history, contains good media criticism. And then the twist actually surprised me, despite being heavily telegraphed. Very well written, and yes I will change my purchasing behavior.

Homo Imaginari

How Gillette Used Gender to Make Face Shaving Needlessly Worse for Men

Let’s start by setting the right benchmark for a perfect facial shave. The undeniably smoothest, cleanest, and best facial shaves I have ever received came from a cheery Tamil man inside a dusty, un-airconditioned shed, amidst the honking, screeching, and clattering noisescape that is any Indian intercity highway before it reaches the rural areas…

Read more

a year ago · 228 likes · 79 comments · James F. Richardson

On my bookshelf as a kid , there were ten Calvin and Hobbes books, and one Far Side book. I never liked the Far Side. Jane Goodall loved the Far Side! I like that for her.

The Leap

Funny or offensive?

Last week, I received an inadvertently hilarious note from a (now former) paid subscriber to The Leap. Inadvertent, for them; they, I’m sure, intended it as serious rebuke. But perfectly timed and perfectly appropriate from my perspective…

Read more

a year ago · 224 likes · 58 comments · Maria Konnikova

Cool car engineering.

We get November baseball! This is the first baseball season I’ve paid attention to since before Covid, because I’ve had favorable timezones up until October. Still, I’ve been waking up around 5am to catch the end of playoff games, and it’s been a great postseason. For a good sense of what MWG has put up with this summer, watch the first 20 minutes of Caught Stealing. This was a great article about falling in love with baseball in Oakland.

Residuum

Beisbol

Oakland, as everyone who cares about sports knows, managed to lose all three of its professional sports teams thanks to a combination of a corrupt and incompetent city government and craven team owners. I love this shithole city where I was born and raised, but my brand has always been sports hater, so it didn’t really break my heart when the Warriors, …

Read more

a year ago · 45 likes · Kaya Oakes

At some point last year, sick as I was, I told my spouse, “let’s just go.” Tickets are cheap, it’s a short drive from our house, and if we got bored, we could always leave early. But halfway through the game I was grinning and didn’t stop. Baseball, as my dad knew all along, is fun. It’s a slow game of strategy, not the relentless pounding of basketball or hockey or the violent onslaught of football. With baseball, at some point, you just surrender to the zen of the game.

I spent the month of October in Maputo, Mozambique. I didn’t leave the city other than day trips into the neighboring city of Matola, a boat trip out to Inhaca Island, and a trip across the tallest bridge in Africa and back.

Maputo is a very pleasant city. It’s easy to get around, very cheap, very safe, and quite fun. I’ve previously lived in Sierra Leone and Gambia, but Maputo feels much more like Mexico City than either of those.

This post is about the city. I don’t think there’s ever going to be much tourism demand for Maputo. It’s very far away for European and American tourists; most of them are going to end up in Cape Town. But it’s an interesting Place, with an interesting history and current political and social climate.1

History of Maputo

My sources: The best history of Mozambique is Malyn Newitt’s 2017 Short History of Mozambique, an abridged version of his much larger 1995 History of Mozambique. I also found the Marine Corps Intelligence Agency’s country study on Mozambique useful. My understanding is that the MCIA produces these 300-400 page books on every country in the world for the embassy guards, and there’s a correspondingly useful focus on cultural norms, good places to eat, and the inadequacy of the Mozambican Navy.

Maputo sits at the mouth of the 50-mile wide Maputo Bay, into which four rivers empty from the Swazi and South African highlands. There are no primary sources for Mozambican history before the 16th century, and few for Maputo until the 18th century.2 Portuguese explorers landed briefly in 1502 and 1544, but mosquitoes drove them off. The city is on a great fishing spot, so there were numerous Bantu fishing villages around, but historians don’t know much about their lives, and none of these sources speculate. In the 18th century, Maputo cycled through Dutch (trading ivory and slaves, losing money, driven off by a local tribe), British (passing by on their way to India), Austrian (the Holy Roman Empire built two forts; again, ivory and slaves), and back to Portuguese hands in 1781.

Drought is the constant worry of southern Mozambique. The Portuguese history of Maputo is detailed from the 1780s, and in 1795 drought struck the colony. During this first recorded drought, the French attacked the Portuguese colony, and occupied it. The Portuguese returned for good in 1799.

How rough was Portuguese colonial rule in Mozambique? I’d say worse than the British or French in West Africa, not as bad as the Belgians or Germans, and about the same as the British in East Africa. Most importantly, Mozambique was the primary supplier of slaves to Brazil and Cuba. Newitt estimates that more than 2.5 million slaves left Mozambican ports for the Americas between 1800 and 1850. Most of the early exploitation (until around the 1890s) was on the coast well north of Maputo.

It wasn’t until the British began moving north from their South African colonies that Portugal really focused on Maputo and the south. There was an indigenous state on the coast north of Maputo, the Gaza Empire. Newitt:

The continued existence of a powerful and independent African state weakened Portuguese authority vis à vis the African population in the south and cast doubts on Portugal’s ability to control its colony in the opinion of the watchful British and Germans. Moreover, the Portuguese wanted to open up the area of the Gaza Kingdom for labour recruitment. Early in 1895 the new High Commissioner, António Ennes, decided to force a showdown with [Gaza Emperor] Gungunhana. A large expedition was sent from Portugal supported by gunboats, cavalry and machine guns and in a rapid and decisive military campaign Gungunhana’s regiments were defeated and the king himself captured.

Three years later, the capital was moved from Mozambique Island, 1,000 km to the north, to Maputo. It was at this point that the central city took its current shape:

The swamps were filled in, connecting the low city with the higher bluffs, and a grid was laid out. The new capital took administrative focus away from the north; while the slave trade was 50 years dead, there was still intense labor exploitation in mining, railway construction, and agriculture. The south saw a relative growth in industry and trade.

Portugal maintained its own form of apartheid in its colonies through the 20th century. The central part of Maputo was de facto white-only. While black Mozambicans could acquire “assimilated” status, “Very few Mozambicans went through the difficult process of acquiring … assimilado and those who did were usually closely connected to the regime.” There was also forced labor through the chibalo system, through the 1960s. For any crime (Newitt highlights tax arrears, “inappropriate dress”, and “indolence”), black Mozambicans could be sentenced to penal labor, or sent to São Tomé or Macau. The police who enforced these laws were black, and were also recruited through compulsion and press-gang raids. There was substantial voluntary and involuntary migration to the mines in the South African Rand.

Mozambique’s war for independence escalated gradually from 1964, and was almost entirely a rural affair. Much like the current ISIS problem in the northernmost province of Cabo Delgado, the guerrillas operated from Tanzania. Maputo and the other major city, Beira, were untouched, but isolated. This is because the country was heavily rural: less than 20% of the country’s population at independence lived in any sort of town. The Portuguese took the war very seriously. Their operations in Angola and Mozambique wrote the doctrine for late-century counter-insurgency. They fielded large Africanised units, and mostly won both the Angolan and Mozambican wars in 1970. FRELIMO (more on them below) attempted to open new fronts in 1973, but it was mostly too late by then. Events in Portugal ended the war for independence permanently.

Portugal was the last of the European states to grant its colonies independence, but they couldn’t hold out any longer. The dictatorial government of Portugal was overthrown in a 1974 coup; the new government put a sudden end to the colonial wars in Angola, Guinea-Bissau, and Mozambique by granting them independence.3 The war of independence in Mozambique, which had been stagnating, immediately turned into a civil war between FRELIMO and RENAMO.

FRELIMO is a pretty straightforward 20th century Marxist-Leninist independence movement cum developmentalist single-party rule. It was founded in Tanzania in 1962 with support from Julius Nyerere and of course, Newitt notes, “suspicions had been aroused that the CIA had been involved.” FRELIMO, as the most prominent of the guerrillas at independence, was handed the government by default; their president at the time, Samora Machel, became president of the country.

There had been no popular insurrection and the most populous parts of the country (the provinces of Zambesia and Mozambique) had seen little fighting. Frelimo had not tried, and had certainly not succeeded, to mobilise the population or to establish itself in any of the main towns.

In the first year after independence, “most (possibly 90 per cent) of the population of European origin left the country.” FRELIMO had been established as a Marxist-Leninist movement, and cooperantes from Eastern Bloc countries began to fill in the gaps. They also seized most industry and established a command economy.

RENAMO is more complicated and no one will give me a straight answer. It was definitely founded and supported by white Rhodesian and South African intelligence agencies. During the war of independence, FRELIMO had developed strong ties to the Zimbabwe African National Union (ZANU), which was fighting against white-minority rule in Rhodesia. After independence, FRELIMO continued support to ZANU, and the Rhodesians didn’t like that. This was the beginning of what Gerard Prunier calls “Africa’s World War”: at the same time, South Africa invaded Angola, further destabilizing it right after its independence. The broader conflict spread throughout southern Africa, and there’s a line of continuity to the conflict in the eastern DRC today. Back in 1976, according to Newitt, “the first sign” Mozambique would be drawn in was “the formation by the Rhodesian intelligence service of an armed guerrilla force to carry out sabotage inside Mozambique,” while “Rhodesian forces carried out major raids on the ZANU camps” within Mozambique.

It almost ended there, with the change to black rule in Zimbabwe in 1979. Of course, the South Africans couldn’t let it be, and took over training and funding the anti-FRELIMO militants. Around this time, RENAMO became the predominant anti-FRELIMO group. This is where I lose track of the civil war. Newitt writes: “It is not clear that the Renamo leaders and their South African backers actually wanted to achieve the overthrow of the Frelimo regime.” So... what did they want? South Africa was definitely opposed to any country ruled by blacks which would help the ANC or SWAPO, but didn’t seem to have any end goals than to sow chaos. As best I can tell, since its founding RENAMO has defined itself as a “FRELIMO-bad” movement.

It was a bad civil war, as these things go.

During the fighting Renamo employed methods that were at that time relatively uncommon. Children were kidnapped to become soldiers and were forced to carry out killings. Rape and the mutilation of civilian captives was widely practised and people were also enslaved to provide Renamo with a work force. Such methods were to become all too common in wars throughout Africa in the following decades but some writers have seen in this behaviour a continuity with the pre-colonial past where much of Mozambique had been subject to warlords and their violent enslavement and exploitation of the population.

Oddly, Newitt quotes, “Renamo must have been the only so-called anti-communist guerrilla force which Washington did not actively support.” My guess is that DC had helped found FRELIMO, and felt silly flip-flopping within a decade. Or they just weren’t pay much attention to Mozambique.

The civil war petered out in 1992. FRELIMO had started to move away from its Marxist roots in 1983, and completely gave up on Marxist-Leninism at its fifth party congress in 1989. This had been one of RENAMO’s major demands, and both parties started seeing a route to normal politics. Everyone was exhausted by the war.

Throughout and since the civil war, FRELIMO was dominant in Maputo and the south of Mozambique. My sense is that, after the mass exodus of Europeans after independence, Maputo was a ghost town and quickly FRELIMO established its socialist command economy of the 70s and 80s there. This is why all the street signs look like this:

Aid flowed into the country after the civil war and the de-socializing. The rebuilding and investment were strongly focused on Maputo, and FRELIMO benefited far more than RENAMO, which was now acting like a totally normal political party that hadn’t done any war crimes. Some notable investments were an aluminum plant in Maputo which consumed 45% of all Mozambique’s electricity and several hydro dams to power the smelter. Since the end of the war, Maputo has been the social, political, and economic center of Mozambique.

Recent history

FRELIMO continues to hold the presidency, 70% of the legislature, all governorships, and the mayorship of Maputo. RENAMO holds a few other mayorships and seats in the legislature, but hasn’t seriously contested the presidency since 2014.

The most recent election was the 2024 national election. The FRELIMO presidential candidate, Daniel Chapo, won with a reported two-thirds of the vote. The runner-up was Venâncio Mondlane, a populist candidate who had left RENAMO for PODEMOS,4 which had been founded by a splinter of FRELIMO in 2019.

The official results reported that Chapo won 65% of the vote and Mondlane won 25%. Almost everyone I talked to in Maputo thinks that Mondlane actually won, and FRELIMO just did a really good job of fixing the election. Surprisingly, the more connected to the current government someone was, the more willing they were to admit these suspicions. Yango drivers and one particularly passionate fisherman were the only people who told me they thought Chapo had actually won the election. PODEMOS’ popularity was strongly concentrated in urban areas, regardless of region, which probably contributes to the bias of my sample.

After the 2024 elections, there were large and deadly protests on behalf of Mondlane and PODEMOS in major cities. The first 2-3 months after the election saw 100+ deaths during protests, and Mondlane fled to South Africa at least twice. Mondlane and Chapo met in person earlier this year, and Mondlane has since founded a new party, ANAMALALA.5

There weren’t any protests while I was in Maputo. At this point, people’s suspicions are mostly resigned; no one told me that they thought Mondlane would have been a better president anyway. More often I’d hear that they wanted a change after 50 years of FRELIMO rule. One topic which came up a lot was the high compensation for government workers, mostly in benefits rather than salaries. This was prompted by the IMF noticing that the government was spending 60% more on salaries than expected in 2023, and requested that the government reduce their bureaucratic wage-to-GDP ratio to under 10% (!!).6 Chapo promised to do this (in English), but also promised (in Portuguese) not to make any wage cuts. Government workers are upset about this, local policy wonks are upset about this, other Mozambicans assume all government wages are just rent-seeking.

How poor is Mozambique?

Mozambique makes the bottom of a lot of lists of development indicators, and that felt subjectively wrong to me. Ken Opalo recently shared this graph:

In public health, Mozambique is bottom-10 in malaria, HIV, TB, maternal mortality. Maputo is definitely an African city, but there’s a visible middle class, and not a visibly high disease burden.

I lived in Sierra Leone for a year. Most of that time was in the capital Freetown, but I also spent four months on the road, visiting every district in the country. Sierra Leone felt much poorer than Mozambique, despite having better economic indicators (GDP per capita, HDI, median income, GINI, etc.). Mozambique also feels richer than Gambia, which is twice as rich per capita! I think there are a few reasons why the stats don’t match the subjective feel of Mozambique.

First, I only visited the south of Mozambique, and mostly the urban area around Maputo and Matola. Mozambique is primarily an agricultural economy (about 85% of the workforce), and unlike Sierra Leone, a majority of farmers are subsistence farmers. The farms that I saw were relatively large and targeting the South African export market. The subsistence farmers are mostly in the north, out of my view.

The north sounds like a different world, and that’s why I called this “Notes on Maputo” instead of “Notes on Mozambique”. Besides the very different economy, the northernmost province, Cabo Delgado, also has a minor ISIS problem. There is no security threat in Maputo.

My urban/regional limit doesn’t explain most of the difference in city-feel though. I’ve never left urban Gambia, and the nicest areas of Maputo are much nicer than the nicest areas of Kanifing, and the worst areas of Maputo are slightly nicer than the worst areas of Kanifing. Again, GDP per capita and median income in Gambia are more than twice those of Mozambique.

A second and important factor is Mozambique’s relatively massive exports. Their exports make up 45% of GDP, compared to 20% in Sierra Leone and 7% in Gambia. Being right next to South Africa also helps.

Third, many Portuguese never left after the late decolonization, or have come back since the end of the civil war. This population creates strong demand for middle class amenities. Almost all of the expats in Freetown are on temporary postings, so fewer nice houses get built, less nice furniture is imported, and there’s less investment in communities. For expats in Freetown, a lot of the social life revolved around embassies. These are the institutions most worth investing in long-term. The social spaces, bars, gardens, restaurants are what made Maputo feel comparatively lively.

Finally, it’s really hard to count stuff in Africa, and all of the development indicators might just be wrong. Oliver Kim

Global Developments

How Much Should We Trust Developing Country GDP?

Poor Numbers: How We Are Misled by African Development Statistics and What to Do About It, by Morten Jerven. 2013…

Read more

2 years ago · 80 likes · 11 comments

and Adam Salisbury make the point well; statistics agencies are understaffed and have different incentives. This could have led to a systematic bias against Mozambique somehow. I don’t have a hypothesis of why this might be, but I am generally not very confident that these numbers are “right”.

A related point is that Mozambique is only at the bottom of lists of a subset of stable countries. If Nigeria’s economic statistics are mismeasured, then countries at war or without a functioning government certainly are doing worse. Afghanistan used to have the worst maternal mortality in the world, but not anymore! Once all the NGOs and aid agencies withdrew, there was no one to measure maternal deaths anymore, and the Taliban aren’t prioritizing it. During the American occupation, there was great data on health, and no government with the weight or incentive to pad those numbers. Even then, there was no convincing evidence that Afghanistan’s maternal mortality was worse than Somalia, Burkina Faso, or even Pakistan. It’s the same with Mozambique. If we had good data from Sudan or Somalia, they’d probably look substantially worse in comparison.

What’s Maputo like now?

Maputo is very green, and the streets are very wide. The old city, the parts built during early Portuguese colonialism, are on a neat grid. While the city is mostly flat, the 18th century part of the city, Maputo Baixa, sits at sea level, and Maputo Alta sits on bluffs about 50 meters above the oldest buildings.

There are also a lot of tall buildings. The central business district has several 20+ story offices, and there are many 10+ story residential buildings. This was really surprising! Freetown had only two buildings >10 stories when I was there. I think Mozambique’s skyscrapers exist because there is especially high demand for living in the old and white part of the city, even since independence. There is no equally desirable area in Freetown.

These incentives are lessening, especially since the big road to Matola (2019), the city to the west, and the bridge to Katembe (opened in 2020) were completed. Many wealthy Mozambicans now live in Matola, and on my trips to Katembe I saw dozens of very large and modern houses being built. I never saw a traffic jam, other than rush hour on the Matola road.

Maputo feels like a high-trust city, relative to other cities in Africa. Drivers follow traffic laws: there was strong adherence to traffic lights, which I haven’t seen anywhere in West Africa. Motorcyclists lane split, but with only a minor death wish. None of them wear helmets. I also didn’t see any difference in how nicer cars drive. The two equilibria I’ve seen are that nicer cars get pulled over more often, and therefore drive more carefully (Gambia), or they have tacit immunity and drive like it (Afghanistan). Even the President’s motorcade only took up one lane of traffic.

High trust doesn’t equal low corruption. I got pulled over in a friend’s car after paying the toll ($1.50) for the Katembe bridge, and the police officers asked both of us (addressing my friend in Xitsonga) for help paying for food. This happened again on a highway to Matola. My Mozambican friends all told me that these road checkpoints are very new, because control of road policing had shifted from the city to national police last month. One secretary to a government official told me that she was “very hungry,” and my meeting would start sooner if I helped her with lunch. Each time, it was easy to politely decline to pay the bribe. The x-ray operator at the airport, on my flight out, also told me, “I’m very hungry. Can you help me?” I remembered that I had 90 metical ($1.35) in my pocket that I had to get rid of, and offered it to him. He looked like he thought he was on candid camera, and quickly explained he had been joking. It seemed that no one had ever taken him up on this before.

There is a good amount of interesting architecture in the city.

Tyler Cowen once put this book on his annual list of favorite non-fiction. I haven’t splurged on it yet, but there are definitely Maputo buildings which I could imagine being in there, unlike Freetown or Kanifing.

Even the average street in the older part of the city is attractive, in part due to the abundance of acacia trees and the wide avenues. But there’s also a lot of remaining attractive colonial-era architecture between the soviet-style apartment blocks.7

Because of low traffic, trees, and wide avenues, Maputo’s streets are the quietest I’ve encountered in poor countries. In Kolkata and Freetown, the streets are so narrow and crowded that one horn will propagate and echo. To make up for its quiet, Mozambique is incredibly loud in social and common spaces. At an open-mic night at the Associação dos Músicos Moçambicanos, the speakers were placed so close to the small crowd that my beer jumped with the bass. I got my first taste of vuvuzelas since 2010 not at the Mozambique-Guinea World Cup Qualifier, but walking past a political rally. Santa Maria beach on Inhaca Island was the most isolated and quiet beach I’ve ever visited, until some Mozambicans offloaded a 3-foot tall, $2,000 speaker and blasted dance music. No one danced.

Besides this, the social scene in Maputo is pretty laid-back and a lot of fun. Middle- and upper-class restaurants and clubs are very well integrated, which is rare in West Africa. The expat community isn’t as separated from the wealthier locals, and even at very cheap restaurants in poor neighborhoods there were white Portuguese. Proximity to South Africa probably plays a part in this.

Getting around

Yango is the local ride-hailing app and makes life really easy. It’s a Dubai-based company, split off from Russian Yandex in 2024. Today it’s available as a ride-hailing app in 30+ African countries. Yango rides are really cheap, cheaper than taxis in Freetown or Kanifing, and quality was much higher. A 15 minute ride from my house to my office cost $0.23 on a motorbike, $0.58 in a tuk-tuk, $1.05 in a car, and $1.20 in the equivalent of an Uber Black car. I usually ended up taking the cheaper car option, because you can’t easily sit next to or talk to the driver of a tuk-tuk.

Yango eliminates the ability of taxi drivers to price discriminate. When I walked out of the airport, the taxi drivers wanted ~$20 for the drive to my flat, and I talked them down to $10. On Yango, the drive is $3. My general model of digital marketplaces is that they enable price discrimination (e.g., Uber surge pricing), but in Mozambique it’s the opposite.

Drivers are often professionals working a full-time job, or graduate students. I met a dentist, an engineer working at the new Total plant, several accountants. They also made up the majority of the Muslim Mozambicans whom I met. Most Muslims in Maputo are first- or second-generation immigrants from the northern provinces. One driver, Ismael, grew up on Mozambique Island, about 1,000km north of Maputo; it has a population of 10,000, but most of the young people move to Maputo. He came down to study engineering and now works at a terminal exporting natural gas.

Because I wasn’t on the local mobile money system, I paid for each Yango ride in cash. The car drivers rarely had change (tuk-tuk drivers always did). Alongside rating the driver in the Yango app out of 5 stars, you have to reply yes/no to the question “Did the driver ask for extra cash?” This led to the drivers being very reluctant whenever I told them to keep the change, even when I showed them I had already selected “no”. So we often went around to fruit sellers asking if they would break a twenty.

Food

The amount of good food, London-quality food in Maputo is surprising. If you’re visiting Maputo, you can reach out for a list of restaurants, as restaurant quality isn’t very correlated with reviews on Google or Facebook. I’d say Lumma was highest quality preparation and presentation,8 the Maputo fish market was best quality ingredients, and it’s impossible to choose best quality for money, because of the $1.50 beef, chicken, and fish dishes prolifically available on street corners. I cooked for myself much less here than in Gambia or Sierra Leone because of the lower prices at nicer restaurants, and the good quality of street food.

Inputs are just really cheap, and that makes food cheaper across the income distribution. One directly comparable setting: I visited the IGC Maputo offices during a training session for enumerators on a property tax project. Back in 2021, when I worked for IGC Sierra Leone, we held a comparable training for the same number of enumerators and same type of project. There, we didn’t pay for their breakfast, and we got them egg sandwiches for lunch. In Maputo, the IGC enumerators got sliders, cake, and a fruit platter for breakfast, and four hot dishes (two chicken, one beef, one fish) for lunch. This is especially surprising given the salary to enumerators is slightly lower in Mozambique: $150 per month, compared to (best memory) $200 in Sierra Leone.

There’s an abundance of beef, mostly imported from South Africa and Zimbabwe. I visited an abattoir north-west of the city, relatively small-scale, which processed in the low-hundreds of cattle per day. Most of it is low quality and hamburgers are an available splurge even in the poorest neighborhoods. Despite this, there is absolutely no dairy. The supermarkets have that horrible shelf-stable milk, which I drank because I love milk, but there were no local options for milk, cheese, yogurt.

The supermarkets were impressive; I put this down to the large Portuguese community. The two I visited were the size of American Targets; they sold bikes, leafblowers, and Christmas trees besides. There are at least two more that I didn’t visit, and they’re not part of a chain. There was nothing of this scale in Sierra Leone or Gambia.

What to do

Go to the beach

I joined a “Maputo expats” Whatsapp group before I arrived. The most common message, every Thursday evening, is “Hello! Is anyone going to Ponta this weekend? Two seats please; happy to chip in for gas.”

Ponta do Ouro is a beach town a couple hours south of Maputo. Until the Katembe bridge was finished in 2020, it required a slow, unpleasant ferry ride across the bay, or a seven hour detour. Now, the number of resorts and dive schools has exploded.

I didn’t get a chance to visit Ponta, but we did take a day boat trip across Maputo Bay to Inhaca and Portuguese Islands. Besides the loud music following us everywhere, these beaches were isolated and very beautiful.

Night life

As I said: loud, so bring earplugs. I wore my airpods on noise-reducing to one live concert. Also, things start earlier in the evening than you’d expect, perhaps because of the timezone.

Visit Mafalala

The Mafalala district is just north of central Maputo, just north of the old ring road built to separate the black suburbs from the Portuguese white city center. It’s a poor barrio; from satellite view, most roofs look like aluminum siding. But it’s seen heavy investment. The main roads are all newly paved with brick, and many of the drainage ditches which should have been dirty streams are now dirty concrete. A big part of this is historical personae and independence mythos: Samora Machel was born there, as was Eusébio, the most famous Mozambican footballer.

The level of infrastructure investment seems massively disproportionate to the history. Best I can tell, most of this is due to a tourism and community-building organization founded in 2009. IVERCA has also been remarkably successful at drawing fundraising for community projects — football pitches, affordable housing, and urban planning from the Danish government; World Bank and UN-Habitat for road and drainage improvements; the EU and German delegations funding a museum and backpacker’s hostel. Best I can tell, none of these would have happened without IVERCA.9

The museum is IVERCA’s crown jewel. It’s a very cool building, located at the highest point in the swampy neighborhood and by far the most colorful entry. The hostel is above the museum, and has two remarkable rooms hidden away in attics. I would recommend the place to stay for a night, but note it is not convenient to any other part of the city, and very difficult to get taxis or tuk tuks out of. The neighborhood is fun to walk around during the day.

There are many further proposals for development of Mafalala. The Danes have partnered with IVERCA and urban planning firms to design health centers, libraries, and community centers, but the museum docent told us none of that is funded yet. The most obviously missing part of the neighborhood was any type of commercial development: there were no markets or shops besides fast food stalls. The residents mostly commute into central Maputo.

Go to a soccer game

I went to the World Cup qualifiers, Guinea vs. Mozambique, at the Zimpeto National Stadium. The stadium is about ten miles north of the city, in a large dusty field next to a relatively new four lane highway. It’s large — at 42k seats, real, plastic seats in the colors of the flag, not bleachers, and built by the Anhui Foreign Economic Construction Group in 2011. Besides a brutalist clocktower on the south end and the colorful seats, there is nothing fancy about the stadium. The scoreboards on either end could only display the score, not a clock, and one-fifth of the pixels were out.

I went with a GIZ consultant on his first trip to Africa, and his first time at a football match outside of Europe. The first surprise came as we approached the steps to the main gate: Guinea scored in the second minute, and the crowd of Mozambicans still arriving heard the cheers, shoved forward, and pushed over the fences around the turnstiles.

I’ve been to soccer games in Sierra Leone and Gambia, in Afghanistan, and, once, to an A’s-Giants game at the Oakland Coliseum. This was probably the most chaotic venue I’ve been to. The ~$2 we spent on tickets was wasted; no one ever looked at them.

The guys sitting next to us really wanted us to stay sitting next to them. At the half, Laurin and I walked around the concourse eating cashew nuts. From 50 meters up the pitch, we could see these ten guys, with their hands up, hissing into the wind at us, to make sure we saw them and could find our way back. Maybe we’re good luck.

We weren’t. After Mozambique tied it at 1-1, Guinea went ahead early in the second half and Mozambique couldn’t do anything for the rest of the match. There were epic levels of shithoussery. Mozambique had a particularly dangerous freekick, the Guinean manager went for a substitution just before, but the player took so long coming off the field (despite his best efforts! an assistant manager shoved him back over the line onto the field several times to delay play) that the official cancelled the substitution. At another point, the keeper passed back and forth to a midfielder three times, just for fun.

Other notes

Mozambique shares a timezone with all the other Southern African countries. As the easternmost country in the timezone, this feels really weird! When I arrived, sunrise was at 5:20am, and sunset about 12 hours later. The early sunsets move business and social events earlier throughout the day. I’m a morning person, but it would usually be considered rude to suggest a 7am in-person meeting. Not here!

The Kalashnikov is a prominent national symbol in Mozambique; it’s on the flag. In particularly touristy areas, hawkers would try to sell me flags, jerseys, banners, scarves with ever-larger AK-47s on them.

I visited the Honen Dalim synagogue in Maputo for Kabbalat Shabbat. They only ever have a minyan during the high holidays, but we got together five jews for this Friday night. It’s a very cool building, and much more fun than visiting Chabad. They’re looking for a rabbi!

Cash is very annoying here, but that’s everywhere in Africa. While there are ATMs everywhere in the city, the three at the airport were all not working when I landed. I joined a line of six shamefaced white people trooping back to the moneychangers we had so confidently brushed off on our way to the ATMs. It is very difficult to get small bills, and this is annoying, and no one has change. I bluffed my way into stacks of 20 metical notes at a few banks, but they don’t last long. Merchants and Yango drivers are just as eager to get their hands on them. If you have a 20 metical note, they won’t give you change even if they have it.

The roads and sidewalks are really good; every road in Maputo is paved (not close to true in Freetown or Kanifing!). But there are a lot of sudden deep holes in the sidewalk. This is because manhole covers are stolen and sold for scrap, according to a police officer.

Outside the airport, there was a big Vodacom billboard saying, “It’s simple to get an eSIM.” It is not simple to get an eSIM. To activate the eSIM you need to first insert a physical SIM. My 2025 iPhone 16e doesn’t even have a SIM slot, which was verified by a dozen Vodacom employees at three stores. They all reacted with shock.

Contents

• Updates

• Reading

• Links

• Previously

Updates

Hello from DC! I am here until the 12th, and in New York City the 13th-15th, before heading back to San Francisco. If you’re in any of these places, please say hi! And sorry for getting this up a few days late, especially to QM and ZR who both texted to ask if I’m alright.

My wonderful girlfriend and I have a new apartment in San Francisco, two minutes from the nearest BART and 52 minutes from Evans Hall, where I’ll be spending the winter and spring semesters as a visiting student. The apartment is large and bright, and it will soon be furnished with a jungle/Mid-Century Modern vibe. The second bedroom will be an office for MWG and an occassional guest room for anyone passing through!

^[1]

I spent much of August in the South Bay. First, I did a few hikes around Portola Redwoods State Park, west of Palo Alto, which has for my money the best redwoods in the Bay Area. The Peter’s Creek loop, which I sweated out in three hours, was completely empty on a sunny Friday afternoon, and ended in a deep dark confluence of three creeks around a lush redwood grove. Arriving here was sudden, a much more dramatic reveal than any other grove I’ve visited.

Sorry—your browser can’t play this video.

Then I spent the last two weeks at Stanford’s Digital Economy Lab, for a summer course on AI run by my friends Phil Trammell and Zach Mazlish. The highlight was the attendees: I learned a lot about how young economists are thinking about the potential of AI. The upshot from the course material is that economic models predict massive change and uncertainty across economic indicators. So there are two possibilities: either things are about to get weird, or the models are wrong.

The general view among the people at this course is that the only thing that can stop this process now is AI capabilities slowing down. Maybe we’ll run out of data on which to train the models, or training won’t be as efficient at scale, or the potential for mass unemployment will lead government to step in and force a slow down. Maybe you think a slow-down is imminent; maybe you think the tech is moving so fast it’ll blast through any obstacle.

One attendee remarked that, in traditional debates of macroeconomics (business cycles, monetary policy, etc.), we just have to resign ourselves to never knowing the right answer. Not so in the economics of AI: in a year or three, we’ll know who was right and who was wrong.

Elsewhere in AI+Joseph news, I’ve been awarded an “Asterisk AI Blogging” fellowship. I’ll be writing a few posts on this blog and elsewhere about the role of AI in poor countries. The rest of the cohort (to be announced soon!) is really cool, and I’m excited to work with all of them.

Reading

Three favorites this month.

Super-Infinite: The Transformations of John Donne, Katherine Rundell

Rundell is now solidly on my “read-anything-they-write” list. I loved her Golden Mole, and her children’s fiction, and now her biography of 17th century English poet John Donne. Borges called Donne “a truly great poet,” even if he had only written these lines:

License my roving hands, and let them go Behind, before, above, between, below! O my America! My new-found land! My kingdom, safeliest when with one man manned.

Rundell provides a more thorough argument, a full biography, with considerations of the love and erotic poetry, the defense of self-homicide in Biathanatos, and the sermons. Donne as lover, polticker, clergyman, and friend.

I thank him for, upon meeting an elephant in London, calling it “The only harmless great thing.”

Disputing Disaster: A Sextet on the Great War, Perry Anderson

A fun historiography of the couses of WWI, with biographies of six leading historians who wrote on the topic. Anderson is a Marxist historian and has no use for counterfactuals.

Indigenous Continent, Pekka Hämäläinen

I had previously read his Lakota America.

The focus of this is on the agency of Native Americans in the colonial story of North America. The author makes some leaps in narrative, which leaves questions – a common path across the book is: look how well this confederation was doing, the English or the Dutch or the French are on the run and then! “In a sudden reversal”, all the indians are dead. Hämäläinen is definitely on the side of the indians; who wouldn’t be.

Links

The Panama Playlists are musical selections from celebrities and politcians’ Spotify accounts. “The Panama Papers revealed hidden bank accounts. This reveals hidden tastes.”

A silly but fun analogy of the Mamluk rule in Egypt to AI takeover.

Where are the [AI] models that :

• • have a knowledge cut-off of 1950?

• • are just trained on the best books?

• • are trained only on open CCTV cameras?

• • are just trained on books and youtube?

• • just on Pulitzer winners and the Criterion collection?

A lot of the bloggers I’ve followed for years are now having kids. Freddie Deboer has always been a kinda embarassing read for me; too polemical, too petty, and focused on the news of the day. If he would only write about things he loved every day^[2] instead of things he’s annoyed about…

So I’m very glad he’s had a kid, because how he writes about love.

I loved the movie High Fidelity.

Dan Wang’s new book has been all over my internet recently, telling you a bit about where I hang out. In his interview with Statecraft, Wang mentioned he’d re-read every issue of the CCP’s theory magazine, Qiushi. The title comes from the phrase “Seek truth from facts”, a favorite of Mao’s. MWG got me a tea mug with this slogal on it, probably due to my Bayesian past; it’s my favorite mug.

A sad and mouth-watering article about drug addict day-laborers in Las Vegas.

Everything they say is as impermanent as steam. One minute they are madly petting each other like teenagers. The next they are screwing other people, breaking bonds like they are toothpicks, finding the soft spot in the other and poking at it until it bleeds.

The coolest sports video I watched this year. If anyone knows a channel which posts these (from different teams, sports, venues), I would immediately subscribe.

Previously

August, 2025

July, 2025

June, 2025

May, 2025

April, 2025

March, 2025

February, 2025

January, 2025

December, 2024

November, 2024

October, 2024

September, 2024

August, 2024

July, 2024

June, 2024

May, 2024

April, 2024

March, 2024

February, 2024

January, 2024

December, 2023

November, 2023

October, 2023

September, 2023

August, 2023

July, 2023

June, 2023

May, 2023

April, 2023

March, 2023

February, 2023

December, 2022

November, 2022

October, 2022

September, 2022

August, 2022

…

November, 2021

October, 2021

September, 2021

July, 2021

June, 2021

May, 2021

1. 

   
   This is what Gemini made, based off of the jungle/MCM vibe and a picture of our living room with only a couch. ↩︎

2. like I linked to last month ↩︎