As Joel Mokyr once said: "The responsibility of economic historians is to remind the world what things were like before 1800."
Read this review to see how bad it was . . .
Spoiler: really bad!
Plague, gruel and bad teeth: why do people idealise the olden days? https://t.co/UGrLqm6Iry
Last weekend, I read Domar’s classic 1944 paper “The ‘Burden of the Debt’ and the National Income.”
As is often the case, I was familiar with the basic result by the time I was an undergrad: you can run a permanent government deficit and still keep the public debt-to-output ratio constant under certain conditions on the real interest rate, the economy’s real growth rate, and the primary surplus. In fact, you can even reduce the public debt-to-output ratio asymptotically to zero and still run a government deficit every year.
But I had never read the original paper. The language is often difficult to follow (yes, leaving the formal model in the appendix makes everything much harder; you’ve got to love math!). Yet the insight per paragraph is truly outstanding. Domar was an incredible economist.
Nowadays, I spend most of my days thinking about how public debt interacts with AI and fertility. Both forces put the fiscal path of many Western European countries in a very difficult position: AI will push up the real interest rate more than the growth rate, and falling fertility will lower the growth rate. In Domar’s terms, the gap between the real interest rate and the growth rate moves in the wrong direction. I truly do not see how France or Spain can avoid a debt crisis over the next decade unless they undertake serious reforms.
To wrap up: read Domar’s 1970 paper “The Causes of Slavery or Serfdom: A Hypothesis” as well. That is a heck of a paper, a true classic. Someone should write a scholarly study of Domar as an economist. If you are working on the history of economic thought and searching for a topic, Domar is low-hanging fruit.
même des collègues chercheurs (pas forcément mathématiciens) sont souvent surpris quand je leur dis que l'empreinte de l'IA n'est pas aussi grosse qu'ils le croient (ce qui ne veut pas dire que ça ne pose pas problème dans des endroits localisés, comme certains datacenters aux
Every time I think European voters cannot be more in denial about pensions, they prove me wrong: they are even more in denial than in my worst nightmare.
The average French person believes there are 6.4 workers for every retiree. The real number is about 1.8.
Hiç abartmıyorum, bu makale iktisat tarihinin ulaşabileceği zirveyi temsil ediyor. Makalede Pirenne tezi, 500 bin para sikkesinin coğrafi dolaşımı üzerinden test ediliyor. Makalenin en önemli bulgusu ise aşağıdaki iki görselde açıkça görülüyor: İslam fetihleri öncesi ve sonrası dönemde sikke akışlarının nasıl değiştiğini gösteren son derece etkileyici iki figür:
Superbe mémo pro-IA du CTO de Proton, envoyé à ses ingénieurs en septembre. C’est le texte le plus clair que j’aie lu sur ce qui arrive au métier de développeur. Très techno-humaniste. J'ai pas lu tout le reste de l'affaire, mais il y a 0 raison de se scandaliser face à ce texte.
Il part d’une question simple. Si tout le monde utilise déjà des LLM, pourquoi la productivité n’a-t-elle pas explosé ?
Sa 1ère réponse porte sur la manière dont les ingénieurs voient leur rôle. Ils utilisent l’IA comme un outil pour coder plus vite. C’est une erreur de cadre : « Votre travail a fondamentalement changé. Il ne consiste plus à écrire du code, et les LLM ne sont pas là pour vous aider à écrire du meilleur code plus vite. Votre travail est d’aider le LLM à écrire du meilleur code plus vite. »
Chaque ingénieur devient donc un manager dont les subordonnés sont des agents. Et un bon manager ne passe pas ses journées collé à son équipe : « une équipe où personne ne travaille quand le chef est en réunion ne serait pas vraiment une équipe ». L’objectif est que l’IA travaille seule et longtemps, et que chaque échec répété déclenche une correction qui profite à tous.
Sa deuxième réponse porte sur les goulots d’étranglement. Accélérer l’écriture du code ne sert à rien si la CI, les validations et la revue de code restent au rythme humain. « Que se passe-t-il quand on élimine un goulot ? Un autre prend sa place. » Il faut repenser les processus de bout en bout. Il va jusqu’à écrire que la plupart des revues de code humaines devront disparaître. Même les bonnes pratiques bougent quand les coûts bougent : si écrire et maintenir des tests ne coûte presque plus rien, on n’a plus de raison d’en écrire peu.
Il défend aussi l’humain sans naïveté. « Le LLM est rapide. Vous êtes lent, mais vous portez une vie entière d’expérience, de jugement et de compréhension. » L’ingénieur reste responsable devant ses collègues « d’une manière qu’un modèle ne pourra jamais l’être ».
Et sa conclusion : « La “programmation” est morte. L’ingénierie ne l’est pas, ni l’expertise. Votre ancien job n’existe plus. Votre nouveau job de bâtisseur vous permettra d’en faire plus et d’avoir plus d’impact que jamais, si vous saisissez l’opportunité. »
Selon lui l’IA a écrit 90 % du code de Proton en octobre. L’ingénieur devient le chef d’une équipe virtuelle, l’écrivain le rédacteur en chef d’une rédaction virtuelle. Ce qui est déjà vrai pour les dev va devenir vrais pour tant d'autres métiers dans les années à venir. Immense opportunité à saisir.
Our robots built this in the city of Almere: curved walls, different types of bricks and bonds, indented text.
Real world deployment, insane complexity, 100% of every brick done by our robots, with another 15 houses to come in the coming weeks.
Lots of debate this week on whether journals should ban AI in peer review.
We actually tested this in our new preprint, and the results were pretty clear to me.
For context, I've reviewed for most of the big journals and evaluated 1,000 human reviews as an editor. I love to read and evaluate science.
TLDR: From now on I won't do any peer review without AI. For clinical trials, I'd even call human-only review dangerous 🧵
Michael Pollan reveals how the introduction of caffeine completely changed human history and consciousness.
"Before caffeine, it was a very different world and a very different consciousness. People were drunk a lot of the time, buzzed almost all of the time because it was safer than water"
"Water was really how you got diseases. If you fermented things, even low alcohol, it killed a lot of the microbes. So people, even kids, drank hard cider for breakfast"
"Tea, chocolate, and coffee all arrive in the same decade in England, and things start to change... no one ate or drank boiling water or hot beverages before"
"Boiling it made it much safer. This gave this incredible public health boost to these places, and then you have the drug that fosters a kind of more linear, rational, focused way of thinking"
"People in the 1600s started writing about it saying, 'Wow, we have this new civil and sober drink that we can use.' And that, I think, makes possible things like the Industrial Revolution"
This chart is insane. This shows Russia's casualties over time compared to Russia's territorial gains in Ukraine over the last five years.
No sane or rational person would say continuing the war to try and take more of Ukraine's territory is worth the casualties.
When George Washington was inaugurated in 1789, he wore a relatively plain brown frock coat. To the modern eye, the garment may look fancy, but it was relatively constrained for a man of Washington's position.
Affluent Americans at the time tended to wear European fabrics, as centuries of textile tradition on that side of the Atlantic meant that their materials were finer, silkier, and more luxurious. Those who could afford to wear such materials often did, not just because they were more comfortable, but because they signaled status.
Washington rejected this convention for his inauguration, opting instead for a plain brown frock coat made from an American-made material, specifically one produced by a relatively new upstart in Connecticut. It's also notable that his outfit for that day was considerably more restrained than those of leaders in other Western nations.
For instance, check out this outfit on George III (slide one). In this 1762 coronation painting, we can see the British monarch wearing a gold damask coat with a velvet mantle lined and edged with ermine and trimmed in gold. Even his shoes featured solid gold buckles.
Washington opted for something simpler because he wanted to send two political messages. First, he wanted to support domestic industry (hence his request that the garment be made from American woolen broadcloth).
Second, he understood that the US was supposed to be a democratic republic, not a monarchy. Back in Europe, monarchs wore fancy ceremonial attire to reinforce their position in society. Washington, by contrast, chose something more restrained, signaling that he was not governing through divine right but rather through consent of the people.
Irish-born architect James Holban also understood this when he designed the White House in the Neoclassical style, drawing from Georgian and Palladian architecture. Neoclassicism was a rejection of Rococo, which had been fashionable among European monarchs and aristocrats in the preceding decades.
Most people will know Rococo from the photos of the lavish French palaces erected during the reign of Louis XV. Much like the elaborate ceremonial attire of European monarchs, these interiors were meant to impress upon people the wealth and power of European leaders. The visual grandeur reinforced the idea that these people fundamentally occupied a different social station from those they governed.
To be sure, the White House was much nicer than what ordinary Americans at the time would have occupied. But much like Washington's inauguration outfit, Hoban's design was meant to balance the dignity of the office with the republican ideal that political authority ultimately comes from the people. Our leaders are not above us; they are here to serve us.
That is why the White House was decorated in this relatively restrained way for almost 250 years, changing only recently. I have seen some on here suggest that glorifying the White House is meant to glorify the state, and thus the citizens. But this logic was historically used by dictators abroad, not US presidents.
Many of you don't understand how serious this is, so let me break it down.
In computer science there is a famous kind of puzzle called 3-SAT. You are given a long list of yes-or-no switches and a set of rules about how they must be arranged, and you have to find a combination that satisfies every rule. The obvious way to solve it is to try every combination, but each new switch doubles the number of possibilities, so a puzzle with a few hundred switches would take longer than the age of the universe to brute force.
For decades, the smartest people in the field found clever tricks that made this a little faster, but every trick still hit the same wall: the time required kept exploding as the puzzle grew.
About 25 years ago, two computer scientists, Russell Impagliazzo and Ramamohan Paturi, working with Francis Zane, turned that wall into a formal idea.
They proposed the Exponential Time Hypothesis, which says this explosion is not a failure of human cleverness but a law of nature. No algorithm, however brilliant, will ever solve 3-SAT without the time blowing up exponentially. Nobody could prove it, but almost everyone believed it, and it became one of the foundations of modern computer science.
Researchers used it as a measuring stick. If you could show that solving some other problem quickly would also crack 3-SAT quickly, you had shown that problem was hopelessly hard too. Hundreds of papers, across scheduling, graphs, logic and much more, rest on this one assumption.
Now an internal AI model at OpenAI claims to have broken it. It says it found an algorithm that solves these puzzles meaningfully faster than the hypothesis allows, and it has checked the proof in a computer verification system called Lean.
And here is the remarkable part: the very people who proposed the hypothesis, Impagliazzo and Paturi, along with MIT's Ryan Williams, are now leading the public review of whether a machine has disproved their own idea.
So what happens next? If the disproof holds, a large part of theoretical computer science has to be revisited.
Every result that said "this problem cannot be solved faster, assuming ETH" loses its foundation, and researchers will have to work out which of those walls were real and which were simply where we stopped looking. It does not mean everything becomes easy. The new algorithm is still slow in absolute terms and this is not the famous P versus NP problem being solved.
But it means our map of what is possible was wrong in a place we were very confident about, and it was an AI that redrew it. If the disproof fails, the hypothesis survives and the field breathes out, but we will still have watched an AI produce serious, checkable work on one of the deepest questions in the discipline, closely enough that the founders of the idea had to take it seriously. Either way, something has shifted.
Quand j'ai déménagé au Japon il y a une vingtaine d'années, la première chose qui m'a marqué, c'est que tout était japonais.
Les téléphones portables étaient Sony, Sharp ou Panasonic.
Les téléviseurs étaient Toshiba.
Les voitures étaient Honda et Toyota.
Les trains étaient Hitachi et Kawasaki.
Les escaliers mécaniques, les climatisations et les ascenseurs étaient Mitsubishi.
Toute l'infrastructure matérielle du pays semblait construite sur place.
Un ami marocain m'a dit qu'il avait eu la même sensation en venant en France pour la première fois il y a fort longtemps.
Des voitures françaises, des trains français, des ascenseurs français. Les téléviseurs étaient Thomson, les ordinateurs de l'administration étaient Bull. Il y avait même des téléphones portables Alcatel et Sagem. (J'en avais un, c'était pas ouf, et pas le sommet de la classe comparé au Nokia 3310, mais ça existait. Sagem vendait plus de mobiles en France que Nokia en 2004.)
En venant du Maroc, tout cela semblait très impressionnant et nouveau.
Nvidia produisait ses puces en France La toute première, en 1995, puis la RIVA 128 de 1997 qui a sauvé la boîte de la faillite : les deux sont sorties des usines de SGS-Thomson, l'ancêtre de STMicroelectronics. Nvidia était alors une start-up d'une centaine de personnes qui cherchait quelqu'un pour graver ses puces, et c'est chez nous qu'elle l'a trouvé. L'année suivante elle a pris TSMC en second fournisseur, puis l'a gardé seul. Nvidia vaut aujourd'hui deux fois l'ensemble du CAC 40.
20 ans plus tard notre environnement matériel et donc industriel est de moins en moins français. Thomson a cédé ses téléviseurs au chinois TCL en 2004, Alcatel ses mobiles la même année, Sagem a disparu des rayons en 2008. Nous avons aussi lâché la course aux puces : l'alliance Crolles 2, qui devait faire de l'Isère la pointe européenne de la gravure, est morte fin 2007 quand Philips et Motorola s'en sont retirés, et ST-Ericsson, dernière tentative européenne de fabriquer des processeurs de smartphone, a fermé en 2013 avec près de trois milliards d'euros de pertes. Les puces qui entraînent aujourd'hui les modèles d'IA sortent presque toutes de Taïwan.
Même histoire en robotique. Nao, le petit humanoïde conçu à Paris par Aldebaran, est devenu le robot officiel de la RoboCup, la coupe du monde des robots. Aldebaran a été vendu au japonais SoftBank en 2012, revendu à un groupe allemand en 2022, placé en redressement judiciaire en 2025. Les robots industriels de Renault, eux, étaient déjà passés sous pavillon italien à la fin des années 1990.
Cela ne fait pas mal par patriotisme blessé, pas même par ses conséquences sur la prospérité et l'emploi, peut-être même pas comme rappel constant des choix désastreux faits en conscience par une élite politico-industrielle sans vision.
Cela fait mal car alors que nous entrons dans un nouveau monde de tensions géopolitiques, notre capacité à maîtriser notre propre infrastructure industrielle est la seule garante de notre liberté.
Ce n'est plus théorique. Les Chinois menacent de couper les accès aux terres rares si nous ne laissons pas leurs produits finir d'achever notre industrie. Les Américains coupent les accès aux modèles IA les plus avancés quand ils craignent qu'ils puissent être trop puissants pour en laisser l'usage à leurs vassaux.
Nous sommes entrés dans le XXIe siècle et ses tensions en bazardant progressivement tout ce qui garantissait notre liberté et notre prospérité, sur l'autel de croyances absurdes. L'idée que la production intellectuelle était supérieure à la production matérielle. L'idée que la mondialisation garantie par l'US Navy serait là pour toujours.
Ces croyances absurdes n'ont pas encore disparu. L'idée simple et évidente que la France ne restera un pays riche et libre ce siècle que si elle triple son parc nucléaire et transforme son énergie abondante et bon marché en centres de calcul et en industries robotiques paraît encore, pour une grande partie de l'élite, incongrue ou blasphématoire.
Le parallèle a souvent été fait, mais il faut le répéter inlassablement : nous sommes la Chine des Qing qui refuse la machine à vapeur, nous sommes l'Empire ottoman des oulémas qui interdit l'imprimerie.
Nous sommes une vieille civilisation qui refuse d'acter sa décadence et les actions indispensables pour retrouver son rang.
Il est tard, mais pas trop tard.
for a generation born into digital age, the quiet tragedy of modern France, it’s watching the nation reduced to a digital colony
while US builds sovereign tech empires and China anchors its civilization in proprietary industrial stacks, europe acts as an open-air data farm, renting its cognitive infrastructure from foreign monopolies while local politicians pretend they still hold the steering wheel….
for my american, chinese and global friends watching the turmoil in france RN, stop swallowing the ragebait served by french engagement-farmers tweeting in english to pander to MAGA confirmation bias
these accounts are monetizing culture-war tropes & blaming immigrant communities to farm impressions while completely misdiagnosing the underlying systemic fracture
french youth across every demographic are taking to the streets bc they are suffocating under skyrocketing inflation, crumbling high school infrastructures and the acute anxiety of a future rendered obsolete by AI and the exact structural reality Taleb highlighted
when the promise of upward mobility vanishes overnight under the weight of the cost of living and technological acceleration, social unrest becomes the baseline response to a system running out of runway, creating an existential fracture that no traditional political party or police measure can contain
It's so daunting on the very slim occasions when I start the daily thread so late.
So much to catch up on.
Missiles hitting a Russian military plant, a POW exchange, another ship sunk in Romanian waters, a 'ceasefire' in Oleshky and the latest on peace talks.
It's Day 1703