Jeremy Wickersheimer

This is one of the biggest geopolitical mysteries of the year: why did China just save the world economy? Because, make no mistake, it absolutely did. Analysts are pretty much unanimous on this: had China not dramatically - and I mean dramatically - reduced its oil imports this year, the price of crude oil would have gone through the roof and dragged the global economy down with it. For proof, read how Javier Blas from Bloomberg - widely recognized as one of the world’s top energy analysts - explains what China did to Jake Sullivan and Jon Finer on their podcast “The Long Game” (thelonggame.substack.com/p/t…). As Blas puts it, China went “from 11, 12 million barrels a day to less than 6 million barrels a day” of seaborne crude imports, and that “freed a lot of oil for other countries, particularly in Asia.” India, Japan, Korea and Southeast Asia “could buy the barrels that China was not buying.” To be clear, China wasn’t the only factor - surging US exports and the IEA’s emergency release helped too - but it was by far the biggest: JPMorgan reckons China accounted for about 74% of the fall in global crude trade (fortune.com/2026/06/10/oil-p…). So it's no exaggeration to say that it is China, first and foremost, that saved the world economy this year. If you remember the early days of the Iran war, none of this is what was predicted. In fact some US think tankers, such as the (always horrendous) Zineb Riboua from the Hudson Institute, were even cheering the war as being “all about China… the opening act in the larger contest against China,” the moment the US would crack “the central pillar of a regional order that Beijing has assembled” (thefp.com/p/the-iran-strike-…). Seven months on, what actually happened is that not only did China come out of the war unscathed economically, and strengthened strategically, it also ended up shielding the world economy, America’s included, from the fallout of the very war that was supposed to be its undoing. Breaking news: if you do something that’s supposed to be a “blow” to someone else - some sort of 3D chess move to gain the upper hand - and they end up having to rescue you from the fallout, maybe it wasn’t such a brilliant strategic move after all. So how did China pull it off? And, more intriguingly, why did it choose to rescue countries, including the US, from a war that Washington’s own hawks were openly selling as a strike against China itself? And why do this when it effectively dampened Iran's leverage considerably, even though they're supposed to be China's "ally" (which, as we'll see, is false)? This is the topic of my latest article. I went very deep researching all the Chinese sources I could find on this, looking at the exact legal mechanisms, and at what officials, oil refiners and state media actually said. The answer turns out to be far stranger than I expected: no-one in Beijing actually ever ordered an import cut. China, effectively, seems to have saved the world economy incidentally, as a by-product of its own domestic policy to protect Chinese consumers. Which is the real story here: China has become so large and so important it can just rescue the world economy as a mere side effect. In fact what I discovered, hilariously, is that the Chinese were just as surprised as anyone else: it’s only weeks after Western media started crediting China with saving the world economy that Chinese media picked up the story. And even then, CCTV’s first take in July (chinanews.com.cn/gj/2026/07-…) presented it as something 外媒称 (wàiméi chēng, “foreign media claim”) 🤣 Almost like a giant suddenly waking up to its own powers. Because, that's the thing: intentions fade, consequences stay. Intentional or not, as Igor Sechin, the head of Rosneft, Russia’s largest oil company, put it (reuters.com/business/aerospa…), the consequence is that it's now China, not OPEC, who calls the shots on global energy markets. The valve of global oil pricing is now in Shandong, not in Zhongdong anymore (Zhongdong, 中东, is the Middle East in Chinese). Full article here, with how China’s 20-year strategic nightmare, the “Malacca dilemma”, has now effectively become its leverage over the rest of the world: arnaudbertrand.substack.com/…

This has got to be the most moronic story of the month. It's genuinely so stupid it frankly boggles the mind. It was broken by Norwegian daily "Dagens Næringsliv" (DN) under the title "Chinese-owned rental cars filmed secret exercise: Saved video of soldiers and codes" (dn.no/forsvar-og-sikkerhet/k…). BUT when you look into the story what actually happened is that, for an exercise, the Norwegian army was short on cars so it rented some from Hertz. Those rental cars had dashcams - as cars do - that the soldiers mistakenly kept on. At one point, some soldiers found out about that: they were clicking through the car menus out of boredom and stumbled on the footage, including clips of themselves typing gate codes in front of the camera. They deleted the footage. There is zero evidence the footage left the car. THAT'S IT, that's the whole story. China hysteria being what it is in Europe these days, this somehow got transformed into a "China spying" story because Hertz Norway is owned by Volvo, itself owned by Geely (a Chinese company), and 2 of the cars were from a Chinese brand, Deepal. The narrative apparently being that Geely, through Volvo, through Hertz, deliberately left standard dashcams on in rental cars in the hope that the Norwegian military would one day rent them and forget to turn the dashcams off 🤦 And now, as is always the case, it's being spread globally by accounts like Clash Report 👇, which even found a way to make the story dumber: their version has the cars as "Chinese-owned Volvos", when they weren’t Volvos at all. Such is the contemporary media ecosystem: one outlet invents the hysteria, the next one strips away the last remaining nuance and we all end up dumber for it.

This is a really funny Xi video, which I've never seen before. I just added subtitles in English to it 👇 He explains the art of giving gifts to foreign leaders. The South Korea anecdote is particularly hilarious: he explains that when China had to find a gift for President Park Geun-hye, they went through her autobiography, where she wrote that if she’d ever had a first love (she was unmarried), he would be Zhao Zilong, a legendary general from the Three Kingdoms. So they had Fan Zeng, one of China’s most famous painters, paint her crush 😂 Xi can’t resist adding that Zhao Zilong happens to come from Zhengding, the county he ran in the 1980s, which makes him "very proud too." Xi’s actual point is that good diplomacy is personal and that you need to do deep research on the people you’re dealing with: reading their biographies, knowing their taste, their history, etc. It signals respect and, as he puts it, is "a very, very effective way of reaching people’s hearts."

This is important: China just launched the "Yau Laboratory Alliance for Mathematics and Artificial Intelligence", an alliance gathering over 1,000 mathematicians from some of the best institutions in the country, and spearheaded by Fields medalist Shing-Tung Yau. If you read what Shing-Tung Yau said at the inauguration ceremony (for instance here: k.sina.cn/article_1644114654…), it actually busts a common misconception about China: that nobody - at least among elites - ever admits weakness and that the only message allowed in public is that China is strong. In this instance, and oftentimes, it's actually the exact opposite. Shing-Tung Yau, as reported in Chinese media, is saying that the whole rationale behind the alliance is because China is behind in the use of AI for cutting edge mathematical research. As Shanghai’s Wenhui Bao reports (163.com/dy/article/L914MG3S0…), he said Chinese AI models are not yet world-leading and that, facing competition from big foreign companies, the alliance is needed to pool large numbers of top talent and achieve a leap forward. The creation of the alliance seems to be a direct answer to the recent OpenAI announcements: last week they published 722 AI-generated maths manuscripts, some of which were genuinely good. Xi Nanhua (席南华), president of the Chinese Mathematical Society, directly referenced this at the alliance's opening ceremony, saying that OpenAI's "practices have caused great trouble for the whole academic community" (finance.sina.com.cn/jjxw/202…) and that a new system of rules and culture needed to be found to jointly promote the development of mathematics and AI. This is actually one of the key drivers behind China's progress: being ruthless about truth and setting ideology aside. In fact, Shing-Tung Yau pretty much referenced this principle directly during the opening ceremony, saying with regard to the use of AI in mathematics that "black cat, white cat: as long as we get to the truth, the result is what matters most" (a reference, of course, to Deng Xiaoping's famous line). And, funnily enough, you're probably more likely to face backlash in China if you show overconfidence than if you admit weakness. For instance in 2018 leading Tsinghua economist Hu Angang famously claimed China had overtaken the US across the board and faced broad public shaming about it, with Tsinghua alumni petitioning for his dismissal and state media reporting about it (asiatimes.com/2018/08/tsingh…). That's why, if you actually read and listen to Chinese political and economic elites, they almost never boast. Another good example of this is Ren Zhengfei, the founder of Huawei. He recently had an interview on the front page of People's Daily - China's flagship newspaper - in which he said: "The US has exaggerated Huawei's achievements; Huawei isn't that powerful yet. We need to work hard to live up to their assessment. Our single-chip technology is still a generation behind the US." (paper.people.com.cn/rmrb/pc/…) This is deep in the culture, one of the central tenets of Confucianism. For instance the Book of Rites, one of the 5 classics of Confucianism, has the line "only by learning do we know our inadequacy... knowing our inadequacy, we can examine ourselves; knowing our difficulties, we can strengthen ourselves" (ctext.org/liji/xue-ji). In any case, long story short, this is why China has good chances to succeed in this new endeavor to catch up in AI-driven mathematics: knowing its inadequacy, it can strengthen itself.

So I looked into the details of this deal because I was very surprised that China supposedly agreed to "halve HEV, PHEV exports" (hybrid cars) to the EU. And sure enough, there's a catch. What China apparently agreed to is to halve exports of hybrid cars **measured against the EU's projections of future exports.** That's according to Šefčovič himself at his press conference in Beijing: "According to the projection we have, that’s more than halving of the exports of hybrids and plug-in hybrids into the European Union under no change scenario" (euperspectives.eu/2026/10/ch…) In other words, what that almost certainly means in practice is that Chinese exports of hybrid cars to Europe will keep growing, but at a slower pace than the EU projected. The numbers confirm this: last year China exported roughly 500,000 hybrid cars to the EU (yicai.com/news/103000226.htm…). Šefčovič said that his deal to halve hybrid exports corresponded to "several million cars over four years" (eletric-vehicles.com/general…): mathematically, for this to happen, the EU's "no change scenario" must assume something like 4 to 6 million Chinese hybrids over four years - that's 1 to 1.5 million a year, two to three times last year's level. Lastly, small reminder that, unlike the US, the EU largely welcomes Chinese carmakers building factories on its soil, and cars made there aren't covered by this deal at all. That's in fact the EU's stated hope: that limiting exports pushes Chinese brands to produce in Europe. Europe played the same game with Japan in 1991: an export restraint that lasted until 1999 and ended with Toyota and Nissan building factories in Europe. Chinese brands are already on that path, with Chery reviving the old Nissan plant in Barcelona (thenewbarcelonapost.com/en/t…) - a perfect case of history rhyming.

We keep seeing videos of clashes between French highschoolers and the police but almost no-one speaks about why these highschoolers are in the street, as if it all happened in a vacuum. They have several complaints: missing teachers (a huge issue, almost 10% of class hours aren't taught at all according to the French ministry of education - education.gouv.fr/depp/la-pa… - mostly due to missing teachers, it's insane, the equivalent of 2 whole days of class a month gone!), run-down buildings, classrooms above 40°C during summer, classes of 32-35 students, etc. But, to me, the most salient complaint has to do with the new French system for getting into higher education, called "Parcoursup", which borders on the absurd and is a perfect illustration of France's Kafkaesque technocracy. So, it used to be the case - when I was a highschooler in France myself - that getting to university or Grandes Écoles was relatively straightforward. If you passed the bac - France's national end-of-high-school exam - you could enroll at your local uni. Dead simple. And, if you wanted to go to a Grande École, you had to apply to a so-called "prépa" - a notoriously grueling two-year preparatory program - then sit the competitive exam. That wasn't simple at all but at least it was clear. There was, admittedly, one issue with this system: since anyone with the bac could enroll into university, the most popular programs ended up with more applicants than seats, and had to resort to drawing lots. In 2017, for instance, medicine in the Paris region received 8,507 first-choice applications for 7,650 places, and 857 applicants were left out by lottery (letudiant.fr/educpros/actual…). A real problem, but a narrow one, affecting a limited number of programs. In typical French technocratic fashion, the government's answer was to take a sledgehammer to crack a nut. They decided to tear down the entire system - which had largely been established under Napoleon - and rebuild it from scratch, for all 640,000 high schoolers who apply every year. Now, you will undoubtedly be surprised to learn that, among other absurdities in the new system, highschoolers still pass the baccalauréat ("bac") but it's become almost entirely useless. It's very hard to describe the new Parcoursup system without giving you a headache, because it is very complex. And that's a big part of the issue. I'm still going to try though. Here's how it works. In your final year of high school, you list up to 10 choices of program (e.g. "law at Panthéon-Sorbonne University") on the Parcoursup platform, and you're not allowed to rank them. This includes prépas for Grandes Écoles as well. Every program then ranks every single applicant according to their own criteria, which are variable and may change every year. First 5 absurdities: 1) As a student you're institutionally not allowed to have preferences when applying, your 1st choice counts just as much as your 10th choice. And you can't game the system by selecting only 1 choice, because it gains you nothing: programs can't see your other choices anyway. Choosing only one program just removes your safety net. 2) You're playing a game whose rules are only revealed once most of the moves have been played. Say you really want to go into a specific program - again, let's take the example of law at Panthéon-Sorbonne University: this program will only tell you which criteria it uses to rank you, and how much each one weighs, in December of your final year of high school, shortly before you apply on Parcoursup. And by then, most of what you'll be judged on (your grades from the previous year, your choice of subjects, etc.) is already set in stone, so all you can do is look at the previous years' criteria and pray they don't change. 3) The most absurd of all. As we saw above, Parcoursup was meant to end the "injustice" of popular programs having more applicants than seats, and of drawing lots as a result. This exact problem has now been compounded by an order of magnitude: since everyone applies to 10 programs without ranking them, each of these programs gets absolutely flooded with applicants, most of whom will end up elsewhere. For instance, in 2025, law at Panthéon-Sorbonne received 19,404 applications for 745 seats (questions.assemblee-national…). In other words, ironically, the new system took a problem that was an exception - hitting a limited number of students - and made it the rule for all 640,000 who apply each year, all in the name of solving it. Completely Kafkaesque! 4) The new system is - obviously - also unpopular with university teachers, who have to do the ranking. I mean: imagine being a law teacher at Sorbonne and being told that your job now includes having to join a committee that has to review and rank some 19,404 application files every year. At 3-4 minutes per file reported by one university committee (letudiant.fr/educpros/actual…), that's 140-180 working days a year! In practice, this means that committees rely on algorithms to pre-rank applicants and only look closely at some files. Which means students produce application files that nobody really reads: funnily, to prove this, a student last year replaced her cover letter with a brownie recipe and still got in (franceinfo.fr/societe/educat…). 5) And, yes, it means the bac - an institution in France, put in place by Napoleon - is now almost completely useless given that, by the time bac results come out in July, you've already gone through Parcoursup. As a result, whether you have a good or bad grade doesn't change your fate one bit, as long as you pass. Getting a 19/20 gets you exactly the same as an 11/20: Napoleon's bac has been reduced to a pass/fail formality. But that's not all, we haven't yet spoken about the process after you apply and start getting responses from programs, which happens starting early June. This one is a real beauty of majestic French bureaucratic complexity. So programs can give you one of 3 answers: yes, no, or waiting list. Waiting list means not yet, but maybe later if another candidate gives up their seat. As a student, you have 3 rules: Rule 1. Once you get a yes, you have two days to accept it (reduced to one day if the answer comes in mid-August), or you lose it. Rule 2. You can only hold one "yes" at a time Rule 3. You can keep your waiting-list places while holding a yes. If a program you prefer later says yes, you can switch to it, and your previous "yes" seat is released automatically. And, funnily enough, once you've received your first answers, you're suddenly obliged to rank the programs you're still waitlisted on, within a four-day window, a few days before the bac. Any you forget to rank get deleted. Are you lost by now? If you are, you're not alone: in one survey, only 6% of French parents find it transparent, and only 5% fair (letudiant.fr/etudes/parcours…). Quite an achievement for a system whose literal objective was to make access to higher education fairer and more transparent. And it's not as if the problem hadn't been solved elsewhere, in ways France could have learned from. For instance in China, once you have your Gaokao score - China's national university entrance exam, roughly equivalent to the French bac - you list your choices for university in order of preference and places are then filled in one go, best scores first: each student lands in the highest-ranked choice their score allows. Done. Procedurally fair - even brutally so - simple and transparent. But no, we French had to somehow decide that - bizarrely - the bac wasn't going to count anymore but previous grades would, that students wouldn't be allowed to rank their choices (until, suddenly, they're forced to in June), that each program would invent its own criteria and only reveal them once most of your grades are set in stone, and that the whole thing would play out not in one go but over weeks of waiting lists and 48-hour deadlines. All ending in a system that piles an immense amount of work, complexity, anxiety and stress on everyone involved. And it culminates with protests against the administrative Kafkaesqueness of it all, and tear gas for the kids complaining about it. How very, very French.

This is a good opportunity to remind everyone that the UK's National Security Act, which MI5 just invoked here 👇 in relation to a Chinese research institute, goes WAY further than Hong Kong's National Security Law that the UK have spent years criticizing as "dystopian". This case is very illustrative of this: MI5 just has to state with zero proof whatsoever - and they didn't provide any in this case - that an organisation is linked to a foreign intelligence service and, from that moment on, as per section 17 of the UK Act, accepting anything from them, money or even mere information, exposes you to up to 14 years in prison. More dystopian still, according to the act, you can be prosecuted even if you never received a penny or never actually did a single thing for the organization: merely agreeing to accept something from them is enough. And there is no minimum threshold for what counts: anything with "money's worth" qualifies. So in this instance, say that you meet someone from this Chinese research institute at a conference and they offer you a coffee: under the UK's National Security Act merely saying "yes, sure" could, on paper, land you up to 10 years in prison, even if it turns out that the conference center doesn't serve coffee, and even if you had no idea about MI5's warning on this institute (according to the act, you "ought reasonably to know" since MI5's warning exists)! But there's even worse: the act was significantly expanded this summer to also include restrictions on speech. Which means that now, merely expressing an opinion supportive of an organisation the Home Secretary has "designated" can get you up to 14 years in prison, if done for a purpose you "ought reasonably to know" was prejudicial to the UK's safety or interests. So say that Shabana Mahmood, the current Home Secretary, following up on MI5's warning, gets this Chinese institute added to the list of designated bodies. From then on, an academic who tweets to defend this institute - for instance saying that they contributed positively to UK research - could, on paper, be prosecuted for "expressing an opinion or belief that is supportive" of it, with up to 14 years in prison - on the theory that they "ought reasonably to know" that defending them was prejudicial to UK interests. And the worst part is that, once an organisation is designated, the accusation behind it never has to be proven in your trial: the designation alone does the work. The designation itself only requires the Home Secretary to "reasonably believe" the MI5 accusation and Parliament to vote it through, and no court is ever asked to decide whether it's true. Which means that the intelligence behind it - or lack thereof - never has to be proven or even shown to you, the court or the public. This all goes way beyond Hong Kong's National Security Law: for instance we just had the trial of Jimmy Lai where the prosecution had to prove beyond reasonable doubt, through witnesses and years of evidence, that he had lobbied foreign governments to act against China. Under the UK's new regime, the foreign-power link just has to be asserted by the government and parliament, and in court this assertion is all that's required without a shred of evidence about it needed before the court. By the way, I'm not making any of this up, you can read the National Security Act yourself here and see that everything I wrote is rigorously accurate: legislation.gov.uk/ukpga/202…

This is interesting: I stumbled upon a 1974 declassified CIA report (available here: cia.gov/readingroom/docs/CIA…) on "East Asian Contested Islands" and it's pretty fascinating to see how the narrative has evolved since back then. So, for instance, the Scarborough Shoal, which is the central flash point between China and the Philippines, and the disputed territory around which the whole 2016 arbitration case originated, wasn't even claimed by the Philippines back then, only by China. In the CIA's words: "The Philippines have not claimed the shoal despite its proximity to Luzon. China includes Scarborough Shoal within the sweeping claims it makes for the South China Sea." In other words the place that has come in the media to define China's supposed encroachment on the Philippines' territory was, back in 1974, considered by the CIA to be solely a Chinese claim, full stop, with the Philippines nowhere in the picture. In fact, the Philippines only started to publicly claim the Scarborough Shoal as part of their territory over two decades later in 1997 when their navy intercepted a China-sponsored amateur radio expedition on the shoal and later arrested 21 Chinese fishermen, the first ever confrontation between the two countries over it. Philippine congressmen Roque Ablan and Jose Yap planted a Filipino flag on one of its rocks that year (irasec.com/IMG/UserFiles/Fil…). As late as 1990, the Philippines' ambassador to Germany, Bienvenido A. Tan, Jr., wrote in a letter that “according to the Philippine National Mapping and Resource Information Authority, the Scarborough Reef or Huangyan Dao does not fall within the territorial sovereignty of the Philippines." (globalnation.inquirer.net/31…) Which raises an obvious question: if a place was exclusively claimed by one country for decades, and the other country only started publicly claiming it in 1997, who is really encroaching upon whose territory? 🤔

The technology of this Chinese company (Orient Sundar, 奥润顺达) is actually super interesting: they've developed full "passive-house" systems where houses/apartments are so well insulated from heat or cold that they require virtually no heating or cooling, even in harsh climates. They've actually developed an entire residential compound with that technology in Gaobeidian, a city next to Beijing, which gets extremely cold in winter (as low as −23.3°C) and one resident interviewed in Chinese media said that her flat stays at about 23°C all year with "no need to heat the house in winter or turn on the air conditioner in summer" (news.qq.com/rain/a/20231204A…). Funnily, one problem they've had with the technology in China is that they underestimated how much heat was naturally produced inside homes, because most Chinese families live with their elderly parents who cook three meals a day at home: due to this, the flats generated around 40% more internal heat than their models initially assumed (journals.sagepub.com/doi/10.…). In other words, with this technology, the home heating system becomes grandma's cooking 😅

This blew my mind: between 1403 and 1407, the Ming Dynasty emperor commissioned an encyclopedia - now called the Yongle encyclopedia, after the emperor - that was so vast in scale (around 400 million Chinese characters) that the only other encyclopedia that finally exceeded it was Wikipedia in 2007, 6 years after its launch! This is yet another example of West-centric our sense of intellectual history is: in French education we're taught that encyclopedias as a concept started with Diderot and d'Alembert's "Encyclopédie" in the mid-18th century, which we're told was the first attempt in history to gather all of human knowledge in one place. Not only was the Yongle encyclopedia completed 350 years earlier, and not only is it 20 times larger than L'Encyclopédie, but it was itself a very late entry in a Chinese tradition of encyclopedias which started with the Huanglan - more than 8 million characters in length - compiled around 220 AD (en.wikipedia.org/wiki/Huangl…). Unfortunately much of the encyclopedia has been lost: out of the original 11,095 volumes, only around 400 are known to have survived (300-ish of which are in China). But we sometimes find new ones, like two volumes that surfaced in 2020 in the attic of a French family, descendants of a naval officer who had received them as a gift from corrupt Chinese officials in the 1870s. They went to auction in Paris with a starting price of 5,000 euros and sold for 8.12 million to a Chinese textile entrepreneur, Jin Liang, who donated them to China's National Archives of Publications this summer (sixthtone.com/news/1019031).