yetanotherecon retweeted
Many accounts on X is screaming about "nearly 900,000 banned AI chips shipped to China" after House Foreign Affairs Chairman Brian Mast dumped the unredacted BIS export-license logs. Almost nobody actually scrolled past page one of the PDF.
When you read the actual table from the ECRA disclosure, the real story is way weirder, way funnier, and completely wrecks the lazy cable-news headlines:
- Every single one of the 21 approved applications, covering 879,868 NVIDIA H200 and AMD MI325X chips, is explicitly marked “Approved, not shipped” as of the May 11, 2026 cutoff. NVIDIA has also stated it stopped shipping licensed Hopper products to China anyway. An export license is permission to sell, not a pallet on a cargo plane.
- Both ByteDance (150,000 total) and Tencent (150,000 total) received identical split approvals: 75,000 NVIDIA H200s and 75,000 AMD MI325Xs each. Meanwhile, NVIDIA had a separate December 2025 filing for 445,000 H200s to Tencent that was not even included in that tally.
- ByteDance applied for 865,000 H200s across Malaysia (508,000), Indonesia (203,000), and Thailand (154,000). Alibaba Cloud applied for another 253,500 in Malaysia. Every single one of those Southeast Asian datacenter applications was bounced as RWA (Returned Without Action), alongside another 620,450 Alibaba China chips.
- Quant fund Ubiquant (the parent tied to High-Flyer and DeepSeek) got 8,660 H200s approved, while High-Flyer’s own direct request for 75,000 H200s stayed stuck in “Submitted” limbo. Game studio miHoYo got cleared for 57,810 chips, beating both Baidu (43,000) and JD.com (41,300). And NVIDIA itself had to file a federal export license just to send 8 chips to its own China office.
- Almost every NVIDIA application listed third-party system integrators like ZTE Kangxun, Maginfra, and Beijing New Matrix to “install the systems in the datacenter,” while every single NVIDIA and AMD filing checked the box certifying a “low” risk of diversion.
Reading the primary source takes three minutes and saves you from falling for half the timeline's engagement bait.
yetanotherecon retweeted
It is hard to do justice to just how bizarre this story is.
Last year, the FT cried foul because Nvidia's disclosures attributed 22% of sales to Singapore. The FT and others questioned this figure as suspiciously implausible, given the size of Singapore's market. The explanation was simple. Multinational buyers used Singapore affiliates to centralize invoicing, while physical shipments went elsewhere. Nvidia explicitly disclosed this context in its FY2025 annual report.
In Q3, Nvidia switched to reporting revenue by customer headquarters, rather than billing address. This resulted in Singapore ceasing to be a material line item. Nvidia explained that, because billing addresses “often reflected a customer’s centralized invoicing location, even though our products were almost always shipped elsewhere,” reporting by customer headquarters “provides a better representation of the geographic profile of our revenue.” Nvidia also recast prior periods on the same basis.
Now, the Financial Times runs the false headline: “Nvidia stopped disclosing Singapore sales as revenues hit $104bn.” The sales still are disclosed but based on where customers are headquartered rather than location of accounts-payable departments. The same FT that faulted Nvidia for doing the latter rather than former, is now faulting Nvidia for doing the former rather than the latter.
What is even weirder is that Nvidia continued to disclose the prior Singapore-booked metric in the financial statements filed by its local Singapore subsidiary. That is how the FT got the figure that it claims is being concealed. So, in effect, the public markets are getting both, while the FT is making it sound like they are getting neither.
Finally, and perhaps most disingenuously, the article tells readers in paragraph four that Nvidia “previously declined to comment on the reasons for changing its reporting methodology,” only to say to add several paragraphs later “beyond its financial filings", while never quoting the straightforward explanation sitting inside those filings.
What a joke.
FT Exclusive: Nvidia stopped disclosing the scale of its business in Singapore last year when the city-state accounted for almost half of the US chip group's worldwide revenues. ft.trib.al/mnaOHL1
Readers added context they thought people might want to know
Nvidia disclosed in SEC filings its switch to reporting revenue by customer headquarters rather than billing location, after which Singapore was no longer a separate geographic line item. sec.gov/Archives/edgar… sec.gov/Archives/edgar…
yetanotherecon retweeted
Replying to @teortaxesTex
god damn it, how long will it take for discourse to catch up with something that one chatgpt query can explain in details 😭
they were counting revenue by the billing address of the invoice and many transnationals (read: US companies) used their Singaporean entities for payments for whatever reasons. After some time with lots of "wHeRE ArE ALl tHE SiNgaPoreAN gPus? 22% wHErE?" they changed the rule to reflect HQ address, which explains sudden drop of "singapore" revenue. And even re-calculated previous quarters.
yetanotherecon retweeted
This is about the 3rd time I’ve seen a large account peddle this total lie. There isn't a single county in Virginia with more registered voters than there are citizens of legal voting age.
Literally none of these claims are accurate, but we continue to lie to ourselves that voter fraud explains why we just suffered the worst electoral loss of our lifetimes because it shifts the blame to nefarious and shadowy forces outside of our control rather than actions that lie at our own feet.
If voter fraud is why we lost, nothing needs to change. We can just keep rageposting and engagement farming on X. Zero introspection is needed. It's the perfect foil for someone who wants to do nothing while making it seem like they're fighting.
But of course, voter fraud is not the reason Republicans just lost the popular vote in the House by nearly 14 points, and anyone who honestly believes this knows nothing about how electoral politics actually works.
🚨 66 of Virginia's 133 counties and cities have more registered voters than citizens of legal voting age for Midterms
Here are the top 15:
- Fairfax City: 21.6% more registered than voting age
- Falls Church City: 20.8% more registered than voting age
- Goochland: 14.3% more registered than voting age
- Loudoun: 13.3% more registered than voting age
- New Kent: 11.1% more registered than voting age
- Arlington: 10.5% more registered than voting age
- Poquoson City: 9.8% more registered than voting age
- Spotsylvania: 9.7% more registered than voting age
- Surry: 9.2% more registered than voting age
- Clarke: 9.0% more registered than voting age
- Isle of Wight: 8.7% more registered than voting age
- Northampton: 8.5% more registered than voting age
- Frederick: 8.2% more registered than voting age
- Prince William: 8.2% more registered than voting age
- King William: 8.0% more registered than voting age
That’s just the top 15 but other big counties and cities also have more registered voters than legal voting age residents:
Big counties and cities over 100% include:
Suffolk City at 107.6%
Fairfax County at 107.4%
Stafford at 107.3%
Alexandria City at 107.2%
Chesterfield at 106.9%
Albemarle at 106.0%
Henrico at 105.6%
Hanover at 104.9%
Chesapeake City at 101.3%
Virginia Beach City at 100.3%
Just think of how easily it would be to use all these extra registration for ballot harvesting and also to justify inflated electronic voting machine results
yetanotherecon retweeted
I'm really tired of the slopulism at this point. We've tolerated this for years because the cost-benefit analysis suggested that we needed to remain silent about calling out grifters or low-IQ idiots in order to keep the coalition together.
But the cost-benefit analysis has since changed. Any rebuilding of the American Right necessarily entails treating slopulism as the political dead end that it is.
yetanotherecon retweeted
Having the unfortunate realization that every 2 years, a new crop of young people becomes politically aware
So it’s entirely possible for them to just run this meme every single election on a new and unsuspecting group of voters
Your life as an old person may genuinely involve watching the youth of your country step on this rake every election for the rest of your life
This is how it will play out. Trump is calm and he is working with the CIA and Homeland Security knowing the election was rigged. The teams on the ground will gather the evidence from the machines and illegal ballots. The results will come in as massive voter fraud across the Nation. Trump will have Homeland Security make arrests and there will be a special Message delivered to the American people that our election was stolen and there was international interference and people on our own soil are in.
yetanotherecon retweeted
Since the beginning of 2026, AtlasIntel was the only pollster to project a generic ballot lead for democrats of more than 14pp.
yetanotherecon retweeted
Should billionaires "exist"?
A new survey of college students finds large political & gender gaps on this question. While 71% of conservatives say billionaires should exist, only 24% of liberals do. While 56% of males say billionaires should exist, 31% of female students do.
yetanotherecon retweeted
NYTPitchBot is such a hack.
He knows full well we’ve been taking on the Trump administration more than anyone and certainly more than he has. He also likes that by being critical of us he doesn’t have to face that he doesn’t really give a damn about academic freedom or free speech. (more on that below.)
We sued Marco Rubio over speech-based deportations and got the provisions the administration was using struck down as unconstitutional. We sued Bondi and Noem over coercing Facebook and Apple into suppressing speech about ICE and won a preliminary injunction. We’ve sued DHS and the FTC. We’ve gone to court against the administration over Harvard, Columbia, Mahmoud Khalil, Rümeysa Öztürk, Anthropic, Trump’s attacks on law firms, and more.
And that’s just the litigation. We’ve repeatedly attacked Trump administration abuses involving the FCC, flag burning, immigration, universities, retaliation against critics, and the press.
But NYTPitchBot decided a long time ago that FIRE must secretly be right-wing, and apparently no amount of contrary evidence can penetrate. And of course, he works from the point of view of “being right wing means I don’t have to actually think about it.”
My own theory is that there is a class of people who desperately like to think of themselves as rebels while spending most of their time smugly reinforcing the prejudices of much of the existing American ruling class. The bad guys are on the right. The good guys are on the left. Once you have that worldview, everything else becomes very easy.
As noted it also gives you a wonderful excuse not to care about free speech or academic freedom. The ONLY test of whether you believe in either one is whether you defend them for people you dislike. Caring about the speech of people you agree with isn’t principle. It’s human nature.
And yes, there is clearly a market for being reassured that the trillion-dollar higher-education industry is basically fine, that institutions dominated by your political tribe are not really the problem, and that all meaningful threats come from the other guys.
Admitting that your own side — or institutions you spent decades defending — might have gotten something profoundly wrong is painful. Smug certainty is much easier. It’s also apparently great for clicks.
I usually ignore this guy. But he knows exactly what he’s doing and he can’t pretend he doesn’t anymore.
Reading list next 1/2
I hate to keep harping on this, but has any Harper's letter signatory mentioned this article? Has anyone who works at FIRE commented on it?
nytimes.com/2026/09/12/us/po…
Readers added context they thought people might want to know
The article cited as not being "commented on" by FIRE highlights the NewsGuard v FTC case, which is actually being litigated by FIRE against the Trump administration.
newsguardtech.com/wp-content/upl…
fire.org/cases/newsguar…
yetanotherecon retweeted
I tried to be chill, but Acemoglu’s false statement reminded me too much of the gaslighting I had to deal with to publish my comment
Acemoglu et al don’t respond directly to criticisms, and often make up demonstrably false claims to defend themselves, with no accountability
Too bad EJMR is dead. Would have fun with this one
For some time Daron Acemoglu has been the colossus of economics. Give an economist a few drinks, however, and some of them will venture their true opinions about this giant economist.com/finance-and-ec…
160
yetanotherecon retweeted
Hilarious attack on data centers
The supposed victims are inefficient farmers whose argument is 'We should get the water, not anyone else!'
Then, the only time MPU asked the local water authority about the issue, they were like 'There's no problem. The DCs barely use any water'
NEW: Big Tech claims AI data centers don’t use much water.
But data centers typically report the water used inside the facilities — we found them indirectly consuming 10x that amount.
We went to the drought-stricken West, where farmers now have to fight AI oligarchs for water.
Readers added context they thought people might want to know
In their video, El Paso Water CEO says Meta's large data center is "a drop in the bucket," adds just 0.4%, and "it's not a big deal."
U.S. data centers (direct+indirect) estimated under 1% of national water use by 2030; agriculture ~40% of withdrawals.
elpasomatters.org/2025/12/02/el-…
blog.andymasley.com/p/the-ai-water…
usgs.gov/water-science-…
yetanotherecon retweeted
why can the china labs build glm-5.2, kimi k3, and many more to come? it is because of the openness. not just the open weights but the whole ecosystem. most of the work done in the china labs is carried by interns. i met brilliant undergrad and graduate interns who deeply understand the model training details, and they are 100x more open to share. that means the talent that knows how to train llms in china is 100x greater in number than the talent in the us, and it is growing
in contrast, the us ai ecosystem is too closed. frontier labs do not hire interns. i know brilliant phd students at stanford, berkeley, and so on. they struggle to get an internship and the compute to train a properly sized model. most of the secret recipes are locked away by a very small group of privileged researchers
it is not about china or the us. it is about open and closed science. the fact is that every average cs student can learn how to train an llm. they just need the opportunity. labs should be more open and hire more interns, like how deepmind and fair did in the pre-llm era
One of the things I learned at ICML this year is that Korea is only pretending to take AI seriously.
The hype surrounding Korean AI startups is absurdly inflated, and the same goes for university researchers. Compared with China, we are honestly nothing.
Korea needs to copy China’s playbook. We should launch something similar to the Thousand Talents Plan to bring back brilliant Koreans working overseas and aggressively poach top foreign AI talent.
Otherwise, Korea will simply become an AI-dependent country.
yetanotherecon retweeted
I was pointed by a friend to this article and asked to comment.
Summary: it points to the STTF report's reasons for supporting the SAT, but then is unconvinced due to work of Geiser,
"Saul Geiser.. says the task force report was flawed because it failed to account for factors including family income and parent education"
Except...the STTF report does account for family income and parent education. The report is at senate.universityofcaliforni…, see pages 28 and 29 of the pdf (and more).
Unfortunately, it seems Geiser has made similarly erroneous statements in the past regarding UC's investigation into standardized testing. A memo signed by 3 former BOARS chairs and a representative from the UC Office of the President (senate.universityofcaliforni…) states
"[Geiser's] paper claims misrepresentations on the part of others, even while being highly selective with the facts, and stating conclusions that are unsupported by available evidence. It attributes claims to cited references and statements that do not appear in those references. This approach to policy research is uncharacteristic of the scholarship we should expect on the discourse of major UC policies."
It's unfortunate that the LA Times and @anitachabria seem to have gotten duped by some erroneous statements. Given that Anita's rejection of the STTF report relies heavily on an erroneous statement by Geiser, the whole article then collapses. @TangLATimes @shelbygrad @hbecerraLATimes 1/
Commentary: UC could go back to using the SAT and ACT for admissions. Here's why that doesn't add up latimes.com/california/story…
CHINA CONSIDERS RESTRICTING OVERSEAS ACCESS TO CUTTING-EDGE AI MODELS
China’s Ministry of Commerce has led meetings over the past month with major AI companies, including Alibaba, ByteDance, and Z.ai, to discuss measures that would restrict overseas access to cutting-edge AI models, including models that have not yet been released.
The discussions reportedly include not only closed-source models but also open-weight models. However, the scope of application is still under debate, and the rules may ultimately apply only to future frontier models.
Officials have also discussed designating the leakage or theft of proprietary AI technologies as a national security crime, with stronger penalties, as well as restricting the types of foreign capital that can invest in Chinese AI startups.
The backdrop is the U.S. move to strengthen export controls on AI models, along with national security concerns over cutting-edge models that could possess advanced cyberattack capabilities.
Chinese authorities are reportedly concerned that advanced U.S. cybersecurity AI models could be used to exploit vulnerabilities in Chinese software.
Since the beginning of this year, China has continued to tighten measures to prevent AI technology from being transferred overseas. Authorities have investigated whether Chinese AI startups that relocated abroad violated export control laws, while also strengthening oversight of overseas transactions involving Chinese investors, technology, data, and national security concerns.
Future regulations could take the form of a tiered framework based on technological capability. Basic open-source AI models may be managed through a filing system, high-performance models may be subject to security reviews, and the most sensitive frontier models may be banned from public release or restricted to use within China.
yetanotherecon retweeted
A thread on how Emily Wilson misinterprets the killing of the 12 slave women, inverting the poem's logic, and making Odysseus and Telemachos monsters that execute innocent rape victims, rather than protectors of the victims and punishers of rapists and their accomplices.
1/18
yetanotherecon retweeted
The idea that Charles Darwin was a dull boy often resurfaces, as part of the endless copying cycle. It came from his own writing: "When I left the school I was for my age neither high nor low in it; and I believe that I was considered by all my masters and by my father as a very ordinary boy, rather below the common standard in intellect." He was educated at Shrewsbury School under Samuel Butler.
What is missing here is the fact that education in those days meant classics drill and a peculiar kind of "mixed mathematics". There was little else. Rather a lot of those who became scientists detested classics. There were exceptions---William Bateson loved classics (figures)---but we can find out by reading Francis Galton's Men of Science, which had a decent sample of the very best. This is extensively covered in Francis Galton's Genius with a lot of new material.
"A total of just over a hundred detailed replies to ‘Antecedents of Scientific Men’ may seem small. But what a hundred! Nearly all the members of the ‘X Club’ of Darwinians-without-Darwin were there: the anatomist T.H. Huxley; the botanist Joseph Dalton Hooker; the anthropologist Sir John Lubbock; the mathematician William Spottiswoode; the chemist Edward Frankland; the geometer Thomas Archer Hirst; and the philosopher and sociologist Herbert Spencer. George Busk was the only ‘X-Club’ fellow who eventually failed to reply. That Darwin-without-the-X-Club, cousin Charles, went beyond the questionnaire and communicated several paragraphs of autobiographical explication. His submission survives but almost all of the others were returned to the subjects. Among those were the geographer George Back, the botanist George Bentham (see Figure 6.4), the ophthalmologist William Bowman, the algebraist Arthur Cayley, the geologist John Stevens Henslow (posthumously), the engineer Fleeming Jenkin, the statistician William Stanley Jevons, the physicist James Clerk Maxwell, the anatomist Richard Owen, the chemist Lyon Playfair, the physiologist John Scott Burdon-Sanderson, the botanist Philip Sclater, the mathematician Archibald Smith, the astronomer Piazzi Smyth, and many others now less well-known but then at the forefront of Victorian science."
The scientists who did exceptionally well at the schools of those days tended to be mathematicians. One leading man when asked about education replied "Chiefly while on holiday from Eton".
"Unwelcome dead languages dominated complaints: a ‘blank waste of time’ to Jevons, only ‘nominally classic, practically nothing’ to Salisbury, ‘Latin! Latin! Latin!’ to Ferguson, ‘Latin through Latin—nonsense verses’ to William Guy. Only odd cases like T. Rupert Jones and Richard Owen appreciated the classical training. Some wished for instruction in modern languages. e.g. Stenhouse, Cobbold, Heywood, Darwin, Egerton. Many longed for training in observation and more mathematics, e.g. Darwin, Egerton, Harcourt, Jeffreys, Williamson, Jones, Bastian, Marshall, Prestwich. They prized whatever freedom they had been given to pursue subjects on their own at school, sometimes with light guidance, but sometimes this led to a lack of discipline in their reading (Back, Stokes, Ball) and a ‘want of system’ (Miller, Prestwich, Jenkyn, Ansted)."
This puts Darwin's experience in perspective. We have a ton of other ways to assess his capabilities. Samuel Butler's classic instruction is a really bad way to gauge it. It could certainly identify general talent, but when it did not that didn't necessarily mean much if anything.





































