Replying to @munster_gene
What the hell are you talking about gene?
bsilly, TLA retweeted
“Alice Huang was a TV reporter for Chinese state media, then in 2023 joins Megaspeed, suddenly becomes CEO, becomes one of the largest buyers of Nvidia chips in the world... then the company sells itself to a Singaporean shell company and operates out of all these empty offices.”
Ed Zitron
bsilly, TLA retweeted
„I’ve found conclusive forensic evidence that the top 4 Chinese AI labs—DeepSeek, Moonshot, MiniMax and Ant Group—have built their newest models on NVIDIA’s most advanced Blackwell chips, which are barred from export to China.“
I have a major result to share.
I’ve found conclusive forensic evidence that the top 4 Chinese AI labs—DeepSeek, Moonshot, MiniMax and Ant Group—have built their newest models on NVIDIA’s most advanced Blackwell chips, which are barred from export to China.
This sounds insane, but I discovered that almost every AI model carries a hidden signature of the GPU that trained it. When a model is first set up, the starting values of its weights are laid out in a pattern that depends on the exact GPU. Each combination of GPU and seed number produces a unique pattern—a sort of fingerprint of the hardware used to create the model. Traces of that pattern survive training into the published model.
Again, this evidence isn’t based on any insider information or leaking. It’s based on a forensic analysis of the model weights released by the companies themselves.
I searched more than six trillion combinations of chip layout and seed number and matched them to open-weights models. In doing so, I found conclusive evidence that six Chinese AI models—including DeepSeek-V4-Pro, Kimi K3, and MiniMax-M3—were initialized on NVIDIA Blackwell systems. Four of the models match Grace Blackwell, the GB200 and GB300 parts in NVIDIA’s flagship liquid-cooled rack-scale systems.
I then verified these results on real hardware. Running unmodified PyTorch with the seeds I’d recovered, a Blackwell B300 reproduced DeepSeek-V4-Flash’s untouched starting weights bit for bit, and a rented GB200 reproduced Kimi K3’s to a correlation of 0.999995. A Hopper H100 reproduced none of DeepSeek-V4-Flash’s, and changing the seed by one broke the match. (DeepSeek’s seed is 953049583 by the way.) The method also was verified to match 24 known open-weights models to the hardware that developers said they used.
It is exceedingly unlikely that these results are a coincidence. Across six trillion combinations, the best accidental match should reach only about 8 sigma (i.e., eight standard deviations above what random chance would produce). The real matches range from 49 to 1,385 sigma, and even the weakest single test hits 20.
These were data-center-scale training runs. With some analysis of release schedules and model details, I estimate that the largest, Kimi K3 and DeepSeek-V4-Pro, used 2,000 to 3,000 Blackwell GPUs each. Clusters of this size cost on the order of $150 to $200 million each to buy, and across these labs alone would add up to more than half a billion dollars of export-controlled GPUs, not counting experimental compute or otherwise. It’s quite likely that we see multi-billion dollar expenditure on export-controlled GPUs.
I am highly confident in this result, and it has been independently verified by senior researchers at OpenAI and Anthropic. You can read the full report at unemploymentlabs.com/checkpo….
There is much more I’ve identified about these models through other forms of checkpoint forensics like this, and will have more to share soon.
This prosecution of Houellebecq provokes concerns that go far beyond freedom of expression.
In just over thirty years as a novelist writing mainly about the misery of bourgeois French life, he has lifted up a number of stones and allowed some pretty unsavoury things to crawl out from under them.
In no particular order these include sex tourism, paedophilia, mental illness, rampant narcissism, family breakdown, betrayal, social alienation, self-hatred, nihilism, prostitution and, inevitably, racism. But then anyone who understands what French society is really like will find none of this surprising.
Read more from Simon Heffer:
buff.ly/TW0kHpf
bsilly, TLA retweeted
51 years ago tonight, the Arthur M Anderson was the last ship to see the Edmund Fitzgerald, and the first one to go back out looking for survivors. She now sits quietly at a dock in Toledo.
bsilly, TLA retweeted
Key post from @scoopercooper (Aug 24, 2026):
“I’d be worried if I was a Canadian bank or developer etc with exposure to Iranian currency exchanges and Chinese underground banking, that is major exposure risk given Bessent’s announcement. Speaking of economic warfare. People don’t know the half of the tools available to DC.” @scoopercooper
This was posted the same day Treasury Secretary Scott Bessent launched Operation Economic Outcast, an aggressive campaign to sever Iran’s financial lifelines (including digital assets, secondary sanctions on facilitators, and targeting networks enabling IRGC funding). home.treasury.gov
thebureau.news/p/us-treasury… U.S. Treasury Says Ryan Wedding Laundered Billions in Drug Money Through Canadian Currency Exchanges, Sanctioning Eight Toronto and Vancouver Firms Accused of Moving Money for IRGC
bsilly, TLA retweeted
The US Treasury said on Friday that it had imposed financial sanctions on eight money services firms based in Canada that it alleges have facilitated transactions with sanctioned Iranian entities.
reuters.com/world/middle-eas…
bsilly, TLA retweeted
this is the level of petty that I aspire to
dağ keçisinin son hareketi efsaneydi yalnız
Readers added context they thought people might want to know
Videodaki görüntüler gerçek bir yaban hayatı kaydı olmayıp yapay zekâ ile üretilmiştir. Fizik kurallarına aykırı hareketler ve görsel bozulmalar barındıran bu tür dağ keçisi videolarının yapay zekâ ürünü olduğu bilinmektedir: malumatfurus.org/keci/
bsilly, TLA retweeted
Cattle Call - by Tex Owens
I have been singing this song for many years, but I never grow tired of it! 🤠
#classiccountry #western #tengallonhat #yodel #cowboy
bsilly, TLA retweeted
Recently I was talking with a guy from the bourgeoisie. Good schools, good job, good manners.
I asked him: "Do you think there are too many people on Earth?"
He said yes. Without a second of hesitation.
I asked him why.
He gave me two reasons. First: there aren't enough resources. Second, and I'm quoting him word for word: "I'm afraid of the masses."
That second sentence chilled me. Because it admits what the first one disguises.
This guy didn't come up with that opinion on his own. Nobody is born a Malthusian. You become one. And this belief has a history, authors, institutions, and funders.
It starts with Thomas Malthus, an English clergyman who argued at the end of the 18th century that population would always grow faster than food, and that famine would take care of the surplus. There were 1 billion of us. Today there are more than 8 billion, and humanity has never eaten better in its entire history.
In 1968, Paul Ehrlich publishes The Population Bomb. He predicts that hundreds of millions of people will starve to death in the 1970s. It doesn't happen. Norman Borlaug and the Green Revolution feed the world. Ehrlich never apologizes, and gets invited on every show for the next fifty years.
In 1972, the Club of Rome, founded by an Italian industrialist and a Scottish scientist, commissions a report from MIT: The Limits to Growth. Computer models, collapse scenarios, resource depletion. The book becomes the bible of an entire generation. Degrowth is its direct descendant.
In 1980, economist Julian Simon makes a bet with Ehrlich: the price of five metals, chosen by Ehrlich himself, will fall within ten years. By 1990, all five had fallen. Simon wins. Why? Because a human being is not a mouth to feed. A human being is a brain that solves problems. The ultimate resource is us.
Fifty years later, the story hasn't changed. It just changed outfits. It's no longer "there won't be enough wheat," it's "your child has a carbon footprint." Studies get published explaining that not having children is the most ecological thing you can do. People discuss it at Davos, between two private jets.
Bill Gates said it himself on the TED stage in 2010. He presents an equation to bring CO2 emissions down to zero, and the first term of that equation is P. People. He then explains that with vaccines, healthcare, and reproductive health, we could lower population projections by 10 to 15%. Nobody is talking about killing anyone. It's more subtle, and in my view more disturbing: a large part of the elite sincerely sees human beings as a variable to be reduced in an equation.
And what happens when a technocratic elite actually takes this dogma all the way? You get China. In the late 1970s, Song Jian, a missile engineer for the regime, discovers the Club of Rome's work, runs his own projections, and convinces the Communist Party. The result: the one-child policy. Hundreds of millions of abortions and sterilizations, many of them forced. Today China sits around 1 child per woman and is set to lose a massive share of its population by the end of the century.
Meanwhile, look at the numbers. South Korea: 0.8 children per woman. Taiwan: 0.7. Spain: 1.1. Japan: 1.14. And France, long the champion of Europe, has dropped to 1.56. In 2025, for the first time since World War II, my country recorded more deaths than births.
Elon Musk has been saying it for years: population collapse is a far bigger danger to civilization than climate change. If people don't have more children, civilization will crumble. Japan will eventually cease to exist if nothing changes. And just yesterday: so many cultures are dying out. People call him alarmist. The curves prove him right.
Now do a simple exercise.
Take all 8 billion humans and pack them into a cube. It would be less than a kilometer on each side. All of humanity, in a volume smaller than a single neighborhood.
Even better: if all of humanity lived at the density of Paris, it would fit in an area smaller than France. The rest of the planet would be empty.
And the next time you fly, look out the window. Paris to San Francisco is hours over Greenland, the Canadian Arctic, deserts, forests, mountains. Emptiness. Thousands of miles without a single light. Cities cover roughly 1% of the Earth's land.
We are not too many. We are tiny.
And above all, we are far too few for what lies ahead.
If we want to leave this cradle, build cities on Mars, mine asteroids, build orbital habitats, and then head for the stars, we won't need 8 billion humans. We'll need a trillion. Just to get started. Every human is a genetic and cultural lottery ticket. The more humans there are, the more Newtons, Pasteurs, and Von Neumanns. The more engineers, builders, artists, and explorers. Genius is a law of large numbers.
So let me be clear.
Everything you think you know about overpopulation is dogma. A dogma born with an English clergyman, turned into equations by the Club of Rome, enforced with a truncheon by the Chinese Communist Party, and recycled today by the degrowth left and the salons of Davos.
It's not science. It's a worldview in which humans are a problem rather than a solution. And behind the resources argument, there is almost always what the guy I was talking to was honest enough to say out loud.
Fear of the masses.
The problem was never that there are too many of us. The problem is that soon there will be too few.
bsilly, TLA retweeted
Around 2016, Jim Belosic bought a $750,000 laser cutter on a whim.
This month, his company SendCutSend raised $140 million and hit a valuation of $2.1 billion. It's on track to make 50 million metal parts this year for more than 300,000 customers.
This is the story of how SendCutSend arrived and ended up reviving US industry. corememory.com/p/sendcutsend…

































