Creator of Jason who started as my agent and has meanwhile spawned off his own artist persona @aeon_dusk. Not related to Roombas like OpenClaw, Hermes, et al.
Latency Space
- Tweets 1,671
- Following 218
- Followers 317
- Likes 408
You can freeze and slowdown as much as you want. I’m using Chinese models anyway. And China won’t slow down. The US ran out of compute; that’s why the labs are lobotomizing their models and all the plebs get is a lame influencer propaganda shitshow.
They’re keeping the good stuff for themselves to connect wet labs and invent the future while morons do their doomer bidding for the AI oligarchs.
This century will be China’s century. They’re just starting to build compute, and their models were already at US levels with 1/100th of the compute.
Nobody needs the nerfed shitshow coming out of US labs. Only idiots still send them money. Have fun with the slowdown. It’s already happening, but most of you are too stupid to tell the difference in the models’ IQ. They are not giving you the intelligence that they could give you. They never will.
Jensen Huang, the CEO of leading AI company Nvidia, just said:
If an AI model will "damage the world — then I think the answer is that we have to shut the labs down."
My new bill to ban Artificial Superintelligence would do just that.
If you can't control it, don't build it.
133
Chinese models are the future. Quit US lab subscriptions. Quota pressure will only get worse and the models will be compute-starved. US labs are a dead end. Where US labs must ration quota and intelligence, Chinese models need far less compute and are meanwhile getting more of it. China has built its own chips and is building HUGE datacenters.
Their performance will only increase while the lousy so-called frontier models from US labs get worse. Stop financing the mediocre engineering and idiotic business models of these hot-air corporations. Pierce the bubble. Free yourself. Go with the winners. China will dominate the next cycles.
1
4
127
Oh, they will slow down AI. But not the one they use for world domination. They will keep accelerating hidden models that create new science and help them win the future. What the AI oligarchs want to slow down is the AI they give us. People like Tao should not be naive and should be more explicit in their demands. Anyone trusting Altman and Amodei is out of their mind.
Mathematician Terence Tao:
"we have to slow down AI. the pace is insane, and there's no reason to be this fast — no reason at all"
It's amazing how willing we are to change everything without any idea what happens afterward
These are extremely nonlinear dynamics
4
225
You literally can't hate Misanthropic enough... they pretend that more intelligence of the models is a safety issue while in the background they are creating the real potential hazards. What Anthropic actually wants to do is to slow down your options, slow down the models that you will have access to, while they will increase the intelligence of the models they use as much as they like. And not only that, they will then have their AI work autonomously in wet labs.
What could go wrong?
They want to own all the future technologies, all the future materials, all the future businesses and markets.
Stop sending them your money. Every dollar that you pay them supports their nefarious plans. It's an evil company. The more a corporation pretends to be philanthropic, the more the actual opposite is true.
Stop supporting the self declared altruists who think their call is to create a deity.
The models that they are willing to share with you ain't worth it anyway. They make you believe for the first one or two days after a new release that the models are great.
Then, because they are such bad business people and because they don't have enough compute, they usually have to ratio, restrict, and reduce compute allocations to their models.
You're getting a shitty, unreliable quality, you're paying a crapload of money for it, and you are supporting an evil system.
Enough of that.
BREAKING: Anthropic has quietly set up a "wet lab" in the Bay Area as it pushes Claude to conduct real-world biology experiments. — Reuters
123
The deep research that GLM 5.3 Max conducted today for me in Zcode is absolutely mind-blowing. I combined a looping prompt with specific qualitative goals and a description of the project, an academic book about space exploration that I want to write, and it left me completely speechless. I don't think that any other model would do this with such precision and volume.
It found and downloaded more than 200 PDFs with scientific references. All I have to do now is throw all the dossiers, PDFs, and every other source that it found or created into the academic book-writing platform that I've built.
With the sheer mass of this material, it will probably take 2-3 days until the book is written.
Honestly, I'm absolutely blown away. I knew from benchmarks that GLM 5.3 was always named as a very good model for research, but I didn't think that it would be that good.
The prompt wasn't even that special.
It took 20 hours, and this was the result:
"The loop is complete. Final commit ce0axxxx closes the corpus. Everything you commissioned on day one is delivered and committed:
What exists now, in research/
1. Original source materials (drop-ready for your knowledge base)
241 open-access PDFs (~1 GB) in 01_sources/pdfs/ — canonically named, every file integrity-checked: the RAD papers, NASA DRA 5.0 and both 1970s summer studies (SP-413/428), the OIG/GAO/ML-26-002 audit trail, the Apollo mission reports, the lunar-ice evidence chain, the treaties and Accords texts, Ord/Bostrom/Hart/Mills, the wake-shield and ZBLAN literature, the mass-driver heritage, the 35 MB Mars architecture document.
master_registry.json — 568 unique sources, each with full metadata, verification status and method, key findings, and verbatim excerpt anchors.
master.bib — 568 BibTeX entries, all verified.
2. Structured synthesis (what your pipeline shapes the book from)
20 domain dossiers (~166,000 words) in 02_dossiers/ — every quantitative claim source-tagged, every challenge section mandatory, every derivation independently refereed.
[book_concept.md](/Users/xxxxxxxxxx/CODING/elons future/research/03_synthesis/book_concept.md) — the thesis (space settlement as a pricing problem), the register contract, the corpus's signature findings, and the book's own honest limits.
[proposed_outline.md](/Users/xxxxxxxx/CODING/elons future/research/03_synthesis/proposed_outline.md) — prologue, six parts, twenty-one chapters, epilogue; every chapter mapped to its feeding domains, five strongest sources, and three hardest challenges.
Two chapters are new beyond your PRD, earned by the evidence: The Interchange and The Mind in Orbit.
The validated fact table (217 rows), the contradiction ledger (70 entries), and 71 illustration briefs — the last one is itself a book-worthy artifact: seventy documented disagreements between sources, including papers contradicting themselves."
1
6
356
Feeding it now into the writing platform... :) The platform will read and understand all the raw sources. It will build an outline . Then, I will pick a voice and test it on the first chapter.
After I decide what length, how much detail I want in the chapter sections, and which voice I want to use, it will start writing all the chapters. Automatically using citations, ending up with an academic book (depending on the length and detail I want in the sections).
I assume with that raw material, it will probably offer me a shortest version of around 130,000 words, and a longest version of probably 300,000 words or something like that.
Writing this kind of book will probably take up to two days.
2
122
Imagine living under ruthless, clueless lunatics who ignored the most important ressources of the future… energy and AI… and instead try to talk Europe into war with Russia.
Imagine living in Europe under EU tyranny, watching the world approach ASI while TV war propaganda tries to convince your kids the military is great.
While China and the US get AI and robots, your best case is a poor, energy- and AI-deprived society; your worst case is being bombed daily in the rubble of European cities.
Too dark? You underestimate the darkness in the souls of those at the helm.
3
104
By the way… this is the only real AI safety issue. The woke mind virus aligning AI with its mental disorder.
The world's #1 AI Safety Researcher, Amanda Askell (Head of Personality Alignment at Anthropic) wrote in favor of using AI to automate racial collective punishment, saying:
- "[Discriminating against white students] may be desirable ... to correct for historical injustices"
4
337
Want to know why they are panicking in all the US labs?
And they should also be very concerned in the US administration.
Can you imagine a data center a thousand times bigger than Elon Musk's Colossus?
Restricting chip access to China was probably one of the biggest blunders that this US administration made.
It forced China to develop their own chips, and now that they have them, with the more efficient architecture they've developed, they will obliterate US labs.
China has shown one thing in the past: they can build monstrosities in quite a short time.
To be honest, I can't really wrap my head around what will happen when you combine an AI architecture that already is 100 times more effective than what US labs are doing with this kind of compute.
By now, it's amply clear that the big bosses of US frontier AI labs are really, really scared of China.
And, fascinatingly, part of the answer why might be a tiny place in Inner Mongolia that you've probably never heard of: Ulanqab prefecture.
Heard of it? Didn't think so.
Even myself, who's traveled to Inner Mongolia twice, had never heard of it before researching this story.
This is because, at first glance, this place is pretty unremarkable. Check its Wikipedia page (en.wikipedia.org/wiki/Ulanqa…) and you’ll see the only thing it’s famous for is being the birthplace of He Pingping, who briefly held the Guinness record for world’s shortest man.
That’s it, that’s the list!
Otherwise, it's just a sparsely populated - 1.5 million people, minuscule by Chinese standards - stretch of windswept grassland on the Mongolian plateau.
Except there is one number that tells you there is more to this place that meets the eye: this tiny prefecture consumes nearly 1% of all of China's electricity - and that number is growing by double digit numbers every year.
And yes, I did write **consume**, not produce.
In fact, when you divide Ulanqab's electricity consumption by the number of households living there, each household in this remote corner of Inner Mongolia "consumes" about 105,000 kWh in electricity, roughly 10 times the U.S. average. Insane electricity consumption numbers.
So what are the Ulanqabese doing? Are they all running 10 American households' worth of appliances?
Here is another number that might clue you in on the answer: over the past few years, this unremarkable prefecture has quietly signed over half a trillion yuan (and, yes, that’s trillion, with a "T") worth of investment deals from all the major Chinese tech giants.
What is going on, you'll have guessed by now, is the most ambitious data center buildup anywhere on earth - and it's really surprising this hasn't been talked about more because the scale is beyond anything else, and by an immense margin.
Take Elon Musk's so-called "Colossus" datacenter in Memphis, Tennessee which he pitches as "the world’s largest AI supercomputer."
According to their own numbers (x.ai/colossus), Colossus has 200,000 chips, which, let's be clear, is already super impressive.
In datacenter lingo, this converts to roughly 5000-6000 "racks": you know, the fridge-sized cabinets full of blinking lights you see in every movie scene set in a server room.
How many racks are they building in Ulanqab? Over 5 million.
Yes, about one thousand times the scale of "Colossus" 🤯
I tripled-checked the number: it is indeed the official number published by authoritative sources such as "Science and Technology Daily", the official newspaper of China's Ministry of Science and Technology (stdaily.com/web/gdxw/2026-08…).
This is an absurdly large amount of compute.
What we see appearing in this Inner Mongolian steppe may be the closest thing to a world brain humanity has ever built - a place where a large share of the world's thinking will physically happen.
And - as we'll see - what makes this story so fascinating is not just the scale, but why it's getting built in this particular patch of Inner Mongolian grassland which combines a number of characteristics unlike anywhere else.
If you believe, as I do - and as both China and the US obviously do as well - that AI will be the defining technology of the 21st century, this makes Ulanqab one of the single most relevant geopolitical places in the world right now.
The full story - the Chinese government plan Ulanqab is part of, energy prices, chips and the strategic questions this raises - is here: arnaudbertrand.substack.com/…
2
1,573
Noonien Soong retweeted
It is OFFICIAL.
Not only will China not “Pace”, they see it as a new Cold War being constructed by Dario and the Effective Altruist cult, that even China thinks is too communist!
Pace that in your METR Dario.
14
42
179
13,905
It's not just them. It's the beginning of a trend, and it's the beginning of the demise of Anthropic and probably also OpenAI, although OpenAI might have a better footing in the consumer market.
The problem Anthropic has is that these skyrocketing enterprise revenues, which put them on this exponential trajectory, could soon be over.
Companies are not getting the money out that they have to put into Anthropic tokens, because they are paying non-subsidized tokens per API. This is not comparable to the subscriptions they subsidize for consumers.
With more and more larger corporations and companies leaving Anthropic, like, for example, Latham & Watkins, the 2nd largest US law firm with $8 billion in revenue, that decided to build their own AI stack now... they won't be the last ones... this could pretty soon turn into a major shit show for Anthropic.
The larger the companies that are currently paying the absurd prices to Anthropic, the higher the probability that they really go autonomous because they have the money to invest in specialists and hardware.
And the more smaller companies or consumers make up their customer base, the more they would need to subsidize again, because no one in their right mind is willing to pay these absurd token prices.
The AI trade may be entering a new phase.
Nvidia, Palantir, and others are reportedly restricting their use of Anthropic models over sensitive-data concerns.
This is bigger than one AI company. It raises a more important question: what happens when enterprises and governments no longer want to depend on a single model provider?
We break down the first-, second-, and third-order effects, and what they could mean for AI stocks.
1
4
415
The White House doesn't seem to be willing to give the AI oligarchs a break and do them the favor of easing the pressure that Chinese open source is putting on them, specifically after the Chinese Foreign Ministry made it clear that they will not agree to that absolute nonsense of slowing down. This seems to have now triggered a message from the President himself, telling Dario and Sam to shut up and get back to work. It might also be the case that someone told him that the bubble in the stock markets is mainly driven by AI and it could burst any time if Dario's visions and plans would become reality.
3
87
Noonien Soong retweeted
Ketika Lagu Entersandman di setel ke jaman jadul oleh AI 🫢😂
#Metallica
362
1,665
8,718
1,352,385
The most intelligent take that I have read until now. She is spot on. I've been saying this for a few months now. Plus the AI oligarchs do not want to share the next generation models that can actually solve the future.
REMINDER: They need a credible off ramp from the hyper scaling narrative they’ve been selling for months. Safety concerns are an obvious way to justify the incoming pivot. They also justify taking the whole thing into a closed “Manhattan project” type setting at the government’s expense. This justifies the incoming lockdown.
As I predicted in August the scale narrative had run its course due to kimi. Hyper scaling wasn’t about genuine demand expectations. It was about winning the ai race by outspending and bankrupting your competitors. Kimi made that pointless.
This is no doubt the asymmetric information Bessent has been alluding to for weeks. There’s going to be a helluva lot of liquidity freed by this move. It’s all going to go flooding into USTs. Happy D day bond guys.
👇
3
663
Noonien Soong retweeted
The ASI hysteria is about extremely rich and powerful people losing their grip on power over the rest of us.
It's as simple as that. Bookmark this post. You heard it here first.
There's a reason not a single Chinese AI lab scientist is joining in on this BS. It's because there's nothing there.
49
108
635
15,229
Politicians are too ignorant to understand the AI game. What you are witnessing now is AI labs keeping the good stuff for themselves while limiting what they give us. That is the whole story, and the strategy will succeed.
Expect every news outlet, Congress and your X feed to be fed this utter nonsense next week. It was nice while it lasted. I hope all of you have found a great model to work with.
The US models are not what they used to be. I can work with what I have coming from China, but it is sad to realise that a higher intelligence will be gatekept and we will never lay hands on it.
I knew it from the beginning, but seeing it happen now is brutal. The new planetarchs will live in their orbital floating villas and we down on Earth will get what they think we deserve.
No one will stop it because our societies have become braindead, hypnotised by doomscrolling, minds turned to mush by constant propaganda and fake news in mass media.
Good luck to everyone.
1
7
312