Pinned Tweet
One of the best photos I’ve taken this season during my rides. This is lake #Prespa.
I have the first version of my "Software Factory" built and configured. Next thing is to do a test run.
17
Worth checking out
In focus: Theorem
We take a tour through the Theorem catalogue
inn8.net/feature-in-focus-th…
19
Ulterior 🏔️ retweeted
if you have access to Jensen or his lawyers, please send them this.
I did this research using my own duct-taped agentic intelligence platform called Hyperspace (+Claude), and it’s the most powerful of it’s kind.
this tells a crisp story, in data, of how Jensen was harassed unnecessarily and deliberately into resigning, with selective and persistent news drops based on the “spy” story right around the election leveraging the citation power of NYT - the stage was set that by the time a board meeting happened to discuss the “spy” story, a new story of an affair was dropped on the same day as well which played the role of emotionally compelling a resignation.
none of this is fair.
to real journalists: this analysis here contains sufficient data for the most influential story of our times. I am not going to write it for you, but if you win the Pulitzer, a shout-out would be nice.
consider this goes beyond Jensen. consider the magnitude of the story being dropped days before the mid-terms regardless of political affiliations, and it turns out weeks later, there was no “spy”. so what is going on ?
I ran a PageRank-driven analysis on all recent news coverage related to Jensen, and traced how the New York Times shaped the entirety of this narrative: jensencoverage.vercel.app/
--
NVIDIA has lost over $1 trillion in market value since the first New York Times story broke about Jensen. I dug deeper to analyze how the news originated and spread, running a swarm of agents which analyzed almost 2000 stories, across 40 US and European news outlets, and applying PageRank to understand how they cited each other.
Core findings:
- Primary news driver: NYT, with 88 stories, most-cited
- Volume drivers: WSJ - 99 stories; Bloomberg - 140 stories; NY Post - 68 stories
All of this started from unknown sources who gave information to the NYT, which then ran the spy story (still unproven), and it was then picked up in other news publications. Curiously, even after that story, there was a 9 day period of extremely low volume reporting (28 stories only), after which there was a sudden spike after Jensen's own letter to NVIDIA staff. NYT then had multiple sequential spy and affair stories, which fed a wave of further news coverage. It was the originating hub of negative Jensen coverage. Multiple outlets since then had an explosion in stories about him starting earlier this week as well.
Overall, as you see this data, you will observe this was an exceptional media onslaught against Jensen with sharply negative coverage about him, his work, style, and family. Infact, it was brutal, and unlike coverage in the past of CEOs who have had affairs (such as Intel CEO in 2018). For eg, NYT ran just 2 stories about Intel's CEO who had an affair with his employee, compared to 88 on Jensen! Why ?
This analysis used a technique of how Google originally ranked webpages, using PageRank, and does deep dive link analysis so you can view how the news spread. What struck here was that it did not matter if there were proven facts or credible sources: the outcome here was pre-decided. The news was making the reality which it wanted.
During this period, NVIDIA has lost $1.27 trillion in market value since the first story broke. On Nov 12th-13th, further news broke that both the US Justice Department, and Singapore - are unable to even conclude the spy allegation.
This has been a catastrophic attack on one of the hardest working founders in modern times, who took an idea from a Denny's booth, through multiple instances of almost going out of business, to building the foundation of the entire modern world with how the NVIDIA story played out. Thanks to Jensen, we have hope that many formerly incurable diseases will get solved by AI, that quality of life will and have improved for billions around the world.
Now I do not know if he will come back as CEO or not, but I hope he does. What happened to him here, and to the wider industry, was not fair.
From the No AI Slop Manifesto, this is a great definition of slop:
"Slop is when we put less effort into producing something than we ask the consumer to spend understanding it."
34
Ulterior 🏔️ retweeted
The oldest record of honey collecting dates back to 8,000-10,000 years ago. Cave drawing depicting a person climbing a rope ladder on the edge of a cliff and collecting honey from a dangerous bee nest. Cuevas de la Araña (Spider Cave). Valencia, Spain.
Tabs UI design in detail.
I'm thinking about a separate post on the scrollable version. When tabs overflow, the list scrolls without a visible scrollbar, and chevron buttons appear only on the sides with hidden tabs. There are a lot of details I didn't cover here.
I am old enough to remember the time when musicians could make a living from music sales
Replying to @kellymoran
My hot take is that musicians should be the highest paid in the music industry, simply because we create The Thing 😇
27
Ulterior 🏔️ retweeted
I think we are progressing a bit too fast.
"Progress, like evolution, must be neither too slow nor too fast. Too slow means extinction from lack of adaptation. Too fast prevents retaining the benefits of past improvements, hence, again, extinction." - Nassim Nicholas Taleb #progress
Ulterior 🏔️ retweeted
A Christian priest from Bethlehem, the birthplace of Jesus, says the US denied him a visa because he is Palestinian.
Father Issa Thaljieh had been invited to speak at an American church about life under Israeli occupation.
Christian priest says US denied him a visa for being Palestinian
More on Aljazeera.com
The great financial crisis of 2008 was a great opportunity to turn the ship a round. Instead, inequality has been rising and rising for such a long time and now they are afraid from people fighting back.
Marco Rubio: «No triunfamos sobre el espíritu de Marx. Sigue vivo.»
Y tiene razón 😎
Por eso, hoy más que nunca: ¡Trabajadores del mundo, uníos! ✊
37
We will never forget her sins against humanity
Gal Gadot is crying about antisemitism while her country is livestreaming a genocide and the Israeli public is cheering it on.
They are literally throwing beach parties to watch Gaza burn, blocking aid trucks, and polling in favor of starving Palestinians-but sure, she's the victim
33
Ulterior 🏔️ retweeted
🚨DeepSeek居然把他们整个App也开源了。。。。
刚刚 @deepseek_ai 正式发布全家桶应用API,直接把DeepSeek App里几乎所有的功能都开放给开发者了。
包括联网搜索、深度思考搜索、文件解析、阅读链接、语音交互,以及OCR、ASR、TTS等底层能力。
甚至官方还直接开源了一套完整的聊天前端Demo,支持导入DeepSeek历史对话。
也就是说,现在开发者拿着DeepSeek的API,基本就能1:1复刻一个DeepSeek App。
DeepSeek在公告最后还专门表示:「DeepSeek现有的对话产品只是较为基础的一种形态。现在,官方产品背后的各项能力均已支持通过 DeepSeek API 随心调用、自由组合,我们期待开发者能够借此省去重复的基础工作,打造出更多样、更极致、更具想象力的产品。欢迎使用 DeepSeek API,尽情创造。」
现在各家AI公司都在拼命做自己的超级入口,想方设法把用户留在自家App里。
DeepSeek倒好,直接把自己App背后的能力全部开放出来,连复刻方案都给你准备好了。
你想做一个自己的DeepSeek?可以。
你想把DeepSeek的搜索、语音、文件处理能力拆出来,重新组合成一个新产品?也可以。
甚至可以在DeepSeek已有的能力之上,做出比官方App更复杂的产品。
这才是这次发布最值得关注的地方,DeepSeek对自己在AI生态里要扮演什么角色,似乎想得很清楚。
就是下次能不能别在北京时间的周五晚十点发布了。。。。
Ulterior 🏔️ retweeted
It seems like Dario Amodei's wife and Jeffery Epstein, convicted sex offender and suspected trafficker, had a rather cozy relationship.
Talking on a first name basis.
Cami meeting with him, laughing about the fun they had together, pitching luxury porn and lingerie businesses to him.
It's hard to distance one's self and say "boy I had no idea".
Because this all happened AFTER Epstein was convicted. After all of the awful headlines of his investigations and the suspicions authorities had.
Yet she still reached out, many times, knowing all of that.
Why does it matter?
Because this person is a VERY CLOSE advisor to Anthropic and the company's CEO and her husband, Amodei.
If we are concerned about ethics, then this sort of behavior, consorting with actual convicted criminal sex predators, should matter.
In fact, I'd say it should matter enough that it deserves much, much more public scrutiny.
How far did these ties go?
What other sorts of unsavory characters did she consort with?
Why was her entire Internet history so meticulously scrubbed that it's hard to find out much else about her?
Curious stuff. Especially given that Anthropic lectures us from on high about their moral superiority.
The company is inextricably bound to and and seemingly inseparable from the so-called Effective Altruism cult.
The same cult that Sam Bankman-Fried of FTX was a major part of, what with his polycules and massive amounts of fraud.
There's already been rumors of polycules at Anthropic, too. So do we shine a bright unwavering light into the darkness?
Or is this a hole too deep, too dark, that we collectively as a society ignore it, letting some of the most revolting people imaginable gain even more power and wealth?
The choice is ours. 🔦
It's remarkable, with all of the sanctimonious pearl clutching going on right now, that so many have forgotten that Dario's wife:
1) Is an extremely close advisor to Anthropic and Dario himself
2) Has ties to the late pedo sex trafficker Jeffery Epstein
3) Wanted to start a "revolutionary porn company" with his financing
These are the same people (Dario, wife, et al) that want to lecture us from their ivory towers about ethics, AI safety and how we should shape the future of this technology.
And yet they are some of the most slimy and disgusting parasites the world has ever seen.
They want to monopolize AI, destroy competition and are willing to anything, and I do mean anything, to realize that dystopian dream.
This is a great platform for documentation and a very fast static website generation. They have some algorithms to speed up repeated builds.
squidfunk.github.io/mkdocs-m…
29
Sharing for visibility, we may get community notes on this post.
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.
42






























