Catmandu (@tradfichad) header
Catmandu (@tradfichad) profile picture

Meow. Janitor at @apolloglobal

Joined February 2016
Photo by Catmandu (@tradfichad)Photo by Catmandu (@tradfichad)Photo by Catmandu (@tradfichad)Photo by Catmandu (@tradfichad)Photo by Catmandu (@tradfichad)Photo by Catmandu (@tradfichad)Photo by Catmandu (@tradfichad)Photo by Catmandu (@tradfichad)Photo by Catmandu (@tradfichad)Photo by Catmandu (@tradfichad)
Theranos but it works
Introducing Helios: measuring 1,000+ protein biomarkers from a drop of blood to help you live younger, longer. Here’s how it works: • Self-collect blood at home with a painless, needle-free device • We measure the proteins carrying signals from cells and organs throughout your body • Your Helios AI agent explains your results, highlights risks, and offers personalized guidance Proudly backed by Y Combinator and Formation VC. Join the waitlist at myhelios.com
26
Hidden lesson that HR is a bane
January 2020. Nik told me the company was burning too much money. Make it profitable by 2021. For the first 3 months I only focussed on revenue. New subscriptions, pricing changes. Then COVID hit. Lockdown hit all of Revolut’s core markets and overnight we lost 40% of revenue. Revenue optimisation was not going to cut it. Declared war mode on costs. Asked finance to show me every single invoice from the last 12 months. Went line by line and labelled each one: renegotiate or kill. Assembled a small team of the best business developers and lawyers in the company. Every cost line got an owner, with one job: exit the contract or renegotiate it. Key vendor contracts were renegotiated. We shut down a whole unused floor of the office just to save on power. We even got rid of office plants because they were leased. It was still not enough. So I made myself the single cost approver for the whole company. Not a single dollar could leave the bank account without my approval. I blocked any spend that did not have justified return on spend. HR was extremely pissed off. The company broke even 2 months ahead of schedule.
30
You may hate Anthropic but you love money more . So you go all in on the ipo anyway
26
♥️ open source is the way
🚨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。。。。
33
Lasth ft gymy ty rrr re
34
I am actually disgusted by the timing of this drop. And this will get a bit more technical. Let’s do a real technical teardown of this "1,385-sigma Blackwell weight fingerprint" report that OpenAI and Anthropic are breathlessly promoting today, because banking on nobody reading past the headline is how DC narratives get manufactured. ​First, look at the actual CUDA maths: when PyTorch initialises model weights (torch.nn.init) with a fixed PRNG seed, curand/Philox parallelises random number generation across the GPU’s Streaming Multiprocessors (SMs). Because Hopper has 132 SMs and Blackwell has 148 or 152 SMs, the exact same seed lays out initial weights in a different spatial pattern depending on the GPU’s SM count. In huge MoE models, certain low-gradient null-spaces or unrouted slices preserve faint directional traces of those step-0 weights even after training. ​So yes, brute-forcing the seed and SM grid can show that a model’s step_0 weight initialisation script ran on a 148/152-SM Blackwell GPU. Now read what the summary explicitly admits in the fine print: 1. Initialisation is not full training: The report literally states that these signatures show how the weights were created, and "support (but do not prove) the inference that Blackwell also participated in model training". Initialising a model checkpoint takes seconds on a single rented dev node before loading the safetensors onto a standard Hopper cluster. The "2,000 to 3,000 Blackwell GPUs" and "$500 million" cluster numbers do not come from the weight fingerprint and Depue admits in his own thread that he simply "estimated" cluster sizes from release schedules and model details. ​2. Zero geolocation and zero proof of smuggling: The final sentence of the executive summary explicitly admits that the findings "do not establish the physical location of the chips, how they were acquired, or whether export control regulations were violated". Depue himself rented a cloud GB200 instance just to run his verification test. Commercial labs that are not on the Entity List renting bare-metal Blackwell nodes in offshore clouds (Singapore, Japan, or the Middle East) is completely legal under BIS rules and leaves the exact same SM fingerprint. Now ask yourself why a former OpenAI researcher dropped this on November 12, the exact day Jensen Huang was pushed out, with OpenAI and Anthropic immediately stepping in to "independently verify" it. US closed-source labs are watching their paid API moats get eaten alive by free Chinese open-weights models. Conflating a step-0 PRNG initialisation seed with a "$500M smuggled Blackwell cluster" hands DC hawks the exact pretext OpenAI and Anthropic want to regulate global cloud compute and kneecap open-source AI.
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.
31
113
15
774
39,568
Catmandu retweeted
the chinese have rightfully earned enormous global goodwill and soft power with their open source model efforts. even most Americans in this space feel more sympathetic to them than with the model companies
189
256
58
4,411
378,486
Been yapping about this for ages. The RSP is PBC is just monopoly shrouded in smokes and mirrors. Closed source is a crutch. Opensource needs to win so that society may win.
1
7
354
Maybe these rumors are true, maybe they’re not. But we are in for a very gruesome societal outcome if opensource fails. Absolute power corrupts absolutely.
6
Catmandu retweeted
Dario Talmudei is responsible for: - Ousting Jensen Huang from Nvidia. - The destruction of rare books for Project Panama. - Blocking the reverse engineering of devices and programs. - Facilitating the mass porting of programs to rust. He somehow managed to be worse than Altman.
This is the man behind the campaign against Jensen Huang. At stake is his IPO. But these barbaric Judeosatanic beasts don't understand that they're going to kill the goose that lays the golden egg by bringing Jensen (and thus NVIDIA) down. The models are not as important as the hardware platform.
124
582
49
3,582
95,278
The US tried for years in many ways to isolate China obviously it’s worked well until… China said fuck it I’ll develop my own ecosystems and now it’s a legit competitor to the US. China has too much grit for this stunt to work
This would mean gpu export controls were not only ineffective, but counterproductive, as a much stronger Chinese GPU market emerged My guess is that seeing this abject failure of policy, the state will double down on it, and the population will generally support it Markets cannot be controlled, nor should they be. Dollar-weighted democracy, without the retardation.
1
113
Catmandu retweeted
asked my russian gf (who is way out of my league) if she is secretly a spy after the jensen news she laughed hysterically for a few, paused, and then said: "you trade animal coins with retarded ppl on the internet" somehow this is even worse than if she were a kgb agent
684
1,609
163
54,453
3,265,947
Wot
Trump on Melania-brokered ceasefire: Both sides wanted to do it
28
Why does she look like shes literally insane
Lucy Guo on why AI isn't shrinking her workweek: "AI is making people work harder actually, because you have to be awake for your agents to be running." She pulled a 26-hour day last week because Claude was shipping code and she refused to miss the loop.
65
This is bull market slop and a shit strategy. Most of the time this strategy is under water esp extended back to 1997 and outperformance relies basically on just a few outliers ie nvda. Another reason why not to trust everything you read online
1/ I WONDERED WHAT HAPPENS IF YOU JUST BUY THE BEST PERFORMING STOCK FROM THE YEAR BEFORE AND DO IT EVERY SINGLE YEAR SO I BACKTESTED IT GOING BACK 15 YEARS STARTING WITH $100 IN 2012: END RESULTS S&P 500 → $800 LAST YEAR'S WINNER → $176,114
47
Bullish unlocks
Hyperliquid is on another level, OTC’ing a year of unlocks to an investor seeking an entry (vs. an LP providing a bid) is a brilliant, masterclass in supply management. Mad respect
61
Zuck is the new Jobs
1
35
Catmandu retweeted
Woah… a new report on Anthropic vs. White House negotiations over Fable says things got so heated that Trump apparently wanted them JAILED. (per Politico) > government tests Fable > Treasury approves its release > 2 days after launch Amazon does a jailbreak > White house tells Anthropic to take it down > Dario says jailbreaks happen for all models and it’s not catastrophic > tries to explain why > Bessent: “Well, I don’t think that’s really necessary, Dario” > Dario said he couldn’t just shut down Fable on an hour’s notice > a frustrated Trump reportedly says “I want to send them to jail” > White House blocks access for all foreign nationals, including Anthropic’s OWN employees > Dario asks if that means they have to take the model down > Lutnick: “Yes. That’s the point.” > Anthropic takes it offline worldwide > they negotiate > Fable returns after 19 days
Politico has posted a long article about the Fable drama, aka Anthropic vs The White House, it's a good read if you are interested in the events that took place during the negotiations.
127
226
45
2,447
693,933
Catmandu retweeted
DeepSeek kernel engineer: “In particular, I do not want Anthropic to control the world's most advanced artificial intelligence of AGI. To put it dramatically, I consider the stakes comparable to Hitler obtaining atomic-bomb technology before the Allies." 🤯
astonishing blogpost from Shengyu Liu (刘胜与, also known as interestingLSY/intlsy), kernel engineer at DeepSeek. The first part is his personal struggle with the fact that his work is about to render his beloved craft obsolete. The second is… well. let's just say we agree.
16
91
3
1,548
98,972
*cough* running *cough* of *cough* compute
JUST IN: Anthropic CEO Dario Amodei says the company will slow the pace of AI model improvements
91