Pinned Tweet
This is 2025
if you can:
- focus for 2+ hours at a time
- delay gratification
- think in first principles
- handle discomfort
you are a mutant in 2025
everyone else is either distracted, triggered, or mentally weak
the new 1% isn’t just the rich—it’s the people who can still think
1
6
1,153
Shabah retweeted
If you ever wonder why my only goal is the enjoyment of technology is because I have been forgiven.
I have nothing to prove or to win. No X arguments or tech arguments to assert why I am the best, nope. I simply get to enjoy software and live the purpose my Creator gave me. I believe my "spiritual gift" is creation. I love creating and want to share it with everyone.
Jesus saved my after 19 years of decently destructive behavior. I would trade everything for this gift.
Jesus saves and God is good. 🕊️
I don't talk about it much, my faith is a big reason I feel so strongly about people who scam, cheat and steal. Right now there's a culture of people who think it's the only way to get ahead in our world. It's not only misguided, but self-defeating. It works short-term but it always ends the same way:
My son, if sinful men entice you,
do not give in to them.
If they say, “Come along with us;
let’s lie in wait for innocent blood,
let’s ambush some harmless soul;
let’s swallow them alive, like the grave,
and whole, like those who go down to the pit;
we will get all sorts of valuable things
and fill our houses with plunder;
cast lots with us;
we will all share the loot”
my son, do not go along with them,
do not set foot on their paths;
for their feet rush into evil,
they are swift to shed blood...
These men lie in wait for their own blood;
they ambush only themselves!
Such are the paths of all who go after ill-gotten gain;
it takes away the life of those who get it.
Proverbs 1:10-19
Shabah retweeted
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.
Shabah retweeted
So is the story that he was yapping at the tomb of the unknown soldier, got reprimanded by the guard there, and when people wrote about it he gave the guy a medal as a way of saying yeah I fucked up?
This is the most graceful thing a Trump official has done
Shabah retweeted
We just shipped AWS Emulator, which gives coding agents a local copy of AWS to test against.
Early in my career, we built test environments like this by hand, one system at a time. Now, agents are writing code far faster than any team can check it in production. Matt Goldberg, a Senior Principal Engineer on our Developer Tools team, wanted to know whether agents could build and test those environments themselves. He gave @kirodotdev a single prompt, and about 15 minutes later it came back with a first version of an SQS emulator. Turing that into a robust solution took much more work
Here’s what I think this means for engineers today: allthingsdistributed.com/202…
Shabah retweeted
I used Gemini 4 Carbon to get Windows 11 running on my iPhone Duo! 💻
Unfold the screen and it's a pocket PC, including Antigravity running coding agents on my phone.
This is the actual desktop ARM OS running locally with touch controls.
Demo + repo below!
Shabah retweeted
I updated my quantization compiler to use e-means as a replacement for k-means in codebook generation and got Qwen3.5-4B from 8.41 GB to 1.23 GB while retaining 91.9% of original performance which I think is approximately SOTA, at least comparable to bonsai 2 but I think we could go lower if someone wanted to donate some GPU time since this is just a hobby for me.
at this size, it's small enough that memory bandwidth isn't the bottleneck to decode speed so I'm interested in seeing how fast we can get larger models on consumer hardware. link below.
Shabah retweeted
What happens when you combine embedding models with System 1 decision models? You get the intersection, which is VENN —> Vector Decision Models, a very cool OSS research project from Mark Russinovich that I was fortunate to review early and am excited to support.
Shabah retweeted
Introducing e-means, an algorithm beating k-means in 10x fewer epochs.
With order of magnitude wall-clock speedups, e-means is released in FAISS and needs no hyperparameter tuning beyond k. It is robust to various scales and datasets. Still skeptical ? Just try e-means [1/10]🧵
Shabah retweeted
> be El Salvador
> send nerds to Haiti for UN thingie
> gang kidnaps Salvadorian nerds
> bukele says "give back"
> gang "nah"
> bukele says "ok we take back"
> gang "nah, also you're a bitch"
> bukele comes to take back
> gang "wtf that's illegal, UN tell them something"
> UN says "idk bro thats crazy"
> bukele goes on social media
> "does anyone know this houses coordinates? we will bomb it"
> crowdsources OSINT nerds
> OSINT nerds deliver exact coordinates in like, 30 minutes
> bombs it exactly as promised
> goes on social media
> calls them pussies
tl;dr OSINT nerds on Twitter called in Airstrike like it's Call of Duty, never been seen before on social media
Shabah retweeted
Replying to @PalmerLuckey @lulumeservey
Shabah retweeted
Fascinating. Evals-as-indexes.
It basically means structuring and storing the output of evals and allowing agents to query by those metrics later.
You need LLM-as-a-judge plus a storage layer capable of storing and querying those indexes along with the raw trace data.
Shabah retweeted
Dario> We are virtuous! (steals the internet)
China> We stole the internet frfr
Z.ai chief scientist Jie Tang just revealed the training pipeline behind the upcoming GLM-5.5 (32B, 122B, and 256B) during a lecture at Tsinghua University today.
Instead of relying on public datasets, their pretraining pipeline pulls from six separate internal engines:
- Web: 1 trillion pages crawled at ~10B pages/day with PB-scale template clustering
- Documents: 2 billion scientific PDFs parsed through GLM-OCR (4x faster with speculative decoding)
- Books: 50 million books with layout reordering and garbled-text filtering
- Code: 500 million repositories turned into executable tasks with real tests and issue descriptions
- Video: 10 billion videos for long-sequence world knowledge
- Audio: 300 million hours rewritten via ASR into trillions of tokens






































