Blanca retweeted
Case in point
Blanca 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.
Blanca retweeted
Fate loves irony, so what if...
- Nvidia fires Jensen
- NVDA crashes
- Elon hires Jensen, Jensen hires his best lieutenants
- SPCX skyrockets, NVDA crashes more
- SPCX can now afford to acquire NVDA
- Jensen back in seat, NVDA board is out
Blanca retweeted
i am very glad the news about jensen imploding his company via an extramarital spy affair came out after october 31st. Because sf halloween would have been so, so much worse
Blanca retweeted
Every year I choose a new, obscure and difficult hobby to throw myself into. Two years ago it was offshore racing. Within a year I went from racing dinghies to VO65s and Maxis.
Some of my biggest breakthroughs in business have come while standing watch in a Force 8 gale at 3 am…
A truck carrying $11.4 million in Nvidia hardware disappeared after leaving Oakland, as organized cargo thieves increasingly target expensive computer equipment destined for the data centers powering the AI boom.
buff.ly/OJs1AXq
Blanca retweeted
It must feel so good to eat small plates and drink natural wine while wearing your Palantir chore coat and your Arc'teryx + Bain Capital VC beanie
Blanca retweeted
from a global allocation perspective, it's terrible that people are letting DRAM sit idle in their phones and laptops. we could automate so much labor if we took away everyone's devices and put them in data centers
sam: local inference is a terrible allocation of scarce resources
“a lot of people want local inference, and somehow they assume that the problem is solved now because we have open-source models, we're going to be able to run them on our computers, and so we're not going to need all the stuff in the data centre
but from a global allocation perspective, actually it's quite terrible to have everybody, like, have a bunch of the world's limited supply of DRAM sitting in their own computer and being used five percent of the time
you want maximum utilisation in the cloud if you're trying to deal with this very supply-constrained environment
somehow many people's intuition is the opposite of that”
Blanca retweeted
As you become an adult, you realize that things around you weren't just always there; people made them happen. But only recently have I started to internalize how much tenacity *everything* requires. That hotel, that park, that railway. The world is a museum of passion projects.
Blanca retweeted
Whoa that’s more than I would have guessed they could raise
Still way way less than the leading American companies but way more than enough to build ASI (especially with lax export controls)















































