Alia Abbas (@_aliaabbas) header
Alia Abbas (@_aliaabbas) profile picture

Don't trust a cook, that doesn't have any vices.

San Francisco
Joined March 2018
Photo by Alia Abbas (@_aliaabbas)Photo by Alia Abbas (@_aliaabbas)Photo by Alia Abbas (@_aliaabbas)Photo by Alia Abbas (@_aliaabbas)Photo by Alia Abbas (@_aliaabbas)Photo by Alia Abbas (@_aliaabbas)Photo by Alia Abbas (@_aliaabbas)Photo by Alia Abbas (@_aliaabbas)Photo by Alia Abbas (@_aliaabbas)Photo by Alia Abbas (@_aliaabbas)
we turned the openai x hugging face incident into a desktop you can log into. everyone has opinions on how ai labs handle safety. almost nobody has seen what it looks like from the inside. Huggingbreach.com/?v=
2
3
13
700
If you want to do this and need batteries that can serve sub 1-20ms, I got you! Dm me
If everyone else doesn’t want datacenters, I did the math on my property: -I can support 50-100k GB300 liquid cooled (I have natural springs we could use) -Nuclear power plant close by -It comes out to about 1-2MW of power per acre I could run a lot of sovereign workloads here We could call it NeoChad Taking investments now
1
51
Alia Abbas retweeted
i recreated THE ENTIRE WORLD in minecraft at 1:1 scale. every building on earth, including your house! to do this, i trained a SOTA world-to-minecraft generator on maps data, complete with furnished home interiors! play and check out your neighborhood: worldcraft.gg
1,717
6,459
2,646
123,740
20,800,996
Past v Present. The future is now. Humanity belongs in the stars.
8
125
5
1,584
20,908
Surreal dedication. All for the love of humanity!
6
325
Alia Abbas retweeted
Made on Earth by humans
6,051
18,102
1,466
227,542
7,339,436
Alia Abbas retweeted
This slide from the Anthropic IPO roadshow was a big factor in my decision to leave. It’s an insane plot and I’m surprised they made it public. It shows the compute efficiency of pretraining over time. The y-axis is a **log scale**: they are seeing continued superexponential algorithmic progress. Mythos was not a one-time deviation, but a point on a superexponential. They could already retrain Mythos with a fraction of the compute originally needed. It doesn’t strictly imply a software-only intelligence explosion- they have been spending more on R+D. But algorithmic improvements do not require ridiculous creative genius and are on the verge of being automated. You do the math.
399
549
296
6,307
1,739,931
colossal beauty
Post launch orbital pass south of Starbase of Ship 42 and 26 Starlinks freshly exposed to the vacuum of space
45
Incredible work by the tile team
42
92
13
2,865
84,912
The FAA has issued a nationwide ground stop for all aircraft within the national airspace system "due to extraordinary circumstances and for reasons of safety" by order of the FAA Command Center.
212
982
354
8,405
1,514,094
Anyone know folks at California Air National Guard's 129th Rescue Wing? Want to chat w them
1
53
Alia Abbas 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.
1,823
6,425
2,210
41,253
6,797,348
General Tibo is now in charge of the internal OpenAI security initiative, and has reassigned 1/4 of OAI's production engineers to improving security infrastructure. Dario said in an interview once that eventually the value of model weights will exceed any possible security spend.
OpenAI has partnered with Cathedral systems. This is the cybersecurity startup founded by former DOGE staffers who very recently hacked Anthropic's Slack.
34
37
9
674
42,257
Anyone got a ticket or two to Fireworks Forge (Tuesday)?? Dm me
1
77
Alia Abbas retweeted
Is this basically one of the largest thefts in history?
BINANCE WALLETS SEE $2.77 BILLION OUTFLOW ACROSS FIVE BLOCKCHAINS AMID DRAIN SUSPICIONS; HACK UNCONFIRMED
70
31
5
869
87,041
Alia Abbas retweeted
SITUATION DETECTED: Hackers attempting to breach the US Federal Election Commission system have triggered an outage, the filing service has been down since 10pm last night, per Reuters.
10
40
12
555
48,512
Alia Abbas retweeted
Replying to @ryangr
<< what our Universe looks like to God
27
9
272
23,968
Alia Abbas retweeted
it's the golden age of the white hats. the world has never needed you more. there should be a huge prize pool for finding flaws in civic infrastructure. their work should be strongly protected by statute
144
160
33
2,633
141,054
Never going to get over that the AT-AT walkers in Star Wars were originally conceived by U.S. Steel.
18
79
13
1,737
156,243
Alia Abbas retweeted
if you have a qwen3.5 4b around, please try steering neurons L14 N8949 +1 and L13 N7956 -1 now you have evil qwen recipe: dose × median(‖resid‖ leaving layer L, tokens 1:) / ‖W_down[:, n]‖` at the down_proj at every input
517
1,320
786
22,390
4,567,212