←
Charlie Warzel34,290 posts
Charlie Warzel avatar

Charlie Warzel

@cwarzel

Staff Writer, @TheAtlantic. I host Galaxy Brain, the first and only video podcast.

◷ Joined February 2009▧ 4,326 Photos and videos
Charlie Warzel avatar
Charlie Warzel@cwarzel·12h

well then

really interesting context for the 'everything is breaking' concerns. i'm even more inclined to take the side of 'the system is working and people are proactively patching stuff to be resilient in the face of an ai cyber threat'

Charlie Warzel avatar
Charlie Warzel@cwarzel·17h

Like—actual questions here—would enough people have noticed or cared about this kind of disruption 4 months ago? Is this an 'outage' or is it more of a thing that happens when systems get updated and it's only in the broader context that people care rn? Questions worth asking! x.com/DeItaone/status/212277…

*APPLE SAYS SERVICE OUTAGE CAUSED BY INTERNAL UPDATE

Charlie Warzel avatar
Charlie Warzel@cwarzel·17h

Feels like 4 things are colliding RN: - Genuine fears about AI agents/cyber vulnerabilities/hacks - Very sensitive media/public to these concerns - A systems infrastructure that is working overtime to build resiliency to respond to a specific moment and perhaps finding vulnerabilities, patching and making new problems, and causing disruptions that are noticeable - An unreliable narrator in terms of the administration But it seems like this tracks with other moments like when everyone worried about the uptick in air travel problems in 2024 after a few incidents or how the press gets more attuned to foodborne illness outbreaks etc.

Charlie Warzel avatar
Charlie Warzel@cwarzel·19h

really interesting context for the 'everything is breaking' concerns. i'm even more inclined to take the side of 'the system is working and people are proactively patching stuff to be resilient in the face of an ai cyber threat'

Pushing back 20 years now, and 2013 Aug was worse it seems.

Post attachment
Charlie Warzel avatar
Charlie Warzel@cwarzel·22h

if this theory is correct what's interesting is that all these disruptions that seem to be freaking a lot of people out would actually be an example of pretty serious responsiveness/resiliency in our various digital systems

There are lots of outages... Why? AI attacks? My more normal guess is that because there have been a lot of AI-discovered cyber vulnerabilities (including lots probably not yet disclosed), companies across all industries are having to do a lot of patches and upgrades, probably faster than usual. Such upgrades can sometimes go wrong, causing outages.

Matt Lanza avatar
⟳ Charlie Warzel retweeted
Matt Lanza@mattlanza·Nov 13

NWS radars are working fine. There is no outage and nothing related to the FAA disruption.

The misinformation about a non-existent nationwide NWS radar outage has gotten out of hand. @Reuters is incorrectly reporting this as news, not having done sufficient due diligence before posting an article about it. reuters.com/technology/us-na…

Post attachment
Charlie Warzel avatar
Charlie Warzel@cwarzel·Nov 13

you really notice in a big, uncertain developing news moments how many people who’ve built up really big accounts here are just deeply unhelpful when it gets serious

Jason Goldman avatar
⟳ Charlie Warzel retweeted
Jason Goldman@goldman·Nov 13

The NVIDIA board was filled with loyalists who had made billions because of Jensen Huang. The reason they ended up in a situation where the government burned Jensen is because of actions that were so egregious even the laissez-faire Trump admin couldn't sit idly by. If the board had not been filled with loyalists and if instead there had been real governance maybe these lapses in judgement would have been prevented. If your takeaway from this affair is "BOARDS SHOULD BE MORE COMPLIANT WITH MESSIANIC CEOs", brother, I don't know what story you've been watching.

@jason@JasonNov 13

RESET ALL BOARDS TO BE RANKED BY SHARES HELD. NO MORE DEI AND VIRTUE SIGNALING BOARDS NO MORE "EXPERTS" ONLY WARRIORS WITH SKIN IN THE GAME

Charlie Warzel avatar
Charlie Warzel@cwarzel·Nov 13

not sure if anyone near the admin has any credibility on this but this strikes me as a good sign that whatever is coming will be messy and protracted

Bombshell story. The spy narrative is collapsing already. Is there clear & convincing evidence that Alice Huang was a Chinese spy? - AG says no. (“no conclusions reached”). - CIA says no. (“not yet convinced”). - DNI says no (“not entirely accurate and over-torqued”). - Ranking Democrat on House Intelligence (incoming Chairman) says no. (“today, at least, I don’t know”). This was immediately after he was briefed. Who argues otherwise? Anonymous officials tell CNN “it strains credulity to believe that she was not acting as an agent of the Chinese government — simply by virtue of the fact that she is a successful Chinese businesswoman working with one of America’s most powerful companies.” This is their proof after a year of investigating and surveillance? That she was a “successful Chinese businesswoman”? At this point, the only crime that we know for sure happened is the leaking of the classified memo. Who did that and why? An investigation should be opened.

Charlie Warzel avatar
Charlie Warzel@cwarzel·Nov 12

eep

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.

Post attachment
Charlie Warzel avatar
Charlie Warzel@cwarzel·Nov 12

Ah! Well. Nevertheless. reuters.com/technology/forme…

The fact he only had one extramarital affair with a state asset after leading the greatest technological revolution in history is what impresses me. For anyone else that number would be double, even triple, digits.