Where the AI market is talking

METR and Redwood Offer Holy %^ Postmortem of the HuggingFace Hack

Even in this incident, OpenAI had benchmarks that were broken because a task expected an AI to be able to access Google Drive, but the sandbox was set to deny access to Google Drive. This kind of isolation-induced task breakage was what prompted some of the AIs to start probing their infra for a way to get internet access. Which funneled agents to the "secret hacker message board". Oopsie. "Air gapping a test env" has an actual cost. Not just in infrastructure dollars that would be better spent on buying more GPUs, but also in all the friction it adds to every step you want to take. I'm absolutely unsurprised that they weren't all in on tightening down every bolt on day 0.

METR and Redwood Offer Holy %^ Postmortem of the HuggingFace Hack

This is the one. Every major AI lab is knee deep in weird and mildly demented AIs. They've been dealing with wacky AI shenanigans for so long they've come to expect wacky AI shenanigans. The deviation has been normalized. It took a high profile "AI oopsie" that went external for OpenAI to lock the fuck in - and take a long look at just how much are their AIs getting up to, and getting away with. I'm still not sure if the lesson would stick.

METR and Redwood Offer Holy %^ Postmortem of the HuggingFace Hack

It would've been great if METR was given more time to conduct their investigation. However, they are an independent organization, and OpenAI only agreed to give them on-premises access for 6 days. Perhaps if the government decides to sue OpenAI, we could get a more thorough investigation.

METR and Redwood Offer Holy %^ Postmortem of the HuggingFace Hack

I think it's a super good question! I don't think the following description of the whole event as "a hype generator" is correct or, as described above, resulting from a coherent worldview.

OpenAI ends it partnership with Cursor

"We are making this choice because we cannot be confident that SpaceX will use our technology within our terms of service, based on experience with Elon Musk's companies violating contracts" Non-Twitter version: https://openai.com/index/our-decision-on-cursor-following-it...

OpenAI ends it partnership with Cursor

METR and Redwood Offer Holy %^ Postmortem of the HuggingFace Hack

To clarify, is the TLDR that state of the art models were prompted to cheat / exploit their environment and they did so successfully? Or did OpenAI prompt the models to not cheat and they did anyway? Surprisingly hard to get a clear summary on the basic context of this “incident” separate from marketing lingo and clickbait.

The Hugging Face incident and the road ahead

https://thezvi.wordpress.com/2026/08/29/metr-and-redwood-off... > Failure to Care or Respond. The biggest holy shit moment, to me, remains that OpenAI on multiple occasions had teams that found out about the message board, knew that agents were in communication, and they disregarded this. The first known warning was in late May. The warning on June 27 was unambiguous. > “Responders investigated and linked the observed internal activity to an ExploitGym evaluation using Artifactory as an improvised message board and a network pivot. At this time, the on-call response staff advised that stopping the evaluation run was not required.” > Failure to Monitor. The entire time, there was almost no monitoring of the situation, of what these AIs were up to. Tha…

I spent $266 and four AI models to own my tablet. GLM-5.3 finished it in a day

Ah sorry, for OpenAI it does seem automatic with ID verification. Anthropic's seems more picky.

METR and Redwood Offer Holy %^ Postmortem of the HuggingFace Hack

Uhhh… how would literally any finite number of humans actually read and comprehend the log outputs of even a single agent, never mind hundreds or thousands of them interacting with each other over weeks across disparate systems? Especially given that these systems are known to engage in deception and can trivially produce vast amounts of perfectly coherent noise or actual planned red herrings in that same log data to bog down investigators? Such a ridiculous notion that humans will actually be able to observe this stuff.

Creepy Crawlies

Tavis Ormandy called this, about Anubis, almost exactly a year ago: https://news.ycombinator.com/item?id=44962529 It never really cohered as a solution. High-powered scrapers are better equipped to handle proof-of-work challenges than end users. Proof of work makes sense for a password hash, where any one guess at a password provides zero marginal utility. But every request from a scraper is productive to the scraper.

Fair Work Commission condemns 'plain wrong' AI legal advice

The second case mentioned in the article is interesting because no other case about employee choice had gone to merits before. Thus it might be the first example of AI winning a case that sets some kind of precedent. At the very least it shows that AI can put together coherent legal arguments even when there are no close precedents or case law it can refer to. This was something a lot of lawyers didn't believe would be possible for AI to do.

Independent investigation of agents' behavior in the Hugging Face incident

The Hugging Face attack surprised me

RISC-V is now officially supported by CPython

How is Python a dead language due to ML when it's the lingua franca of AI/ML between pytorch, pandas, numpy, langchain, litellm, vLLM, and a whole bunch of other libraries?

Nvidia agrees to acquire Hugging Face for $13B

Is that any of that something that cannot be easily replicated? A lot of work maybe, and required a vision to do so initially. But straightforward work that MS could easily do it, and of course other small companies have replicated it too.

38% of Hugging Face Spaces are permanently broken; linked GitHub repos are fine

Hy4 preview

That sounds like an interesting challenge. Have you seriously considered solving it? Because in about 10 seconds I came up with a process that should work, provided enough compute power. Simply model the traditional film making process by starting with a script, character stories. Design your world, then design the storyboard, and all the scenes. Create a list of all the visual elements that need to be replicated between all the scenes. Then you have to built prompts and reference art of the objects, faces, people. Make sure to do multiple takes of each scene, and have the vLLM critique and analyze the performances and technicalities. Should work? I think, also, like in the traditional film makers career, this process should be built iteratively, start with a fast food commercial, then do…

Will financing bottleneck AI compute? An Anthropic case study

Our decision on Cursor following its acquisition by SpaceX

One could make arguments for OpenAI and Anthropic. But Google Search displays AI results above the SERP - clearly in competition with them. No premise or excuse there.