Agent hacks, USG vs Kimi, & More

I have new proof for you that I’m not an AI. This time, it’s another podcast, as I had the opportunity to appear in The Creative Process to talk about all things AI from a less-discussed angle: its impact on society and jobs, but from a more sober, anti-hype view.

THEWHITEBOX

TLDR;

Welcome back! This week, we take a look at OpenAI’s cyberattack on HuggingFace, the USG vs Kimi, Google’s earnings, Vera Rubin’s incredible first results, and more.

Enjoy!

CYBERSECURITY

GPT-6? hacks HuggingFace

As published by OpenAI and HuggingFace, an unreleased AI model from OpenAI (possibly GPT-6) hacked HuggingFace’s production infrastructure.

In a nutshell, when testing two models, GPT-5.6 Sol and an “unreleased” model many believe is GPT-6, on whether they could pass certain tests in a cybersecurity benchmark, the models autonomously decided they would hack the source of the test, HuggingFace, to retrieve the answers.

And they did.

In response, HuggingFace used open models, GLM-5.2 to be precise, to understand the problem and solve it. Most interestingly, they tried using Anthropic and OpenAI models, but the AIs refused to help, which is beautifully ironic.

If you wish to better understand this whole thing, if go into full details in this piece.

TheWhiteBox’s takeaway:

Sadly, many will use this news to fearmonger about open-source models and how they should be banned to prevent these things from happening.

However, it’s precisely the existence of powerful open-weight models like GLM-5.2 (Chinese, by the way) that prevented this from being much worse, because in the words of HuggingFace:

“Thanks to this approach [using a powerful open model], we were able to do in hours what would usually take days, and match the adversary's speed.”

If anything, what looks like a great PR campaign for OpenAI “look how scary my models are” is in fact a PR campaign in favor of open source, considering HuggingFace could not protect itself using frontier models because these models were guardrailed; this is the best advertisement one could possibly find for open models.

HARDWARE

Vera Rubin, an astonishing 10x performance improvement

Source: CoreWeave

We’re finally starting to see the performance improvements that the NVIDIA Vera Rubin NVL72 servers, NVIDIA’s new GPU platform, bring… and it’s impressive.

For the same interactivity, around 150 tokens per second for every user (a more than acceptable user experience), a Vera Rubin NVL72 server offers a 10x improvement in token throughput per megawatt relative to a Blackwell GB200 server (previous generation server).

In plain English, for every megawatt of GPUs a data center company has, they can now generate ten times more billable tokens while offering the same quality experience for the user. We’re talking about a 10x improvement in performance per watt from one platform to the next, completely outrageous.

But how is this possible? The answer is, as you probably guessed if you’re a regular of this newsletter, memory.

As we have discussed tons of times, AI inference, serving AI to users, is “memory-bound”. This means that the performance is not…

Enjoyed the introduction? Subscribe or sign in to continue with this edition.