Old Fucking Punk @lwriemen@librem.one

I actually like coding and talking about code. What’s the point of the job if you take away the fun part?

No, OpenAI's new magic models did not autonomously hack Huggingface.

Per OpenAI's PR blog post (https://openai.com/index/hugging-face-model-evaluation-security-incident/):

"This incident occurred during an internal evaluation which prompts models to pursue advanced exploitation using complex attack paths, in an effort to quantify their cyber capabilities. We estimate maximal cyber capabilities by running this evaluation without production classifiers used to prevent models from pursuing high-risk cyber activity."

So they prompted their model running without guardrails to hack some shit. OpenAI told it to do that, the model didn't do shit autonomously (no LLM ever does anything autonomously, it's always prompted).

The whole story is PR. We know that from Anthopic's Mythos: "Look we have this super secret new model and it's so powerful. We are scared ourselves. And you will soon be able to rent it!"

I’m surely not the first to come to this realisation, but the LLM industrial complex is to tech-life what MAGA is to politics: erode any and all trust in established institutions, people and practices for the benefit of the very few and worst people on the planet in the (poor) guise of freedom.

The other issue is more personal: you can’t have conversations about code with people any more. Ask them for their personal opinion on a piece of code and you’ll just get a chatbot text extrusion back

Makes the job a whole lot less fun and personable.

This kinda bothers me more than the bug misrepresentation

One disconcerting thing I encounter when working on projects with LLM users: their models do find some bugs (though also a tonne of stuff in the ‘that’s not a bug, that’s just software” category) but the explanations are often bullshit. As in, they misrepresent how the code works. The bug, though it exists, isn’t as described and the proposed fix would just make things worse

The understanding of anybody relying on these tools, AFAICT, could slowly drift away from factual reality of the code

This is #AIhype disguised as a confession. OpenAI wants to pretend that its product has supernatural powers. The LLM didn't "go rogue" and it didn't "infer" anything. It did what its programmers designed it to do, when they gave it the necessary information to target Hugging Face and failed to isolate it in a sandbox.

https://www.nytimes.com/2026/07/21/technology/openai-attack-hugging-face.html?unlocked_article_code=1.zlA.kI7n.HgoniJc3OqNw&smid=url-share

It's essential that people understand the flagrant contempt for law that Elon Musk demonstrates on an ongoing basis. A current example: dozens of fossil fuel generating turbines with no permits and poisonously placed near black communities. https://www.reuters.com/world/americas/pollution-musks-unpermitted-xai-power-project-hits-hardest-black-communities-2026-07-14/

#pitchforks

I keep seeing people who have become very dependent on LLMs for everything make the point that we can’t complain about AI because the environmental impact of the meat industry is worse, but

1) A lot of them aren’t vegan/vegetarian themselves
2) It’s not because the meat industry is bad that other things can’t be bad in addition
3) We should have the capacity to be against two evils simultaneously, reduce impact on several fronts
4) There are lots of other issues with AI besides water usage

😩

RE: https://salisburyai.com/2026/07/20/bbc-bias-in-gaza-conflict/

‘BBC is systematically biased against Palestinians in Gaza war coverage‘

Good morning Fedi friends,

I have been feeling a rising anxiety over the state of the open web and the fediverse - after seeing efforts by politicians all over the world (in Europe too!) to introduce age verification laws for social media.

My way of coping?

Teaching people how to self-host their own Fediverse profile (using the superb #GoToSocial) via #YunoHost.

Chapter 1 in this series is available here:

🔗 : https://blog.elenarossini.com/a-newbies-guide-to-self-hosting-gotosocial-part-1-installation-and-set-up/

#resist #blog #ActivityPub #selfhosting

A former prison warden argued that a young defendant should be spared execution because prisons “need that labor.”

Calvin John Smiley examines what this call for “sweat equity” reveals about incarceration, exploitation and punishment.

https://inthesetimes.com/article/julio-foolio-unions-and-sweat-equity

Join EFF's Cooperating Attorneys List!

This mailing list is where we post short case descriptions, and interested attorneys simply drop us a note to get connected with the client.

Interested in joining? Send an email to info@eff.org with a subject that says: "Add Me to Cooperating Attorneys."

This book is a guide to help workplace organizers get started, keep going, and get unstuck when things aren’t going as expected.
Get 20% off this new title with code JULY thru 8/1 here: https://pmpress.org/index.php?l=product_detail&p=2146

DHS expects to spend $1 billion in taxpayer money to nearly triple the number of surveillance towers along U.S. borders from 830 to 2,300, expanding a surveillance network that threatens the privacy and civil liberties of everyone in the border regions. https://www.eff.org/deeplinks/2026/07/explosion-surveillance-towers-coming-us-borders-costing-over-1-billion

Orbital AI datacentres come with latency issues. If they're in LEO, they will have moderately low latency to a moving point on Earth, but for a fixed client the latency will increase as more relays are needed to account for curvature, so the jitter is high. If you put them in geostationary orbit, the latency is fixed, but large.

The solution to this is to put not just the AI datacentres, but also the AI bros in orbit. If you put an AI bro and a small radio in LEO, in the same orbital trajectory as an AI datacenter, they will have very low latency - indeed, significantly lower latency than ground-based routing if both they and the datacentre are on the ground.

Importantly, AI bros have a lot less mass than datacentres. As such, they are much cheaper to launch. As a pilot programme, to assess feasibility, I recommend launching a test constellation of AI bros into LEO. Note that, unlike datacentres, AI bros are made out of non-toxic (at least, in the chemical sense) substances. The environmental impact of uncontrolled atmospheric reentry is significantly lower, making the risk from the failure of a pilot deployment significantly lower.

Decoy is an ambigious font that "hides" from AI - https://boingboing.net/2026/07/17/decoy-is-an-ambigious-font-that-hides-from-ai.html also humans...

#EVs outsold petrol and diesel in Germany for the first time. The birthplace of the car just changed its fuel. - https://thenextweb.com/news/evs-outsold-gas-diesel-germany-first-time but too late to save the industry...

Tom Homan Again Promises That ICE Will Keep Murdering People Until Democrats Stop Calling Them Murderers - https://www.techdirt.com/2026/07/21/tom-homan-again-promises-that-ice-will-keep-murdering-people-until-democrats-stop-calling-them-murderers/ great headline

Argentina Joins Global Treaty to Fight Illegal Fishing - https://www.pew.org/en/research-and-analysis/articles/2026/07/17/argentina-joins-global-treaty-to-fight-illegal-fishing well done, #argentina

Iranian women being shut out of engineering fields, official warns - https://www.iranintl.com/en/202607209913 well, that's stupid... #iran