Old Fucking Punk @lwriemen@librem.one

So @protonprivacy has been in and deleted the top upvoted customer comments about how to improve. “That idea has been deleted.”

Example:

https://protonmail.uservoice.com/forums/945460-general-ideas

All the ones highlighted a few days ago are gone.

They flat out lie to get them built. All of the centers, everywhere, are pitched by ruthless liars.

Lanarkshire AI datacentre to fully rely on National Grid despite renewables claim

DataVita has revised plan for Chapelhall development to be wholly powered by 1GW of renewable energy

Kat Jones, the director of the Edinburgh-based charity Action to Protect Rural Scotland, said DataVita’s apparent change in direction was “emblematic of the hyperbolic claims being made by datacentre developers in Scotland”.

“Hyperscale AI datacentres simply cannot run on renewable power alone, they absolutely need a highly reliable source of grid power,” she said. “It is not surprising that the goalposts have now changed.”

https://www.theguardian.com/uk-news/2026/nov/07/lanarkshire-ai-datacentre-datavita-chapelhall-national-grid-renewable-energy

#humanemacs #emacs

I've just read a lot of the human emacs thread on the Emacs mail list.

It's seems virtue is lost.

1. To use LLMs is to support techno fascists and proprietary software. This is against the foundational tenets of gnu and free software.
2. Supporting techno fascists supports their anti free software ideology.
3. If gnu/emacs is true to software freedom, it cannot support proprietary software in it's use or creation.
4. Using LLMs creates the strong possibility of dependency which would completely undermine software freedom.
5. Techno fascists and their LLMs are anti freedom.
6. Anti freedom of mind, people and software.

Enough reasons.

"Attack and dethrone god" (ENG: English)

Source: instagram:@churchofsatananarchist

Original: https://gateway.ipfs.anarchiststickersarchive.org/ipfs/bafybeigvxlq5tkwvpfc22xcma5o7qjpti6bhe5oa7ynufefqsieacpo32i

Permalink: https://anarchiststickersarchive.org/stickers/8185

#anarchism #stickers #archive

irregular reminder that if you're on #Codeberg and you need more CI than the hosted offer provides, i might be able to help

RE: https://digipres.club/@alexandra/117576651383899283

llm user: uh just b/c llms cause harms elsewhere, that does not entail they will cause harm in our project

also llm user: I have adopted a moral framework wherein I am compelled to profit from slavery

You know the jokes about "keep your books from before 2021"?

I just got a hard lesson. That is not a joke. Keep your 2021-era textbooks.

Someone asked me to have a look at some online learning materials, "Last Updated : 8 Apr, 2026" because something seemed wrong.

So I read it. The text does not parse as LLM output; there are no telling phrase structures, no em-dashes, no alerting idioms.

But the content? The actual thing to be learned?

It is coherent and completely wrong. The method and maths are right, but the intrinsic meanings of parameters are both completely plausible and utterly wrong.

It's like giving walking directions to someone in a car. Yes, the stairs to the beach are there, and they do get from the parking lot to the water. Yes, cars are things for going places. But if you try to do that - and you can! - you get a car submerging in the surf, not a 50kph-capable vehicle with no stoplights ahead.

And since there's no physically obvious problem, no car filling with water, just a number that does indicate things in other contexts but here is not linked to reality? Then decisions could get made, statements published, if nobody picks up on the extraordinary mis-use.

Keep your textbooks, especially anything with methods or maths!

#LLM #AI #LLMs #slop #knowledge

More evidence for the pile:

> ...[AI] has yielded __zero financial payoff to corporate America outside of the tech industry__. This is the conclusion of an analysis by prominent Danish economist Torsten Sløk, who is a partner and chief economist at Apollo Global Management and has previously held senior positions at organisations including the OECD and the IMF, where he was responsible for the analyses featured in the World Economic Outlook.

https://ing.dk/artikel/leading-economist-shatters-illusion-only-tech-giants-are-profiting-ai

Apparently now in macOS 27, when you type a mathematical expression, you get an incorrect result in large text at the top, and then the correct result in smaller text underneath. Awesome.

And before you dismiss my scenario as made-up nonsense, remember that Patent Trolls exist. All these companies do is find/buy/license patents, and look for products that violate those patents. They do this by painstakingly reading patents and finding ways that seemingly random products could be violating those patents.

Well, guess what we just spent two trillion dollars on that can read a huge bunch of source materials to find potential matches? That’s right: Spicy Autocomplete. Even with low hit rates, these stochastic pattern-matchers will yield enough potential matches to make a shake-down business possible. It doesn’t matter if you actually copied the code; it only has to be *plausible* that you *might* have copied the code—and using LLMs means it is definitely plausible, because the original source is likely somewhere in the training set of whatever model you used.

I think it’s not going to take that long until this happens.

I’ve never felt comfortable using LLMs to generate code for anything that gets released to the public. Not just for provenance and licensing reasons (although those alone should give you pause), but also for copyrightability reasons. AI art is already deemed un-copyrightable (https://www.theverge.com/policy/887678/supreme-court-ai-art-copyright)—not that that stops even large companies from generating AI slop, which boggles my mind—so it seems to me that we are a small step away from finding out that no code generated by LLMs can be copyrighted at all.

The way I (not a lawyer) see it, using LLMs in my (at least published) software projects puts me in a risky legal situation. I simply don’t do it.

#ai

Look, we all already know that any code that an LLM generates is plagiarized, basically by definition. And if that plagiarized code was copied from a codebase with anything other than a CC0 or Public Domain license, your copy has a distinctly nonzero chance of violating some sort of license requirement; perhaps some copyright, attribution, or re-licensing clause.

I think it’s only a short matter of time until someone develops a code-scanning tool to look for potential license violations in popular FOSS projects found in commits created by LLMs. They would then use those commits to shake down projects for money, demand licensing fees, or demand IP ownership of certain features, based on those commits.

If you’re a closed-source software behemoth, maybe you’ll never deal with this, because nobody can directly inspect your code for violations. And besides, it’d take too much money to sue you. But if you’re a FOSS project with any grant or foundation money behind you? You’d look like a pretty juicy target.

#ai

Just to state really clearly what is going here.

The New York Times is leading an information warfare operation to mislead the public that Chuck Schumer has the votes, and is the inevitable next Majority Leader.

This is not the case. Schumer has significant opposition both from new and incumbent Senators.

We have see the New York Times use this disinformation tactic many times before.

Do not be misled.

https://mastodon.online/@mastodonmigration/117536299911092203

All right, we now have plenty of French translators, thank you!

We need more translators for:

⁃ Polish
⁃ Spanish

I would like to get translators for:

⁃ Mandarin Chinese
⁃ Hindi
⁃ Japanese

#noAI #translation

https://sfba.social/@drahardja/117576862013078469

The change dot org petition to not elect Hakeem Jeffries—the one with more than one million signatures—has an AI image on it.

Way to send mixed signals, gang.

Here's why garment factory workers in Bangladesh are striking:

Take a woman working a 12-hour shift in a textile factory in Bangladesh. During her shift she sews more than 150-200 t-shirts for Zara, Levis, Walmart and H&M. Her pay for the entire day? Less than $5.

H&M takes those shirts and sells them to European consumers for more than €20 each. That's over €4,000 in retail value. Generated by her hands. In a single day.

H&M runs massive ad campaigns in Europe talking about "conscious" fashion and "sustainability." But there's nothing sustainable about a system that uses their workers up like a cheap commodity.

This is why Bangladeshi workers are shutting down garment factories all over the country. They aren't just asking for a few extra cents. They are striking for the right to be treated like human beings.

#H&M #Walmart #CheapFashion #Bangladesh

@oneiros

Ayanna Pressley would be a great Speaker of the House.

We can’t have a speaker who’s taken $150,000 from Peter Thiel’s Anduril, tens of thousands from executives of Palantir, hundreds of thousands from AIPAC, and secret meeetings with Jared Kushner.

You would have supported Clarence Thomas for the Supreme Court, with that kind of thinking.

https://futurism.com/future-society/hakeem-jeffries-donors-corporate-controversial-tech-companies

People are taking this democracy thing further than the Dem Party is comfortable with.

https://truthout.org/articles/petition-urging-dems-to-reject-jeffries-as-house-speaker-passes-1m-signatures/

District-installed spyware on computers assigned to kids isn’t only offensive, it’s ineffective. Anybody who has written a swear filter or attempted to automate moderation will tell you that it’s a fool’s errand. You can’t automate and surveil your way to compliance with district policy.

What's crazy/depressing/fascinating is seeing a chatgpt conversation someone sent me that they'd had about some software/reverse engineering I'd done and asking for some advice...

*I* am the only source of the information, there is no one else, and yet the chat is confidently chirruping plausible sounding shit....

"well what a lot of people have found is..." ! a lot of people??? there ARE NO OTHER PEOPLE. just me.
"yes it's quite common on these devices for..."??? Quite common??? NO-ONE does this! my blog is THE ONLY SOURCE OF INFORMATION.
"it's not unheard of for some people to have success with..."??? yes.. yes IT *IS* UNHEARD OF.

A bit like those situations where you see news reporting of a story you're familiar with, it's fascinating seeing complete fabrications of information that are structured and explained in a plausible sequential manner.