Old Fucking Punk @lwriemen@librem.one

Stop the surveillance state, I want to get off. https://patreon.com/BrianMcFadden

It’s hard to convey the scale and scope of how cooked a generation of people are by LLM slop.

“You can use AI to improve efficiency but make sure you double-check all its output to make sure it isn’t just hallucinating bullshit” thanks that’s actually more work and thus less efficient than not using an LLM at all.

“You have to double-check LLM output, but also if the task is too complicated or complex to figure out on your own, you should use an LLM” thanks but if it’s too complicated to figure out on your own, then you can’t double-check the output.

Just absolutely cooked.

A woman can be shackled, forced to remove all her clothes and menstrual products, and subjected to an invasive search—yet the government does not consider that sexual abuse.

Khawla Nakua reports on how strip searches are normalized “in the name of security.”

https://inthesetimes.com/article/investigation-prison-abolition-invasive-humiliating-routine-violence-incarceration-women-strip-searches

There's an EU feedback consultation open on the usage of LLMs in creative and cultural spaces. You don't need to be an EU citizen to participate (but you do need to make an account). If I read it correctly, it ends in like 3 hours (I've just been made aware like 10 minutes ago)!

https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/18592-Artificial-intelligence-strategy-for-the-cultural-and-creative-sectors_en

CC: @tante @davidgerard @zzt

#AI #Math

The most recent case of the finite time solution of the forced Navier-Stokes equation has been discussed a lot in the recent days, on this network and in the mainstream press,
but I will try to give a perspective on what happened recently.

This is prompted by a friend informing me this morning of two annoying additional stories that taint the recent public success of AI generated proofs and will hopefully lead the mathematical community to address the question of predation and plagiarism.

However, I won't address here the even more important questions posed by AI companies beyond merely quoting some of them:
- its energetic cost;
- its impact on rare earths and other matters;
- its direct and immediate consequences on the ecosystem (water heating, pollution, land waterproofing, etc.);
- its impact on the work market, in particular those prompted by (actual or alleged) productivity gains;
- its impact on learning, cognition, etc.;
- its impact on the economic structure, tremendous announced gains/losses of companies that look like a Ponzi scheme;
- its impact on the power structure, who owns what, who gives access to what;
- its impact on the political structure, the surveillance capitalism, etc.

These are fundamental issues that should be addressed more clearly in the mainstream medias, and that, in my view, suffice to object to the existence of AI.

Anyway, what does it mean for academic math?

1/n

If being "clearly at odds" with your values is not enough to stop you from using the slopbot, your values aren't worth much, are they now?

"Although using the big AI platforms, as the study notes, is clearly at odds with the solidarity economy’s values, the lack of an alternative means many organizations within the sector are using it anyway..."

The alternative is NOT F'ING USING IT!

https://canadianworker.coop/study-concludes-a-co-operative-alternative-to-big-ai-platforms-is-both-necessary-and-feasible/

If you can’t write code without a particular IDE or LLM, it’s because you don’t know how to write code.

Learn! It’s fun to learn how to code.

🔵 This week's Tom the Dancing Bug
-
These Are the People in Your Neighborhood: This week we meet Mr. ICE Officer
-
Brought to you by the the letters I and H, which stand for theInner Hive, including long-time member Brian F. and new member Wendy the Flee
👉 bit.ly/theInnerHive
-
READ IT RIGHT HERE👇

RE: https://neopaquita.es/@berniethewordsmith/117246249020074543

"Disabled people need AI! Why are you so ableist?!"

#Apple Doesn’t Want You to Worry About the New #Apple Watch's Listening Features

The new #AppleWatch includes several “intelligent” listening features that have #privacy and #security baked in. But the protections can’t change the facts of what the tools do.

https://www.wired.com/story/apple-doesnt-want-you-to-worry-about-the-new-apple-watchs-listening-features/

California is deciding who may verify AI, and one investigation already cost $400,000 in tokens - https://thenextweb.com/news/california-sb-813-independent-verification-organisations-metr-400000-tokens-openai-paid-eu-ai-act-scientific-panel-article-68 "#OpenAI supplied the credits that paid for the inquiry into its own agents, while Europe appointed 60 independent experts in June"

The Problem with Intellectual Property - https://c4sif.org/2026/09/the-problem-with-intellectual-property-springer-2026/ "explores the nature of property rights and argues that IP is incompatible with genuine private property rights, liberty, and justice, and distorts and impedes artistic creation and innovation." #intellectualmonopolies

US government argues that fair use applies to #OpenAI’s training of #LLMs in key #copyright case - https://walledculture.org/us-government-argues-that-fair-use-applies-to-openais-training-of-llms-in-key-copyright-case/

#Forgejo 16.0.4 and 15.0.8 were just released! They are critical security releases.

We recommend that all installations are upgraded to the latest version as soon as possible.

Check out the release notes and download it at https://forgejo.org/releases/. If you experience any issues with this release, please report to https://codeberg.org/forgejo/forgejo/issues.

#forgejorelease

@ambiguous_yelp @Xtreix @rhelune

From what I read, they were talking about doing LLM-based code review, not code authoring. This involves a different set of problems, which are more subtly.

They acknowledged that the LLMs produced a lot of inaccurate and misleading feedback, but claimed that the good reports in the middle made this an acceptable tradeoff. This would be true, if developer time were unbounded.

The problem is, with the available effort, you are never going to be able to eliminate all bugs during code review. Even if you do formal verification, you only avoid bugs that aren't present in your specification and your specification complexity is such that it's pretty much guaranteed to contain bugs.

Even if you ignore the ethical issues surrounding LLMs, introducing a tool with a high false-positive rate into code review comes with a number of problems:

It distracts attention away from addressing feedback from deterministic tooling. Or, if your commits are gated on these, you move time away from fixing backlog bugs to reviewing the output from the LLM to find the things that aren't false positives.
It means human code reviewers will pay less attention, because there are already a lot of comments and some of them will be ones humans want to do.
It causes people to ignore the real bugs that it finds because they're hidden in the middle of false positives. We saw this 20 years earlier with Coverity: it found real bugs, but it also found a lot of not-actually-bug things. We recently found a bug in a certified MISRA C component that was hidden because someone had explicitly silenced the static analyser warning about it because they thought it was a false positive. The more common false positives are, the more likely people are to miss the real ones.

And all of that adds up to a project that doesn't think it has a duty of care to its users. And that's fine, unless you're a project that's telling everyone to switch to you because of security.

@Xtreix @rhelune allowing LLM contributions invites security holes - legitimizes the environment,culture&reality destroying surveillancec machine that is generative AI. And it encourages deskilling of an entire generation of security researchers ultimately placing all the security work (and trust associated with it) in the hands of the richest companies on the planet. I use FLOSS to escape the control of these people.

#LLM #LargeLyingMachine #GenAI #AISlop #VibeCoding #FLOSS #GrapheneOS

@zzt And if your own project is pro AI and pro cryptocurrency the slop pit that is github will still be happy to have you.

grapheneOS is now falsely claiming to be affected by the codeberg ban on cryptocurrency projects because they take donations in monero

of course these shitheads have a problem with codeberg banning slop and cryptocurrency projects, why wouldn’t they

I was just reflecting this morning on the push to integrate LLMs into the software development process. Thanks to LLMs, everything now is much more expensive (devices, energy, food), as well as the code. Every unit of code now costs many multiples of what it used to, and there is more of it.