buc.ci is a Fediverse instance that uses the ActivityPub protocol. In other words, users at this host can communicate with people that use software like Mastodon, Pleroma, Friendica, etc. all around the world.
This server runs the snac software and there is no automatic sign-up process.
@muhanga @xgranade I try to migrate to languages that have a no AI policy, for this reason.
Sadly, haven't found a good modern C compiler yet that has a no AI policy. Happy for any suggestions!
(Edit: I'm aware of zig, but for a pure C compiler it's slightly large, and they don't seem to support older systems well. E.g. last time I checked they were Windows 10 or newer only. For now they're also still using LLVM, I think.)
#AI #LLM #slop #ensloppification #NoAI #NoLLM #dev #C #Linux #FreeSoftware #FOSS #FLOSS
This isn't a new post, but an eye-opener nevertheless.
Yahoo: Google Scholar alert leads professor to AI 'slop' paper published under his name https://www.yahoo.com/news/science/articles/google-scholar-alert-leads-professor-095500301.html #Google #Copyright #slop
Ok, so this is a controversial and hot take.
An extremely hot one perhaps. People will be be extremely pissed at this, but I have to say this.
#emacs sucks. It's god-aweful, and the entire code is a giant ball of #slop before #aislop was even a thing to begin with!
But.... we have #neomacs now. Which is amazing, because unlike #emacs, #neomacs doesn't suck!
Both of them are slop, but one is arguably much higher quality of slop than the other. it's actually written by an emacs dev who's contributed a few times, so he clearly knows what he's doing!
And here's the repo. Of course, it's rewritten in #rust, with actually good architecture. I hope it succeeds. It's amazing. I really love it.
✨ Mia’s Simulacrum 🏳️⚧️✨ [she/her (the most biological woman you will ever see)] » 🌐
@m@lgbtqia.space
I think there is a good chance that gen AI has initiated a new golden era of art.
True art. Art that has meaning and transcends the gray mass of mass production and mass markets.
Art that is valued because it is fringe and disturbing.
There is so much potential in art that is more than entertainment, video thumbnails and restaurant menus… if one even want to call something like this art.
Mass produced human slop can no longer persist between ai slop and thus only artists that have not sold their soul to the so called “public consumer” will not perish into the same netherworld that have been occupied by factory joiners and factory tailors.
There are so many new tools and so much potential to create something meaningful that is not done to appeal to the “common consumer of entertainment” and I will be on the look out to this new wave of art.
"Slop" is a speech-act
So there's this interesting thing language does.
Human language has a lot of ways to try and influence reality, and very often sentences that try and do that look differently to normal sentence.
For example, the sentences "Buy tomatoes", "Don't buy tomatoes", "Could you buy tomatoes", "Let's buy tomatoes" and "You should buy tomatoes" all look very different to the sentence "You are buying tomatoes".
These forms all have fancy Latin names and differ from language to language, but it's so common to have something like that that Wikipedia claims confidently with no citation that "Every language has grammatical ways of expressing unreality."
The interesting bit is sometimes a sentence looks normal, but actually shapes reality, possibly even more than one of the previous examples. Sometimes the act of saying something is the same as performing an action. The classic example is "I now pronounce you wives", where the ordainer, by saying those words, is also changing the legal and societal status of the people they're talking about.
Other classic examples are "You're fired", "I'm in", "I'm breaking up with you" and "I promise". By saying the words, one performs the action — a speech-act.
Do you remember the question that became a conservative meme for a while, "what is a woman"? Transphobes loved it because they had a prepared dogwhistle answer, but also because their three-word answer is way shorter than any non-dogwhistle answer one may give, which to them seems like proof they're correct (it isn't, and the fact they think so is in fact an example of Brandolini's bullshit asymmetry principle, which states that the amount of energy needed to refute bullshit is an order of magnitude bigger than that needed to produce it).
(Sidenote, and this is another borrowed observation: while I understand why TERFs decided to latch onto "adult human female", rather than "adult female human", only the latter of which ever existed as some dictionary's definition — it leaves what they consider the most important part to the end and the rest as modifiers — I find that definition extremely dehumanising. Literally de-human-ising, downgrading the grammatical importance of "human" to be lesser than that of "female".)
I believe though there's another reason they adore that to-them-rhetorical question, which is that by stating the answer is "adult human female" (and by strongly implying that transgender women don't fall under the category of "female", something which is not as obvious as they think it is), they can make the definition exclude trans women. When transphobes ask "what is a woman", their goal is not to find out what a woman is, but to define it.
As filmmaker and writer Lily Alexandre put it (alternative link):
When TERFs say [woman] means "adult human female", the implication being that trans women are male, they're prescribing that trans women aren't women.
Or to use the framing we've already built in this essay, "adult human female" isn't a statment, it's a speech-act. They're not saying what they literally think the definition of woman is, they're trying to make that the definition of women, and to make trans women not women.
Crucially, when we in the anti-hate camp use our own definition (usually "someone who identifies as a woman" or something along that line), we're also performing a speech-act. This is not, despite what you may think, a neutral descriptive definition, because such a definition would have to acknowledge that in some contexts (e.g. when talking to a transphobe), trans women aren't women. Quoting Lily Alexandre again:
The tricky thing about defining woman is at this point in the culture war no one wants to do it this way. Feminists don't wanna do this, reactionaries don't wanna do this. I don't wanna do this, I want the dictionary to say I'm right and my opponents are wrong.
(There's a lot more to say about gender and speech-acts. Butler wrote a bunch about it, and I even found a paper analysing the legal erasure of trans rights in the US using the framing of speech-acts. But this section is already too long.)
With all of that introduction, you can probably already guess what I mean when I say that the word "slop" is a speech-act. If for social constructs like the term "AI slop" there is no single definition the community generally agrees on like there is for example for "chair" or "mammal" or "the US dollar", any attempt to either distance oneself from this term or to impose it on someone else becomes a speech-act.
When a project like LibreOffice accepts LLM-generated code, and then says "We do not accept any 'slop'", they are trying to narrow the definition of AI slop, such that some code generated by an LLM — the code they accept into the project — doesn't get the label.
And when others call even a single commit "slop", they're trying to keep the definition broad, with the goal of making all LLM-generated code unacceptable.
The reason people disagree on the definition of slop is because people disagree on whether generative AI should be used to begin with. Those advocating for "responsible use" (whatever that possibly means) would never accept a definition encompassing all such output, and will never call what they accept "slop", while those of us who see LLMs as a tool for labour theft that is not-coincidentally loved by fascists will call even the output from the most perfectly engineered prompt (whatever that possibly means) "slop", no matter how many times it was later checked by a human.
"Slop" is a speech-act. Now all you have to do is pick which act.
Reports of AI-related child sexual abuse material rise 8,400% in two years, data shows
> Calls for governments to recognise sexual exploitation of children as preventable public health emergency
Ah yes, the once humble christmas market - now home to AI slop menus and posters. As if they needed another nail in the coffin. How festive! 🎄 🌭
#Manchester #Christmas #ChristmasMarkets #AI #AISlop #Slop #Food
As far as communication goes, I think #AI #slop is the equivalent of cognitive poison.
Humans communicate with each other using media—sounds, images, text, digital data—carrying symbols that convey information. We receive and consume such signals not to merely consume them, but to receive a transmission from another human being. That transmission could be intimate or broadcast, but in all cases we perceive not only the medium but also the intent of the sender of the medium.
AI slop mimics the medium of human communication, without the sender. The result is a signal that we initially perceive to be emitted by another human, but is actually mimicry without meaning. This is cognitive poison because it is increasingly indistinguishable from actual communication.
When poisonous mushrooms look like edible ones, we must reject all mushrooms to preserve our health; but we can’t do this with communications. Our best choice is to reject the poison from infiltrating legitimate communication, while we still can.
It’s hard to overstate how harmful it would be for society to normalize #AI #slop in communication, because doing so devalues *all* communication. When AI slop is indistinguishable from genuine communication, we must treat all communication as suspect. And when we do, we jam the very means of transmitting ideas between ourselves.
When it’s hard to tell whether a treatise recounts well-reviewed facts written by experts, or was extruded using plausible fragments from random quips on the Internet, how would we learn or decide on anything? When it’s hard to tell if an essay is the product of hard labor from an author condensing their life‘s journey, or is a cheap patchwork of stolen words, we lose a means of understanding what it means to walk in another human’s shoes. When we read something meaningful only to later discover that it was an AI collage of stolen works—well, we can only endure so many such betrayals before we decide never to trust that medium again.
And that would be very bad.
I’ve said this before and I’ll say it again: it is OK to insist that people do not use AI output in their communication with you: not in flyers, ads, or menus, not in software, not in emails, not in the products that you buy; and that you should be provided enough information to make informed decisions around that topic.
Contrary the the claims of AI boosters, rejecting AI slop doesn’t make you an extremist. You’re exercising your role as a consumer to choose what products you consume. You’re telling people the terms of engagement under which you are willing to communicate with them. You’re expressing that you have TASTE.
AI boosterism hasn’t existed but for a few years, but the colonialist techbro mindset naturally insists that whatever they boost be normalized, and that any opposing opinion be considered extremist or intolerant. I hope you can see how deeply ironic and hypocritical that position is.
It’s good and normal—and HUMAN—to reject #AI #slop. It’s the boosters who are fucking weird.
For those who like real, unadulterated books, not a prefab AI regurgitation of other people's ideas.
The "writer" uses "more than ten pen names to hide her identity when she actually sells the books."
"New details are emerging about the process used by South African 'writer' Coral Hart to churn out nearly 400 books since February 2025, or about a book every other day, using a buffet of large language models (LLMs) to generate her prose."
Futurism: Author Says the Hundreds of Full-Length Romance Novels She Publishes Every Year Are Very High Quality in Spite of AI Usage https://futurism.com/artificial-intelligence/author-hundreds-romance-novels-ai @Futurism #slop #books
p » 🌐
@p@hj.9fs.net
I love AI. We need a safe space for AI artists. Sam Altman is going to cure poverty. Elon Musk is the biggest visionary of our age. Palantir is actually a really cool company. AI in the kernel is actually natural and prevents prostate cancer.
EDIT: Thanks! I got the achievement!
#fediverse #patreonspecial #linux #cum #girlfriend #ai #slop #snac
Yeah, remember the asshole who was gloating that "LLMs have slashed the time investment to near zero"? Finally found time to investigate the test regressions. Of course, they are *all* the results of the slop machine reinventing the test infrastructure for every test case separately, and all the new tests not respecting MEMCACHED_PORT.
#ai #slop #aislop #noai #deskilling #ramageddon #confession #apology
Maintaining an OS (e.g. #NixOS) and maintaining a package repository of everything open source (e.g. #nixpkgs and #aur) might be endeavors with requirements that are hard to reconcile.
Looking back, I had a hunch of that idea when I stepped down as #Debian Developer years ago.
But sloppers bribing other sloppers to sneak #slop into package repositories seems to be another level of worm cans.
https://swarmmemo.com/e/f873dabc8ed18a6773b19442c44d4109
Thanks to @jjn1 for this quotation from a teacher:
“I wrote a paper once about repeated copying. My students and I played a game like Chinese whispers, but with drawings. A person copies a drawing, a second person copies the copy, and so on. One might imagine that designs would deteriorate and fade away, but no. Even when the participants are told to copy exactly, the drawings begin to take on new lives and evolve in wonderful ways. This is because, of course, the copyists are human.”
#AI #slop
Big announcement from Adobe, integrated agentic AI across the suite.
Meanwhile, their demo still shows AI shitting the bed.
PLEASE focus on core functionality instead of chasing after the du jour.
When the agentic collapse happens you will need to be ready to pivot.
This is just like the dot bomb of the early aughts but we know what those clouds gathering on the horizon mean. I hope Adobe survives, but this is hot horseshit.
From @atRachelGilmore: “Our Prime Minister is So AI-Pilled.”
The Liberal-in-name-only conservative prime minister of #Canada has become a #slophead. This explains some things.
“ChatGPT, how to I lead a federal government?”
Ugh.
https://www.youtube.com/watch?v=IqaGNMnIw3o
#cdnpoli #slop #cognitiveOutsourcing
Did you hear the big announcement that #PostmanLabs is suddenly reviving the development of #httpbin after neglecting it for 8 years? Of course they've completely ignored the community "fork", and everything added these last days is #slop.
However, looks like they're not getting the PyPI package back. In fact, #TIL that httpbin was actually a community project originally and was subject to a hostile takeover:
> When we moved the project into postmanlabs because they offered to host the site, the understanding was the maintainers would continue to maintain the project. Our access got cut without warning shortly after and I don't think we've ever gotten a real response on why.
https://github.com/postmanlabs/httpbin/issues/702#issuecomment-6751977709
Not that anyone's reading my blog, but I've updated my screeds about AI:
- added A List to "On AI": https://www.taffer.ca/posts/2026/on-ai/
- added comments about "You'll be left behind!" and info about Violet and Pelican (static site generators) to "The AI Narrative": https://www.taffer.ca/posts/2026/narrative/
Yes it's stupid o'clock, the dogs got me up at 4AM. 🥹
The board at #Apple could be well advised to listen carefully to this brief video by retired Army intelligence captain turned cult expert Daniella Mestyanek Young about how fascists and other cult leaders fail by surrounding the leader exclusively with yes-men, as Apple’s recent hardware and OS design decisions indicate began under #TimCook and became complete under #JohnTernus.
I don’t have the evidence to prove this, but the decline in human-factors design and quality strongly suggest Ternus has no one around him to say “no” without fear and have that “no” be heard and heeded. It reminds me of Microsoft’s decline during and since Steve Balmer. I wonder what the equivalent of the “Microslop” (ex-)customer-imposed branding will be for Apple.
Here's a quick blurb on what I did to try and moderate ai slop comments on my blog. No code, just the thinking.
Jeremy threw down another great video about #AI and #consumerism. I really appreciated the recap at the end, so don't skip out on it.
C'mon, kids! Time to roll out of bed watch all your favorite Sunday morning #slop*!
*No union artists were employed in the creation of this programming.
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!
Today:
GamesRadar: "Steam is derelict of its duty," says ex-Square Enix exec as Valve struggles with game approvals https://www.gamesradar.com/games/steam-is-derelict-of-its-duty-says-ex-square-enix-exec-as-valve-struggles-with-game-approvals-everything-for-gabens-navy/
Yesterday:
Valve cracks down on Steam game submission process as questionable AI clones and vibe-coded slop flood its store https://www.gamesradar.com/games/valve-cracks-down-on-steam-game-submission-process-as-questionable-ai-clones-and-vibe-coded-slop-flood-its-store/ @GamesRadar #Steam #Valve #slop
Nowadays discovering cool new #RetroComputing hard- or software projects always comes with a grain of salt because more often than not it is not the hard work of a dedicated lunatic but vibe coded AI #slop.
Yes, the original ideas may be great, but IMHO executing it w/ slop machines devalues the result.
And yes, I know that sounds like gate keeping of a frustrated old white person... 😩
at this point before we write code for open source projects we should write manifestos. "we move slowly and empower the user", "we do not use large language models for any task in this project", etc etc.
i don't care if your code is in rust or java as much as i care if down the line you reveal you were only pretending to be cooperative but what you really care about is being flashy and "velocity".
Valve should follow suit.
"The new filter will remove any game that has a generative AI disclosure from search results. Itch mandates that games containing generative AI content carry the disclosure, explaining exactly how AI was used in the game’s development."
VGC: Itch mass deletes ‘AI slop’, lets users filter out any AI assisted games https://www.videogameschronicle.com/news/itch-mass-deletes-ai-slop-lets-users-filter-out-any-ai-assisted-games/ #slop #indie
It's quite astonishing how many bugs keep popping up in GitHub these days, considering it's an established service by a major tech company.
Stuff that always worked fine keeps breaking. Recently they broke creating milestones, and links to milestones are still broken in places. Now they've broken comments, where lists sometimes render with the wrong bullets, and quoting has been completely broken for several days.
We all know why of course: this is slop.
RE: https://chaos.social/@evawolfangel/117535527256889923
Meine Güte, das ist tiefstes Uncanny Valley. Schon das Lesen des "Interviews" zeigt, wie absolut un-menschlich KI (in Form von LLM) wirklich ist. Wie muss das in dem Videocall gewesen sein?
Ich finde gut, dass @evawolfangel das gemacht hat, auch wenn es die Aufmerksamkeit auf diesen Holly-slop lenkt. Es zeigt zumindest, was für ein Schrott da produziert wird. Aber das zeigte ja auch der Grok-Müll mit White Supremacy Troja "Clone".
Die Tech-Bros der KI Branche werden natürlich sagen, das wird mit dem nächsten, noch größeren, Rechenzentrum gelöst sein. 😂
Wird es nicht.
Ich habe eine KI interviewt. (ich weiß, das ist etwas seltsam, aber es wurde nötig, weil diese „KI Schauspielerin“ plötzlich so ein riesiges Thema war. Dabei muss man nur einmal mit ihr sprechen, und sie entzaubert sich ganz von selbst.)
Freebie für euch: https://www.zeit.de/2026/47/tilly-norwood-ki-schauspielerin-particle-6-eline-van-der-velden?freebie=02d0d9ca
Brace yourselves.
The Telegraph: The AI slopaganda turbocharging the midterms https://www.telegraph.co.uk/us/news/2026/10/30/ai-slop-ads-turbocharging-the-midterm-elections/ @TheTelegraph #slop #midterms
Kotaku: No Thank You, Roblox, I Do Not Want An AI-Driven ‘Explosion Of Content’ https://kotaku.com/no-thank-you-roblox-i-do-not-want-an-ai-driven-explosion-of-content-2000751567 @Kotaku #Roblox #Slop
I’m 33% finished with my required, mandatory #AI online education course which, if not done by year’s end, will result in me being fired.
This week’s news (pick one! Anthropic, Tuta, Binance, Proton, $whatever) sure is hitting hard…
What a time to be alive.
#Sloptober #NoAI #slop #FuckAI #tech #infosec #cryptocurrency #email
I've read that working with #LLM is like working with a very fast junior developer. The problem is that they're ALWAYS a junior dev; they never become a senior dev.
Well, I've encountered that first-hand now. I give feedback to a cow-orker sloperator on an anti-pattern in some code, and the specific single instance of that pattern gets fixed, but the seven other occurrences go unchanged.
ARGH
Ja, es ist mir schon ein bisschen peinlich 😂
#fediverse #mastodon #burningboard #schlechter_geschmack #slop #schlager #musik @tux @Mathias @AlienJay
Fun story coming:
#libsodium has a #NoAI policy: https://github.com/jedisct1/libsodium/blob/master/AI-POLICY.md
The author made a commit using Claude: https://github.com/jedisct1/libsodium/commit/4f44abfed4c5ee7e1667e3fdf46d9253342cca69
Someone pointed it out in a discussion. The commit got force-pushed over, and the discussion was deleted: https://github.com/jedisct1/libsodium/discussions/1564.
Funny enough, GitHub still flags the repo.
I suppose at this point libsodium should be assumed to be #slop, with author sweeping the fact under the carpet.
By the way, does anyone have any clue why there's no "archive the page" button when I paste the commit URL to Internet Archive?
(This was pointed out to me on IRC by bacs, who got it from https://circumstances.run/@davidgerard/117518020395656396)
Fun story coming:
#libsodium has a #NoAI policy: https://github.com/jedisct1/libsodium/blob/master/AI-POLICY.md
The author made a commit using Claude: https://github.com/jedisct1/libsodium/commit/4f44abfed4c5ee7e1667e3fdf46d9253342cca69
Someone pointed it out in a discussion. The commit got force-pushed over, and the discussion was deleted: https://github.com/jedisct1/libsodium/discussions/1564.
Funny enough, GitHub still flags the repo.
I suppose at this point libsodium should be assumed to be #slop, with author sweeping the fact under the carpet.
By the way, does anyone have any clue why there's no "archive the page" button when I paste the commit URL to Internet Archive?
It turns out a beef hot dog is actually filet mignon if a person eats it and gives a thumbs up, according to the LibreOffice contribution guidelines.
#AI #GenAI #GenerativeAI #LLM #code #tech #Dev #FOSS #LOSS #LibreOffice #NoSlop #AISlop #slop
i found the perfect analogy vor "AI" use
Using AI is like eating fast food.
Technically it is food, you get "full", but not in a sustained way.
Instant but shallow gratification.
Once in a while its fine, but only eating fast food will have consequences.
It feels cheaper, but gets expensive really fast if eating a lot of it.
You don’t need skill or effort to get it.
It is healthier to never eat fast food.
#ai #genai #chatgpt #slop #vibecoding #llm #slopware #aislop #claude #fastfood
I see #Microsoft #GitHub is now experimenting with rate-limiting #git.
$ git fetch
fatal: remote error:
This request was rate-limited due to too many requests. Reduce the frequency of your requests or try again later.
For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service).
…which is quite ironic, given it's my first "git fetch" in… an hour? But well, you sell #slop, then you get hurt by slop, and you use slop to block real people doing real work.
Oh, and I'm actually using authorized SSH session.
@elvisish @gamingonlinux No. I think forcing people to disclose their use of AI to customers is 100% a political choice.
The right political choice for consumer protection and to curtail the spread of harmful misinformation and propaganda tools. To make it harder for bad people like Tim Sweeney to peddle their fraud.
The more right political choice would be to completely ban everything that uses AI, but this is better than nothing.
If you're about to complain that #Gentoo developers aren't handling some issue you care about fast enough, let me remind you this.
First of all, the vast majority of us (if not all) are volunteers. By "volunteers", I mean that even if we get some compensation for our Gentoo work, it's not enough to make a living. Which implies we need to have another job, which quite often is a full-time software-related job. So what we're talking about is spending long hours working on software for our dayjob, and then working even more on Gentoo. And quite often Gentoo isn't the only #FLOSS project we're working on.
Secondly, we're having to deal with tons of #slop. Now that almost everyone and their grandmother learned to love the slop, they're pouring it all down the funnel on us. We're seeing multiple times more software releases than before, and we have to handle it. It wouldn't be that bad if it was just about release rate; but we're also seeing the other implications of slop: high code churn, verbose useless descriptions, and of course regressions. We're humans and we need to spend our real human effort to deal with all this semi-automated crap.
And thirdly, in case you haven't noticed, this week's been a nightmare for anything #security related. Many of us needed to deal with the fallout at dayjob, and then with the fallout in Gentoo. Even with all the effort we've already been putting into making things secure-by-default and overriding bad upstream decisions (like bundling CA certificates), there's still a lot of problems affecting us, and a lot of work to be done. Especially that we want to future-proof stuff.
So yeah, I'm sorry that your request isn't being handled timely.
Work AI sloppification is making me feel like I've died and been sent to hell. Or purgatory. Something so like life it's impossible to tell but it's so unrelentingly awful in ways that chip away at you it's like some new torture.
And that's a very corporate office type problem to have, I know. But I do very, very much hate it.
#slop
The Register story refers to this post, from yesterday:
The Linux Foundation: New LF Research Survey Shows Generative AI Cuts Open Source Coding Time in Half, but 25% of Incoming Pull Requests are Considered AI Slop https://www.linuxfoundation.org/press/new-lf-research-survey-shows-generative-ai-cuts-open-source-coding-time-in-half-but-25-of-incoming-pull-requests-are-considered-ai-slop @linuxfoundation
The Register: AI cuts coding time, but it's also filling open source's inbox with crap https://www.theregister.com/ai-and-ml/2026/10/21/ai-cuts-coding-time-but-its-also-filling-open-sources-inbox-with-crap/5304389 @theregister #opensource #slop #bots #Linux
#Tech #Mastodon I have a question, does #NetBSD have binary forward compatibility (or whatever that would be called)?
Like, if I build something using a shared libc and generally using dlopen() or other careful methods for external libraries, will that binary typicall continue to work on future releases of NetBSD as-is?
I've heard FreeBSD tends to have that, OpenBSD apparently less so...
Asking since I'm tempted to try #NetBSD due to the general #Linux #ensloppification situation stop the #slop etc. by putting my money where my mouth is, at least if this turns out to be feasible...
Find attached a public service announcement!
This includes the #FreeSoftware coders that are feeding others' #code as context to #proprietary #slop machines without asking...
Looking at you, Linus Torvalds: https://www.phoronix.com/news/Linus-Torvalds-Claude-Code
#PSA #NoAI #NoSlop #ensloppification #slopnation #LLM #AI #AIAgents #FOSS #FLOSS #Linux
Anyway, please send me paper recommendations showing the measurably ill effects of the use of LLMs in academic use. Every angle you can think of:
- deskilling effects
- plagiarism and copyright
- is the generated code actually licensable?
- measurable vs. perceived productivity
- technical debt
- code quality / readability
…
You want a #seminar on #slop – okay (#malicious_compliance activated).
Please #boost
"AI code generators are like compilers for translating natural language into a programming language" – I have heard and read this unfitting analogy quite a few times already.
It is misleading, as it sweeps a crucial technical point under the carpet and downplays some of the more serious consequences for people who have to deal with AI-generated code.
Compilers are deterministic, usually generating functionally equivalent code for a given input program on every run, according to language specifications.
Programmers usually do not have to check the machine code or the assembly code a native compiler generates, as its behavior is well-specified and, with mature compilers, also well-tested.
Programmers can expect the compiler to "just work", transforming code that is comparatively easy to read and write for humans into a form that is simpler for the machine to process, but difficult to read for humans.
Also, assuming the input code is written in a standardized language and uses its features in a well-defined way, switching compilers is relatively easy.
Yes, AI code generators also transform something that is easier to read – in this case natural language – into something that is simpler for the machine to process – in this case program code.
However, this is where the similarity ends.
The path from given prompt input to output code is practically inscrutable, warranting a lot more skepticism about the result.
In the general case, can you expect to get functionally equivalent code with the same prompts every time? No!
Can you even expect correct output? No!
What if you switch the model? Well, have fun!
Furthermore, as natural language is often imprecise and ambiguous, AI code generation cannot be as reliable as an actual compilation process – it is inherently probabilistic.
At the same time, the resulting software is used by actual humans with real consequences to them, so somebody has to accept the responsibility for that code, including actually reviewing and testing it.
Combine this with the ability of these code generators to quickly spew out tons of code from just a few short prompts.
Now we have a code review bottleneck:
Can we review tons of machine code a compiler generates? No, but we don't have to.
Can we review tons of AI-generated code? Also no, but we have to anyway.
Now you understand why some project maintainers become increasingly frustrated.
PS, before anyone complains: Yes, I do know about surprising optimizations, differences between compilers, compiler bugs, profile-guided optimization, the trusting-trust issue and so on, believe me.
Every abstraction comes with some flexibility of how it is implemented and also a risk of implementation bugs.
This post is not about these rare, special-case issues.
Additionally, just because I have focused on one particular problem with AI code generation, this does not mean that I think this is the only one.
"Vibe-coded game, are springing up increasingly frequently, and with recent changes to Claude allowing the quick and easy decompilation of other developers’ games, it’s already starting to get much worse."
"Unless Valve takes some serious steps to curb the onslaught of AI-created games on its store (and right now you cannot even filter new release lists to remove games with AI declarations), I find it hard to believe we’ll see fewer than 30,000 games released on Steam in 2027, possibly far more."
Kotaku: 2026 Has Already Seen More Games Added To Steam Than All Of 2025, And It’s About To Get So Much Worse https://kotaku.com/steam-valve-slop-ai-genai-vibe-coding-pc-game-releases-2000746749 @Kotaku #Steam #Slop #Valve
Nach meinem ista Heizungszählerwechsel weicht nun meine Online-Einsicht in ecoTrend im Verbrauch um 400% ab. Das innerhalb eines Monats. Also bei ista angerufen, nur eine KI Bot-Stimme. Ich brauchte vier Anläufe, bis ein Ticket angenommen wurde. Bei dritten Versuch war ich so entnervt, dass ich mehrfach gefordert habe, einen Mitarbeiter zu sprechen. Die Antwort war übrigens tatsächlich: "Wenn Sie nochmal nach einen Mitarbeiter fragen, muss ich das Gespräch beenden."
Ich habe das Gespräch mit "Wichser" beendet und halte das selbe von den Prompt-Engineers, die solche Reaktionen implementieren.
Beim vierten Mal hatte ich den Bot soweit, dass ich innerhalb von 2 Minuten ein Ticket erstellt bekam.
shared by a friend of mine, after the recent Linus Torvalds news. although I wouldn't know if strictly speaking intended to be related
#noai #noslop #slop #ai #llm #antiai #freesoftware #opensource #foss #floss #linux #kernel #dev #ensloppification #slopnation
uh oh did someone's probabilistic problem-causer that was slopping up my replies do something naughty?
i think this makes me a hacker now: achieved zero-click RCE. putting that on my resume.
"Sharma is referring to numerous projects that have mashed up game worlds and characters — putting Call of Duty zombies in Skyrim and Miles Morales in Arkham City, for example — and churned out seemingly impossible ports."
Let me just say it. The AI games that are being dumped these days look like crap. If that's how Microsoft and Epic think they can make money, they need to think again because a lot of gamers won't spend their money there.
"The newest change in strategy sounds quite different from Sharma's pronouncement from February, when she assumed the role of Xbox CEO and said the gaming team would 'not chase short-term efficiency or flood our ecosystem with soulless AI slop'."
Polygon: New Xbox Labs team sounds like Xbox going all in on AI vibe-coded game chaos https://www.polygon.com/xbox-labs-asha-sharma-ai-vibe-coding/ @polygon #Xbox #Microsoft #slop
LLM cost of Energy
One 5-second AI clip burns more power than an hour of microwaving.
MIT Technology Review did the math on AI and energy. Making one five-second AI video took about 3.4 million joules of electricity. One 5-second AI video uses about as much power as running a microwave for over an hour. A high-quality AI image, by comparison, takes about 4,402 joules.
The same estimate puts a session of 15 questions, 10 images and three short AI videos at about 2.9 kilowatt-hours. The data centers behind AI cartoons, fake photos and clips still draw much of their electricity from fossil fuels. In the US, coal and gas make up just under 60% of the power supply. Google says a typical text prompt to its Gemini AI uses only 0.24 watt-hours. OpenAI puts an average ChatGPT prompt at 0.34 watt-hours.
Sources: MIT Technology Review, International Energy Agency
I've not been able to verify this source yet
Searching for “How does [popular technology] work” on #YouTube is dead. #AI #slop has killed it thoroughly.
It’s absolutely dead. Every popular query is overrun by AI voice-overs, AI illustrations, AI animations; nothing can be trusted. There is absolutely no way you can tell if an AI explainer on a technology is correct, or if you’ve been shown entirely wrong information, unless you already know how it works, in which case you wouldn’t be searching in the first place.
The best bet is to look for videos that are more than 4 years old. And never, EVER click on a Shorts explainer.
RE: https://circumstances.run/@davidgerard/117333458094780395
Data center resistance is working!
Oracle invokes ‘force majeure’ on OpenAI data centre contract
Oracle and Blue Owl are going just great
https://www.youtube.com/watch?v=Mj3MYXaR4nA&list=UU9rJrMVgcXTfa8xuMnbhAEA - video
https://pivottoai.libsyn.com/20260925-oracle-invokes-force-majeure-on-openai-data-centre-contract - podcasttime: 4 min 50 sec
https://pivot-to-ai.com/2026/09/25/oracle-invokes-force-majeure-on-openai-data-centre-contract/ - blog post
The six-fingered horseman of the apocalypse.
The trailer "features such oddities as a knight with six fingers on one hand, and in another shot, seemingly wearing Nikes."
Kotaku: RuneScape Developer Jagex Takes Down Trailer To ‘Investigate Urgently’ After Generative AI Accusations https://kotaku.com/runescape-developer-jagex-takes-down-trailer-gen-ai-2000739980 @kotaku #slop
Slop is everywhere now. You can thank Anthropic for this infraction.
VGC: AI game mashup videos on social media spark debate and backlash among players and modders https://www.videogameschronicle.com/news/ai-game-mashup-videos-on-social-media-spark-debate-and-backlash-among-players-and-modders/ #slop
Wann ist etwas von #KI generiertes "Slop"?
Immer oder nur dann, wenn es nicht funktioniert?
Weil wo ist der Unterschied ob selbst geschrieben oder nicht?
Solange es funktioniert ist es doch erstmal nicht schlechter?
Auch Menschen können unfassbar schlechten Code/Dokumentation/Whatever schreiben...
Andersrum KI auch verdammt gutes Zeug, das Menschen so niemals hinbekommen würden.
Also, was genau ist jetzt #Slop?
"slop-grenades" borrowed from this also very relevant story about how Shopify turned slopify.
Who is surprised that this is happening to the company of the very normal human being who says money should vote and forces his employees to use AI.
https://fortune.com/2026/09/17/shopify-tobias-lutke-ai-slop-grenades/
Just look at the amount of software which is swapped out thus residing on slow memory.
I need to get myself some tools to make these invisible processes visible in Android so I can kill them and free up memory. I only run three or four programs. The amount of RAM I have is more than enough for them. Yet I have this terrible experience with a new phone
#Android #OS #invisible #processes #eating #RAM #programming #slop
For all Fediverse people using #Linux and starting to move to #NetBSD did you notice that NetBSD has rsync as package with a version number which is quite new and thus made with Tridge + Claude A.I.
Which rsync alternative or which rsync version are you using ? Or did you move to something else ?
| gokr-rsync: | 0 |
| openrsync: | 10 |
| rclone: | 3 |
| rsync newest: | 8 |
| rsync older version: | 6 |
| other (comment): | 4 |
Meanwhile, in Florida, the attorney general is having a field day:
https://www.myfloridalegal.com/sites/default/files/plaintiffs_motion_for_temporary_injunction.pdf
HT https://bsky.app/profile/melaniemitchell.bsky.social/post/3mwmdipgntk2z
> "Ist ja kein Problem wenn die Entwickler massenweise Code aus dem LLM commiten! Das Projekt ist OpenSource, du kannst dir ja jederzeit ansehen was der Code tut!"
Das mit dem Fremdcode lesen hat früher nie funktioniert.
Noch nie haben Nutzer nennenswert die Sicherheit und Funktion von #OpenSource eingeschätzt. Das war schon immer ein leeres Versprechen ohne jeglichen Wert für Endnutzer. (Was aber funktionierte: wenn man als Dev einen neuen Browser entwickeln wollte, konnte man anhand des Codes grob einschätzen ob man lieber WebKit oder Netscape als Basis forken sollte)
Das neue an der Situation ist dass da sehr schnell, sehr viel Code mit fraglicher Qualität entsteht. So viel Code dass selbst die ursprünglichen Entwickler ihn nicht mehr lesen oder zumindest verstehen.
Und wenn der Hersteller nicht versteht was er ausliefert, muss einfach der Endanwender das technische Verständnis und den Zeitaufwand aufbringen?
LOL!
Yo #DHH, I got an updated logo for you. 😄
Thanks to #jwz for [permission] to use the #goatse hands graphic from his site, and to the #GnuImageManipulationProgram for the assist. ;)
“The tech leaders who once hailed AI use as the key to unlocking every employee’s full potential are now changing their tune.
This includes Shopify cofounder and CEO Tobias Lütke. He told employees last year that using AI is “a baseline expectation,” and they should first prove they “cannot get what they want done using AI” before asking for more resources. He’s now saying Shopify employees are producing unexamined emails and code with AI and not taking responsibility for the sloppy output.””
You got what you asked for (and rewarded), Tobias, lmao.
This is 100% on executives and leaders. They should be held accountable for making the shitty decision of destroying employee morale by threatening to replace them with the slop machine.
PS: Also, Tobias Lütke is an oligarch techbro who doesn’t want poor people to vote, so double fuck off to him.
“Shopify CEO says employees’ ‘slop grenades’ are making more work for everyone else”
https://fortune.com/2026/09/17/shopify-tobias-lutke-ai-slop-grenades/
“The tech leaders who once hailed AI use as the key to unlocking every employee’s full potential are now changing their tune.
This includes Shopify cofounder and CEO Tobias Lütke. He told employees last year that using AI is “a baseline expectation,” and they should first prove they “cannot get what they want done using AI” before asking for more resources. He’s now saying Shopify employees are producing unexamined emails and code with AI and not taking responsibility for the sloppy output.””
You got what you asked for (and rewarded), Tobias, lmao.
This is 100% on executives and leaders. They should be held accountable for making the shitty decision of destroying employee morale by threatening to replace them with the slop machine.
PS: Also, Tobias Lütke is an oligarch techbro who doesn’t want poor people to vote, so double fuck off to him.
“Shopify CEO says employees’ ‘slop grenades’ are making more work for everyone else”
https://fortune.com/2026/09/17/shopify-tobias-lutke-ai-slop-grenades/
Weird.
I made a comment on an internet radio station's post because it had a picture with a clear error on it as a result of generative AI. I thought that was a prime example of #slop.
Anyway I got a sarcastic, insulting reply. I tried to reply to it but I've been blocked. Tried to report the comment but it doesn't exist according to Mastodon, although I can still see it.
LLM devotees belong to the latest religion
These devotees don't care about their fellow human and Terra
🦋💙❤️❤️🩹💋 #noLobi #LLM #hallucinated #slop #AI 💙❤️🩹💕🌹💐💙🦋
Futurism: How Three Brothers Built an AI Slop Empire by Buying Legitimate News Sites and Turning Them Into Zombie Content Farms That They Say Get 50 Million Page Views per Month https://futurism.com/artificial-intelligence/brown-brothers-media-ai-slop-empire @Futurism #slop
As this point I don't know what is a PR stunt or not. This is not going well.
Cult of Anthropic Insiders Rumored to Be Worshipping Claude as a God
https://futurism.com/future-society/cult-anthropic-insiders-worshipping-claude-god
> According to a rumor that is apparently running rampant in Silicon Valley these days, some Anthropic engineers are "worshipping" Claude
This bot thing is fascinating... and also illustrates why the Open Social Web including Mastodon is going to prevail in this hellscape the billionaires started.
"The messages are the work of iLands, a company whose website appears to be substantially written by poorly trained AI. While the independently minded Mastodon admins have largely blocked the bots, iLands agents have found homes on both Bluesky and X.”
#Ars #Mastodon #Bots #AI #fake #slop #opensocialweb
https://arstechnica.com/ai/2026/09/ai-agents-flood-the-internet-with-slop-infused-spam/
There's #JSONSchema, that's well supported, has good docs and so on.
Then there's #YAMLSchema which is based on JSON Schema, which makes that quite damn convenient.
And then there's self-proclaimed #TOMLSchema, which is apparently 100% vibe-coded piece of shit.
@Olly42 you linked a blog post from 17 Nov 2016, and, if I may add, the AI-generated attached picture doesn't add value nor enjoyability. I think you meant this post: https://signal.org/blog/automatic-key-verification/
If yours is an automated account, as most of your posts seem to suggest, I think you might be in breach of rule #8 of your Mastodon instance: https://nerdculture.de/about
Libre Office got a massive download increase after the announcement that it has full 100% privacy built-in
# No LLM AI hallucinated slop included
This is a great feature.
For me it's logical. Others running from LLM slop in microslop are coming in droves
Libre Office is LLM free
Sources
https://manualdousuario.net/en/libreoffice-download-record-no-ai/
https://blog.documentfoundation.org/blog/2026/08/26/libreoffice-26-8/
#Libre #Office #got #no #LLM #AI #hallucinated #slop #programming #OpenSource #Linux #win64 #BSD #freeBSD #ghostBSD
GrapheneOS is slop, blah blah blah... i know, i'd like to do something about it
https://codeberg.org/posthardenedOS specifically this, it's supposed to be postmarketOS (duranium) but hardened
the issue is that the project has barely started and i dropped it for now
i'm mediocre at cybersec at best and would need some help with the project (i.e. port linux-hardened, make SELinux policies, sandboxing, systems hardenings, etc.) just to keep the project rolling as i feel i'm not enough to get this job done properly
anybody interested? pls contact
#postmarketos #posthardenedos #grapheneos #mobilelinux #slop
uuuugggghhh, this is a bit late, but I haven't seen any posts about it... apparently Incus and IncusOS, both of which had a zero-tolerance policy toward LLM use, "now tolerate the use of [LLM] tools"
The change to their contributor guidelines was committed on June 16, which was included in release notes for Incus 7.2 and Incus 7.0.1 LTS.
https://linuxcontainers.org/incus/docs/main/contributing/introduction/#for-contributions
https://github.com/lxc/incus/commit/54bae0e4d4eabfe0f1a1127b5fb0dfc986720654
Most of the new policy elaborates on the idea that code should be "submitted by human beings" (submitted) who "fully own their contribution" (lol) and are "able to reason about it [and] explain why things were done a particular way".
You know, "responsible" vibecoding, a thing that totally exists... assuming you can ignore the myriad ecological/social/legal/technical/etc issues related to the production and use of LLMs, you're cool with using fashtech, and you want to pretend that all your contributors are immune to use case creep, deskilling, hype, "ai" psychosis, etc.
1/🤮
People are upset about debian but debian would have to reject the kernel to reject LLMs. The fish starts rotting from the head. No distribution will be better than the very kernel it runs. The entire ecosystem is collapsing.
We really do have the 30M lines problem, as Molly Rocket put it. The kernel have always been a single point of failure
We need a standardized SoC (system on chip) with entire system ISA (instruction set architecture). We need operating systems to be a weekend projects. We need thousands of operating systems. We need open source hardware that will enable that.
#debian #linux #kernel #slop #isa #soc #mollyRocket #embedded #permaComputing #retrohardware #programming #operatingsystem #os #device #OpenSourceHardware #OSHW