"I suggest the social action: shame. Shame these people. Their actions are shameful. AI bros hate being shamed and they try to make rules against shaming — because shame works.

None of these people doing shameful things deserve a place in polite company. They should be shamed for their shameful choices.

The AI industry is also running a PR push to redefine common word usage. The word “slop” is also super effective — so the bros really hate “slop” and try to define it to mean slop they don’t like.

But all AI output is slop. It’s social spam. That’s what it’s for. It’s slop! Say “slop” every time."

pivot-to-ai.com/2026/10/14/linux-kernel-goes-full-ai-slop-code-hides-chatbot-commits/

in reply to tante
@hbons
Alright, I'm in agreement about casting shame on the guilty and choosing the words with traditional or dictionary meanings. Would you say we should keep the word slop for the new meaning and henceforth call the messy meat sandwiches Intelligent-Joes?
in reply to tante
@hbons I love this and am stealing it. “Punch as high as you can”. Yes that way you hit the tallest nazi ideal or just the nearest fascist you can reach. We are all fighting at the bottom, yelling at the top most days, so this is nice. We need to aim carefully. Unlike the current US department of war… we don’t want to harm our fellow earthlings.
in reply to Hylke Bons 🥜

@hbons
I very much view slop flyers for restaurants through a different lens based on the price of the food and how many locations they have.

The cheap take out on the corner: whatever, I might complain a bit if I think they'll listen but unless there is a comparable one without a slop flyer nearby I'm not actually changing anything

The local chain with 10 locations around the city: actively looking for alternatives

The global chain: probably already boycotting them for other reasons

in reply to chloe
@hbons
And no 55 isn't exaggerating this is in Archbald Pennsylvania and there are at least 76 data center proposals and 4 proposed full size gas plants and hundreds of diesel generators being pushed in my county. Just in my county. 55 in one town.
in reply to mx alex tax1a - 2020 (6)
@atax1a @peteriskrisjanis
Because not all vibecoded changes are slop. You can't call it slop just because it's made with IA. You can find a lot of slop code not made with IA at all.
It is slop when the code is not properly reviewed or tested.
Esta entrada fue editada (viernes, 16 de octubre de 2026, 13:22)
in reply to jo the disgraced
@jo
There are many rational reasons to be anti-AI, trained on stolen data, driving climate change, techbros being assholes aso.
But saying: "- No matter how good code it produces, it is still slop!!!!" , that is just going to make you look quite irrational.
in reply to Mattias Eriksson ⛵🎸🚵
@snaggen @jo Nope. It is all slop. Slop is a recycled mix of things of unknown provenance that have some superficial resemblance of what was actually wanted or needed.
in reply to Cassandrich

@snaggen @jo "No matter how good code it produces---"

No, stop right there. It does not *produce* code. It *re*produces it. And mixes it to obfuscate the source and launder copyright, the same way a crypto mixer launders money.

So what you are actually saying is:

"No matter how good the code I get to copyright-launder and reuse without attribution or following the licensing terms..."

And that is pretty damning.

in reply to Mattias Eriksson ⛵🎸🚵

@snaggen @jo
"Code generated by genAI is slop" is a tautology, that is the definition.

Regular developers just want to write code and use tools but there is a legit war over the future of the value of human life and labor and they are in the middle, whether they want to be or not.

The nuance of asking "yes, but what if it only does 'X'. Is that still slop?" is getting into the opposing ideologies. One reason to ask this is so someone can produce code while using genAI while still technically complying with whatever the "No AI" policy is. GenAI zealots will use this as a wedge to creep over the line, get buy-in that it's actually okay just this once, and then keep doing that until the "No AI" policy is meaningless.

The Epstein class, fossil fuel billionaires, and the effective altruist broligarchs have joined forces to deskill us, manipulate us, devalue our creativity, surveil us, and even murder us when it serves their ends. Generative AI is the largest, most effective tool they've ever had. Choosing to use genAI benefits them more than it benefits you.

in reply to Mattias Eriksson ⛵🎸🚵

@snaggen @jo Even if it is irrational, it’s not nearly as irrational as being convinced (and trying to convince everyone else) that the slop generator is somehow alive, has feelings and morals, and is likely to destroy humanity.

And yet, those people got to spend ONE TRILLION DOLLARS on their slop machines.

I’ll stop being irrational when they do. Or at least when they’ll stop bleeding us all.

in reply to Mattias Eriksson ⛵🎸🚵

@snaggen @jo I would say more that there should be a separation, whether something works at technical level and whether it's a morally acceptable thing to do...in a perfect world...

That said, irrational expression could still be effective.

Somehow though I don't feel the shame even while accepting many of the moral tenets, so I'd guess not effective for people that don't.

I think based on the Debian vote, most people have accepted this is the new reality that may get toned down a bit after the economic bubble bursts. Obviously there are many communities that are anti-llm too, like codeberg, so I may be wrong here, though OS communities do immediately see the harm of slop submissions from people, who do not even try out, what they submit etc. so they'd expected to be more anti-llm than norm YMMV.

in reply to Mattias Eriksson ⛵🎸🚵
@snaggen@jo@tante yeah, also referring to all LLM output as slop would end up redefining the word as something inherent to LLMs which effectively makes the word a meaningless, neutral term
in reply to nicole
@mkljczk @jo @snaggen nah. consider: referring to the product of human digestion as “gówno” does not make the word a meaningless, neutral term, quite the opposite.
in reply to nicole
@mkljczk @jo @snaggen what are you talking about, there is literally no reason it would do this. That's not how language works in the real world.
Esta entrada fue editada (jueves, 15 de octubre de 2026, 17:36)
in reply to nicole
@mkljczk @jo @snaggen so we should stop coloquially referring to AI output as slop because non-native speakers don't know the official definition of the word "slop"? They've not read enough english to have seen something described as "sloppy"? They are incapable of connecting the word "slop" to "sloppy"? They are not capable of looking up words they don't know when reading a foreign language? Because that's what i do as person who reads multiple languages. This is a non-issue.
Esta entrada fue editada (jueves, 15 de octubre de 2026, 18:40)
in reply to jo the disgraced
@jo I had a consultant explain to me once how they were "still working on implementing another layer of AI to filter out the mistakes and hallucinations". So the fact that their demo was producing mostly garbage was nothing to worry about. 😂
in reply to jo the disgraced

@jo I use slop to describe generated code that has not been reviewed at reasonable level and has not had other methods in use to validate code quality independently of the primary generation task, it has nothing to do with quality except as indirect consequence.

You probably count me as AI bro or whatever, while I do accept much of the criticism at some level.

in reply to tante

This too (Meta interference), appears like an attempt by Meta to take over Infrastructure. They’ve already tried as an ISP, and they have bullied their way into fiber deployment across the world. If I’m reading this correctly, this is very worrisome.

"Linux is not a social warrior project, it’s a Meta project! When your networking fails, remember to say: Thank you Meta!”

in reply to Cassandrich

@dalias
Yup. I have a similar private note for that account.

I should probably follow through on my desire to call these accounts out on the Fediblock hashtag. Second strike == raise the alarm. Seems fair. I'm pretty good about checking the nature of their other posts before commiting to the initial block.
@Profpatsch@mastodon.xyz

in reply to tante
It seems to me that some people’s opposition to AI is downright fanatical. It’s one thing to have silly pictures of cats doing human activities; it’s another thing to use AI to help create software, where it’s actually useful.
This black-and-white rejection of generative AI is childish.
in reply to Michal Špondr
@michal It isn't creating anything, it's stealing labour from uncredited uncompensated humans. Microsoft themselves described it in those terms in a leaked document: techcrunch.com/2026/09/17/microsoft-exec-called-ai-scraping-the-largest-theft-of-labor-in-human-hist…
in reply to Michal Špondr
@michal Pirates pass on unauthorised copies, they aren't trying to claim credit for creating anything. GenAI/LLMs are plagiarism laundering machines with no knowledge about what their output means.
in reply to Michal Špondr
@michal @FediThing NO. First of all, GEN AI / LLM are software, not 'technologies' per se. Second, these softwares are tactically designed by these fascist corporations for its merely purpose of doing what it is intended to do, no matter how people use it, even if they 'self host it'.
in reply to Michal Špondr
@michal @FediThing how would you train those local open-weights LLMs? With whose text, images, information, audio would that be? Are you gonna scrap all that without conscent from other people's authorship over their own work? On whose propietary hardware are you gonna run those for them to properly work?
in reply to Michal Špondr

@michal@spondr.cz @xaq @FediThing @tante you people are fucking insufferable. I don't give a shit about your workflows or whatever research & development you're streamlining with your teeny tiny baby LLM who's never chugged way too much water & power bc he's just a little guy with little bitty training data sets. UwU

Fuck your LLMs at every scale. I don't care. You're a shill & my kids have no future on this planet you're making excuses for destroying with these stupid fucking technicalities. Shut up.

Esta entrada fue editada (jueves, 15 de octubre de 2026, 18:42)
in reply to Michal Špondr
@michal @FediThing The way I think about it is: You use it, you are the problem. I don't care how you use it. I don't care if you claim to always check its output. I don't even care if you sometimes check its output. I don't care if you think the output is better than what you would produce without it. You are still the problem just by using it. It's really not that difficult.
in reply to Michal Špondr

@michal acceptable piracy steals from soulless corporations and gives to individuals; AI steals from individuals and gives to soulless corporations.

Also, genAI is an existential threat to so many communities¹, and any use of it is unethical (though there are *some* uses that are acceptable, IMO, such as as an accessibility aid).

¹ all of Fedi, for instance

@FediThing @tante

Esta entrada fue editada (jueves, 15 de octubre de 2026, 15:50)
in reply to Michal Špondr

@michal @FediThing Pay close attention to the words.

Labour is what was used. The point is not the product.

The commodification of art is also an atrocity, which has been pushed and perpetuated by copyright maximalists.

See the Golden Calf and its follow-up volumes if you're legitimately interested in learning about that topic.

in reply to Michal Špondr

@michal You wanna talk about fanaticism? Look up TESCREAL and how the people at the top of the slop pile are essentially doomsday cultists rushing to develop AGI to make humanity obsolete.

And no, AI helping people write code is not a good thing by default. You're the one thinking in black-and-white.

in reply to Michal Špondr

@michal Nevermind the ethical problems or the fact the technology is fundamentally unreliable and cannot have its risk of introducing Trusting Trust issues & Hypocrite Commits removed, as well as the cognitive damage known to occur to its users regardless of how well they know how it works.

You would literally be better off binge-drinking. It comes with more warning signs.

Esta entrada fue editada (jueves, 15 de octubre de 2026, 17:49)
in reply to Michal Špondr
@michal why do you assume that i arrived at my utter rejection of autoconfabulators & similar technologies by throwing dices and making my mind based on the result?
Esta entrada fue editada (jueves, 15 de octubre de 2026, 18:02)
in reply to Michal Špondr
@michal it’s religious, really. I silence these zealots most of the time. And secretly I enjoy Linus doing this, I wonder which OS they’re going to run to now. Or in 5 years, when almost all software is llm coded. Ah well.
in reply to tante

man there are layers to this shit onion for sure.

Don't like the meta involvement. Don't like the pay-to-play attitude for kernel contributions (although I agree that contributions should not be made easy. They shouldn't be made exclusionary as well. It should just depend on code quality). Don't like the overall tone this is worded ("social warrior project")

Unfortunately Linux is still the least bad alternative in terms of compatibility and support compared to windows and MacOS.

in reply to mr_harm
I've seen people pushing for people to move to BSD distros. Which in some senses are better than Linux (resistance as a community to external pressure, secure-by-default distribution guidelines) but in some others it's lagging far behind (software and hardware support, desktop compatibility, a license that makes proprietary forks much more likely). It's a pick-your-poison type of scenario.
in reply to tante

[1 of 2] The wording is a bit too strong there, blaming wrong people for wrong things. Slop is a result of a developer ignoring the code output. It was there before LLMs, though people have to try harder even if they just copy-paste from StackOverflow without thinking.

Author assumed there are AI Overlords, while in reality it’s just a bunch of corrupted and greedy (and mostly crazy) technocrats.

in reply to an actual bus
@renardboy Exactly. Also, various corporations contributed tons of code to the Linux kernel. Google, Meta and so on. The problem with such purisms is that they don’t see shit beyond one phrase, word or action of one person at some point in time. Suddenly, that one thing summarize that person.
in reply to Mister Dave

@misterdave Nobody is doing anything bad. He used Claude to make something faster in Linux kernel. He’s not calling to do erhnical cleansing.

There are real villains like DHH, if you need one of software world.

in reply to Oleksii

@oleksify@mastodon.social
The cognitive dissonance of putting "vegan" in your profile and still thinking your AI apologia is welcome in an AI critical thread. May all your vegan advocacy threads be filled with carnivore reply guys, fuckface.

(no disrespect to vegetarians or vegans who understand the irony here; I keep my mouth shut in your threads)
@tante

in reply to tante
I continue to call it slop at work. I am not well liked by some coworkers but I don't give a shit, I don't respect them at all. We'll see how long it takes to get reprimanded by HR I guess, but I ain't stopping.
in reply to paelnever

@paelnever

He was known for being toxic and verbally abusive on the kernel mailing list for a long time. He was "celebrated" in "men's rights" forums and on Breitbart for some of his public statements.
Yes, he got therapy and apologized, that's a first step. But he still has work to do and his statements about AI and the elitism in them don't raise much hope in me.

@tante

in reply to tkissing

@paelnever

lkml.org/lkml/2012/7/6/495

"retroactively aborted" is not someone in his position, even in 2012, should write and it's way beyond "having a bad day".

Want some more? "Whoever wrote it should be shot." maybe? How about "Hardware designers are f*cking morons that should be given extensive education about birth control and how not to procreate."?
The topic of abortion was quite popular with him in the 2010s, but there was also plenty of "brain damage" and "retarded".

in reply to tkissing
@tkissing
I would not call somebody "toxic" for having a dozen of bad days in more than 30 years. Must of the comments in the mailing list considered "toxic" are for simply defending the quality of the code inside the kernel, maybe with a little bit too much passion. I never saw a racist, xenophobic or sexist comment from him and if you have even a single example of it please show me.
Apologize is much more than a "first step", is one million light years away from 99% of the people is able to do.
As far as i know their statements about AI are something like "i don't care how you write the code you commit but you must understand the code, understand the kernel architecture and you are responsible for your commits".
About the "elitism" i don't know what are you talking about, please teach me.
Esta entrada fue editada (jueves, 15 de octubre de 2026, 17:53)
in reply to tante

It would be good for all sane protein-based maintainers to stop commiting. LT is claiming that bots can. Let him vibe code with bots only. Do your tree to the drawer for a longer while. IDNK how fast it will be to get datacenters to crawl then crash due to the slop influx. Oh, it would be nice to see.

As for the LT 'infatuation'.

> the kernel project only wants serious developers. That includes having access to enough resources to function in this world.

AI pushers can pay more than a hundred of grands to take over a YT programmer tutorials channel. So yes, now came the times a maintainer of mid-popular FLOSS project you can sell out, buy a house, pay your kid's tuition. And embrace for the new brave world you helped 'them' to acquire.

Just personal 0.02 of any currency used.

in reply to tante

I guess it's not all that surprising that Linus would develop a preference for interacting with AI, nor that he'd just belittle and denigrate those who disagree with him.

As a C++ developer--and not just that, one that would drive him NUTS--I feel proud to have earned his abject hatred without him even knowing I exist. A smart man with some very silly notions he's quite unwilling to be skeptical of.

vOv

I have liked his system but honestly it's been pissing me off of late.

in reply to jesterchen42

@jesterchen
If you want to have your freedom spoiled by irresponsible traitors to FOSS, yes.

IMHO it's rather trowing out the baby with the bath water.

I wonder why people are always exaggerating the consequence (hey, why not totally refrain from using electricity, this was certainly where the problem originates), thus putting everybody of from doing the right thing, because this was just pivoted into extremism.

I worked pretty well with my computer 2010 without LLM.

in reply to tante

NetBSD is committed to not using AI code - take that information with what you will.

Hopefully a more consumer friendly NetBSD based disto becomes a thing.

Esta entrada fue editada (jueves, 15 de octubre de 2026, 15:25)
in reply to Javensbukan
@Javensbukan I'm installing NetBSD on an old laptop right now to give it a try. I try to avoid slop as much as possible, and am very disappointed with the Linux Foundation's decision to allow slop in the kernel.
in reply to Keith
The media in this post is not displayed to visitors. To view it, please go to the original post.

@keydelk I'm trying GhostBSD now which is feeling very similar to Debian/Ubuntu experience wise.

It's based on FreeBSD, which does allow AI tools for editing documentation and stuff, but not for code submits.

Better than Linux I suppose.

in reply to Javensbukan

@Javensbukan people seem to be getting their arse in their hand about the precise wording of that commitment, rather than actually prepared to see how it's playing out in spirit...

or allow for the fact that in May 2024, before the AIpocalypse, the copyright implications were the only really visible ones and therefore the only ones worth making policy about - and doing so then was damn near prescient. i think we've all forgotten how quickly that changed.

Esta entrada fue editada (jueves, 15 de octubre de 2026, 17:00)
in reply to VoidZeroOne

@TheOneDoc Like, any regular old consumer can install it and use it as a daily driver on a PC.

Ala Linux Mint or Ubuntu.

This already exists for FreeBSD, but not NetBSD to my knowledge.

Esta entrada fue editada (jueves, 15 de octubre de 2026, 17:40)
in reply to tante

"Torvalds’s first publicly released chatbot code was in January this year, when he vibe-coded a project in Python because he didn’t know enough Python. The programming language kids learn in high school. The guy who runs the kernel couldn’t get his head around Python"

I hate when people write like this It's genuinely poisoning the discourse imo.

in reply to tante
in reply to tante

I looked at NetBSD and it scares me. I run s shit-ton of home stuff and some commercial products on Trixie and just don't see what a viable option is. Even if I went NetBSD (unlikely) - what about my Proxmox(es) and VPSs - it's fucked.

I guess from now on when asked, I'm running the Linux sloperating system.

in reply to tante
This article is indeed pure "slop". The Linux kernel is not going "full AI slop code". Linus Torvalds have stated very clearly several times that after every commit to kernel code there must be a person responsible for it and that person must understand not only the programming language but the whole architecture of the kernel and the implications of the changes committed. Behind this filthy campaign against Linux there are interest of the major privative software companies (the main suspects are as usual microslop and apple) that always try to trow all kind of false accusations and fears against Linux. They have being trying it for more than 30 years in every way possible and that's only the last try using the anti-AI rage as a weapon. Linux is not the culprit of AI slop, neither free and copyleft software. In fact free and copyleft software is the best tool that the normal people have against the use of AI tech by the billionaire class.
Esta entrada fue editada (jueves, 15 de octubre de 2026, 22:23)
in reply to mx alex tax1a - 2020 (6)
@atax1a
Yes, precisely the people who convinced you that the kernel is being vibecoded now must have some dark objective because that's objectively false. Don't believe everything you hear, specially if what you hear is against free and copyleft software.
in reply to tante

dude - i get this everywhere i go.

a librarian was complaining about me being in the library i visit 3-4 times per week.

she said 'and he talks back to me,' as if i were a peasant and she were royalty.

i get that from security guards, too. and store employees. and workers in non-profits that provide services for homeless people.

there's been a lot of shit-talking about me (and my mother) at the center for positive aging, because i was using drugs there. on a day i wasn't there!

in reply to tante
The media in this post is not displayed to visitors. To view it, please go to the original post.

shaming is good. I'm sure you saw how Abdul ElSayed shamed Tommy "I'm a Christian!" In Michigan. Managed to get Tommy's tail between his legs in under 5 minutes. What a dick, Tommy. eupolicy.social/@jmaris/117461810385145539
(Tommy's been outed as a former staffer of Josh Hawley, not necessarily something to be proud of). The Onion will have a field day with Tommy.


It's very easy to get depressed by the state of politics right now, but then you see interactions like this one, and it rekindles your faith. A new generation of politicians will come forward and they will be different.

#AbdulElSayed #USpol #Politics


Esta entrada fue editada (lunes, 19 de octubre de 2026, 11:57)