Okay, I'll give you that: LLMs aren't the root of the problem.

Capitalism is. The idea of infinite growth. The idea that people can't just live, they must with 40+ hours a week to justify their existence, and they must be purchasing something all the time. Companies must keep selling new stuff. All the resources must be tapped into and exploited.

And companies are making software. They must keep selling new features and pointless complete redesigns nobody wanted. The code must keep being churned over and over again. Programmers must justify their existence by churning out absurd amounts of meaningless code. The companies must exploit them.

Then, companies are entering the #OpenSource "market". They are acquiring and enshittifying. They are hiring and exploiting. And then so many volunteers just jump on the bandwagon and keep cosplaying them. And they too churn out useless code, "sell" pointless complete makeovers, "profit" off their users (even if they actually aren't making any real profit).

And then come LLMs, perfect tools for the job. Perfect tools for exploitation, for churning out useless code, for creating addiction, and for turning everyone into mindless corpospeak bullshit machines.

#AI #LLM #NoAI #NoLLM #AntiCapitalism

Okay, so apparently there's been some "shuffle" between a cyclist and an old lady. The police's looking for the cyclist now, and shared a camera footage looking for help in finding them. Except that the footage is such a low resolution it's practically useless.

So helpful people from the internets used "#AI" to enhance it. So now we're looking at an angry mob looking for a person whose face was generated by an #LLM. Or well, multiple independently generated different faces apparently, but would that stop a mob from lynching a random person?

This fucking crap needs to be outlawed immediately. And whoever's selling it should end up behind bars.

#NoAI #NoLLM

New #slop recognition criteria just dropped: a project bumps the dependency to `setuptools>=61.0.0` (valid but most people don't care) and adds a `wheel` dependency while at it (wrong).

https://github.com/aio-libs/aiodns/pull/252#discussion_r3252294971
https://github.com/micheles/decorator/commit/05480a17ebd4c4c81ca16526c3add53a16a7831f#r185646957

#NoAI #NoLLM

Oh, great. I'm sure it's great idea to boil oceans in order to have some random bullshit machine repeat what I said in the description. And also add a bullshit suggestion with bullshit explanation that's plain wrong.

https://github.com/aio-libs/aiodns/pull/252#pullrequestreview-4303099647

#AI #LLM #NoAI #NoLLM

Of course, we have it.

Now someone's commenting on my big blog post, agreeing with its points and at the same leaving link to a website about "AI Text Tools & Writing Assistants".

#NoAI #NoLLM

Yesterday, I've read a vibe coded script for the first time in my life, and I've cried.

It wasn't ugly. "Ugly" is not the right term. It was as if someone wasn't able to comprehend beauty, but badly tried to mimic it. It felt like "malicious compliance" to beauty. The kind of awful verbose pedantry that feels wrong every step of the way.

It's the kind of code you'd expect in a corporate environment when you know that the code would be read by the top suits who have no idea about coding, but judge it by the volume and expect science fiction level of make-believe.

It's the kind of code is abstracted away into the tiniest details. Every function returns a complex dataclass explaining precisely what it did, for no reason at all. What would be two lines of code is a function. What would be a function is a whole module. It's a caricature of good programming practices.

I was supposed to add modifying a second field on the same object via GitHub API. I've guessed it would take me about an hour to figure out the code enough to be able to do that — what ought to be 2-3 extra lines. I suspected I'd discover that most of the code does precisely nothing. Just meaningless API exchanges that are absolutely unnecessary. It felt like the kind of parody of bureaucracy where you have to file 10 forms to do something, and only one of them actually means anything.

What used to be "do one thing well" became "doing ten totally random things is fine, as long as one of them happens to be what I need, and the whole thing doesn't blow anything up in an obvious way".

Perhaps it's just because this way a throwaway script. Maybe "production" stuff takes more, err, prompt refining? Maybe it actually can produce stuff that's comprehensible.

But if that code was any indicator, then I'm not going to believe that any big LLM contributions are actually reviewed by humans. A review will take more time than rewriting from scratch. This is a ticking time bomb. That LLM-generated code isn't introducing exploits right now is either a statistical accident, or it's just that nobody bothers.

#NoAI #NoLLM #AI #LLM

If you're looking for another thing to thank #LLM techbros for: #OpenAI is now acquiring #Cirrus Labs, and #CirrusCI is going to shut down in <2 months.

https://web.archive.org/web/20260407101724/https://cirruslabs.org/

#AI #NoAI #NoLLM

Remember how people gave the term "#AI" to use for their #LLM crap, and then started using "AGI" for the old AI?

Apparently techbros are now selling LLM crap as "AGI": https://futureagi.com/

Also, my "work on GitHub" is apparently "directly relevant to what they're building". Enough to justify a #spam mail anyway.

#NoAI #NoLLM