But all AI output is slop. It’s social spam. That’s what it’s for. It’s slop! Say “slop” every time
This writing is far, far too emotional to be considered anything but an opinion piece
sir this is a blog post
It’s so weirdly emotional that I’m sure I’m saving some folks the click
do people not see the url? the tagline? the smug clippy?
Well fuck off back to Windows then (I’m sure you can determine whether or not the code for that has been near an LLM… oh wait). Or fork it, if you can. If neither of those stop being so entitled about what tools someone uses to build the OS they give you for free!
“You don’t like Trump? Well maybe you should go back to the middle ages. Or just go and make a new country, see if that’s so easy, huh”
Since when is the argument of “other things are even worse so we shouldn’t care about something becoming worse” a thing?
Nevermind, don’t answer that, as usual the answer is “Since the rise of the ‘AI’ cult”
No it’s been an argument since forever, not always a good one but that’s not really the point. And the use of AI does not make things worse anyway you absolute nitwit, but your performative hate and wilful ignorance do - and they lead directly to shit like trump! YOU are the problem
So the fact that LLMs are having a negative impact on already collapsing environment, are leading to people like Trump and Musk to become even richer and destroy society at a massive scale, while also killing individuals by literally making them go insane, is “not making anything worse”? All of that so a bunch of dumb fucks can go and generate garbage “pictures” and broken code?
You say that like it’s true. It’s not. None of it. The few bits that are close to true are old school corruption not AI. Twit.
Ah ok, full on cultist then
Just someone who actually understands the situation,. unlike you
Yes, and we’ll all be left behind if we don’t join the sect, and LLMs are here to stay and replace the world as we know it.
… If you are a competent developer, he said, an LLM makes a great tool, but only after that effort has been put in to set it up. He resisted suggestions to post his LLM-training materials, saying that each developer needs to go through that process separately.
This is the vibe-code-two-step. “AI” is useful because it helps inexperienced users to code. But you also have to be an expert to use it properly.
Same thing could be said about a fishing pole. Or car, or rifle, or a bicycle.
It helps inexperienced users to catch fish… But with some knowledge and experience can get more out of it.
Not the same at all. You don’t have to be a gunsmith to kill someone with a gun. You do have to be a knowledgeable engineer to produce anything that isn’t slop with “AI”.
A MUCH better analogy is construction. Do you want a bunch of teenagers and Main Character Syndrome dipshits who cannot even read blueprints nor understand regulations and building codes let alone know them, to build your house? Nono, it’s OK! They have nail guns and modern lumber, they don’t need to know all that!
Sure you can get a shed or even a small house out of that, but you’re never going to get safe, modern, multistory construction like that…
It’s a tool, not a product.
LLMs and other machine “learning” are tools.
Most of what people talk about when they talk about AI… are talking about products. ChatGPT is a product being sold. Claude is a product being sold. “AI” being shoved into other products also makes it a product being sold. etc.
They are not even mutually exclusive words…
You’re right, they’re not mutually exclusive words, but I’m going to thoroughly disagree that people are talking about those specific companies when talking about AI as a technology, which we certainly are here.
You might not like the product that is Claude or whoever, but I don’t like Irwin hammers, doesn’t mean it’s not an effective tool. And it doesn’t stop the world from utilizing it. And it doesn’t automatically make the job any less correct.
No one is out here trying to have the AI be their OS, they’re using it to build things, so the house thing isnt any better. Keeping those teens off the jobsite is only going to stop that house, not the industry. You don’t want to live in it, that’s fine, but that doesn’t mean it’s not going to be exactly what someone else needs. And those kids did get some experience, even if its just the walls falling down around their ears. Next one will be better.
Would it be ideal if everything was inherently safer and more thoroughly understood by the masses? Yes. Is it a reality that is achievable?
lol defending teenagers willy nilly building a house and ignoring building codes…
OK
Your attitude is beyond pathetic. Your attitude would still have us working 10x6 days with no insurance anywhere to be seen. Your attitude reinforces all the shit you think “AI” can fix.
Yours would have us using bones as clubs.
You do realize that all code for the kernel is code reviewed and tested by so many people right? The patches are probably some of the most human reviewed code on the planet…
This feels like the time when every armchair expert became experts on immunology…
The article is literally about how they have so much AI slop that the code is no longer reviewed by people it’s reviewed by a Meta chatbot…
The article is one angry person’s rant and you are further going beyond what it actually says.
We are completely overwhelmed, of course. The glimmer of hope is that we secured sufficient LLM budget and access (thank you Meta!) to run reviews with multiple frontier models on each patch. This eliminates some hallucinations.
That’s not the author’s words, it’s a quote from the Linux mailing list…
Yeah. Pretty much.
The actual policy is that it all gets reviewed by a human I think when I was looking into it.
The bots are likely most useful to reject obviously bad patches which have obvious bugs or don’t follow guidelines. And identifying bad code is something which AI tends to be good at.
In our code at work, we use weakly typed languages which aren’t strict and on those it tends to be even better. We catch things like table keys that are typed wrong, etc. On C, things like pointers it probably does a good job with identifying as wrong
If it passes AI code review it still gets checked by human for things that are wrong and needs fixing. But if it fails, it probably needs fixing regardless.
A good language server / ide will also complain about incorrect types. That specific example is a deterministic process, no llm is required for it at all.
We do use one. But, in practice, they can’t catch a lot of things which AI will
At least in our case too based on what’s available, it also can’t check stuff like coding style well (but our development relies on that less).
Some might though
I find that very hard to believe, if you’re truly just talking about typing. If this is true you’re using an awful language server / linter / whatever.
Coding style is also a deterministic process, for which every widely used language has a linter or formatter or whatever. Also no LLMs needed.
As a professional software developer of 8 years, I honestly have no idea what you’re talking about
This is Lua code. LuaLS is basically the language server available.
And sorry, I should have said “dynamically typed”. But, weak tables.
There’s a lot of things that don’t get caught, such as concatenating a boolean which causes a lua error in our environment. LuaLS won’t catch that.
Formatting isn’t a problem for us, but it does affect the kernel. But, the VSCode one we use does a lot of weird things honestly.
