Yeah, the sellout.
From reading all the comments from the community, it’s amazing (yet not surprising) that all these managers have fallen for the marketing of all these LLMs. These LLMs have gotten people from all levels of society to just accept the marketing without ever considering the actual results for their use cases. It’s almost like the sycophant nature of all LLMs has completely blinded people from being rational just because it is shiny and it spoke to them in a way no one has in years.
On the surface level, LLMs are cool no doubt, they do have some uses. But past that everyone needs to accept their limitations. LLMs by nature can not operate the same as a human brain. AGI is such a long shot because of this and it’s a scam that LLMs are being marketed as AGI. How can we attempt to recreate the human brain into AGI when we are not close to mapping out how our brains work in a way to translate that into code, let alone other more simple brains in the animal kingdom.
I agree with almost all of your comment. The only part I disagree on is:
How can we attempt to recreate the human brain into AGI when we are not close to mapping out how our brains work in a way to translate that into code, let alone other more simple brains in the animal kingdom.
An implementation of AGI does not need to be inspired from the human brain, or any existing organic brain. Nothing tells us organic brains are the optimal way to develop intelligence. In fact, I’d argue it’s not.
That being said, it doesn’t change the conclusion: We are nowhere near AGI, and LLMs being marketed as such is absolutely a scam.
One of the best written comments I’ve seen about this. LLMs are cool for what they can do, but anyone comparing them to AGI is just shilling and trying to make a fortune off of selling pickaxes in a gold rush with the only gold being fools gold.
From reading all the comments from the community, it’s amazing (yet not surprising) that all these managers have fallen for the marketing of all these LLMs
This is probably related to automation bias and wishful thinking
I don’t think LLMs will become AGI, but… planes don’t fly by flapping their wings. We don’t necessarily need to know how animal brains work to achieve AGI, and it doesn’t necessarily have to work anything like animal brains. It’s quite possible if/when AGI is achieved, it will be completely alien.
We don’t necessarily need to know how animal brains work to achieve AGI, and it doesn’t necessarily have to work anything like animal brains.
100% agree. Definitely thinking inside the box, inside the brain, when I went down that path.
I think better way to explain my thinking is that LLMs can not operate like a human brain in that they fundamentally lack almost all qualities of a human brain. They are good but not perfect at logic just like humans, but they completely lack creativity, intuition, imagination, emotion and common sense, qualities that would make AGI.
Without humans being able to understand how our brains process those qualities, it will be very hard to achieve AGI. But again, very wrong of me to think we need to translate code from our brains to achieve AGI.
Aircraft wings operate on pretty much the same principle as bird wings do. We just used a technology we had already developed (fans, essentially) to create the forward movement necessary to create the airflow over the wings for lift. We know how to do it the bird way too, but restrictions in material science at scale make the fan method far easier and less error prone.
I can’t wait until billionaires realize how worthless they actually are without people doing everything for them
We can’t to wait for them to realize this themselves. We need to demonstrate this by actively creating a society which excludes them.
Yes we should start arresting them and when they flee to Israel we sanction and embargo it with every country on earth
its either israel, the uk, or RUSSIA. trump is going to russia.
Trump is in a lot deeper with Israel. He is preparing to spend trillions invading Iran ( a country that poses no threat to us) to assist Netanyahu with his settler colonialist expansion project or Middle East manifest destiny. He bypassed congress in January to Gift Israel a new fleet of fighter jets from Boeing he also sent them an additional 8 billion in aid. Russia is the straw man. Ghisline and Epstein were Mossad and they worked for Israel to collect blackmail on the leaders of the west. Israel is violating their ceasefire and withholding aid from Gaza. Journalists still can not get in the open air prison that is Gaza and Gazans are locked in so they can’t flee. ISRAEL literally owns our politicians WAKE UP
Might be a minute. The brain damage that lets them think they’ve “earned” those billions kinda hides the work of others. Especially the poors.
Yep they’re definitely mentally ill
Very few billionaires built from their own effort. Authors and Musicians? The guy who made Minecraft? Khaby Lame?
Eh, as the world goes to shit there will always be desperate people willing to work for them, probably cheaper than before even with the AI failures, so they wouldn’t care.
up until they stop paying them, once they are in thier bunkers, islands, the “servents” will feast,.
They will never realize that, they will blame any failures on others naturally. They truly believe they are better than everyone else, that their superior ability led them to invest in a company that increased in value enough for them to become filthy rich.
Surrounded by yes men and woman that agree with everything they say and tell them what a genius they are. Of course any ill outcome isn’t their fault.
“All my successes are thanks to my superior intellect and skill! All my failures are the fault of bad serfs who didn’t follow my vision!” - Every billionaire
When you think about it, it’s not too different from how some people treat the current crop of AI, so it makes sense that they’re so hypnotized by the promises.
Wouldn’t hold my breath for it.
“Top-down mandates to use large language models are crazy,” one employee told Wired. “If the tool were good, we’d all just use it.”
Yep.
Management is often out of touch and full of shit
“often” they are almost always full of it.
If the tool were good, we’d all just use it.”
Eggs-mothafucking-zackly!!!
There are no daily pressure campaigns to convince you to use a laptop or a smartphone. The value of those are self-evident.
AI on the other hand… -_-
You wanna know who really bags on LLMs? Actual AI developers. I work with some, and you’ve never heard someone shit all over this garbage like someone who works with neural networks for a living.
There’s this great rage blog post from 1.5 years ago by a data scientist
Oh, the guy from Hermit Tech! He’s great, and his blog is hilarious and poignant in turns (though sometimes both at the same time).
Missed this one, thanks
our company renamed out ML team into AI team, just to please investors, they been around for over a decade and never touched an LLM
That’s me, but for QA…
Management: “No, that doesn’t work, because employees spend so much time doing the actual work that they lack the vision to know what’s good for them. Luckily for them I am not distracted by actual work so I have the vision to save them by making them use AI.”
Something an idiot would say. Jack Dorsey is precisely this type of idiot. He’s not the only idiot, though
I had a meeting with my boss today about my AI usage. I said I tried using Claude 4.5, and I was ultimately unimpressed with the results, the code was heavy and inflexible. He assured me Claude 4.6 would solve that problem. I pointed out that I am already writing software faster than the rest of the team can review because we are short staffed. He suggested I use Claude to review my MRs.
One big problem with management is their inability to listen. Folks say shit over and over but management seems deaf because we’re not people to be listened to. We’re the help. And management acts like they know better.
This has been my life for the last nine months. I’m thinking of getting of software development all together for fear that no other place will be any different regarding AI.
This is a major issue with capitalism. It is a massively inefficient way to organize society. The people with the most money do not necessarily make good decisions. They usually make selfish decisions.
If you were so smart you’d have wads of cash like them. They got where they are through sheer grit and bootstraps and a paltry $50 million from their family.
next time, tell your boss that Claude should replace him, not you.
It’s not like MIT and the Harvard business review have published studies that have shown that AI is actually best suited to replacing executives and management in order to flatten organizations. But unfortunately, management and executives make the decisions on who AI replaces, and they don’t want to be replaced. Hell at the company I’m at right now they’ve been axing low level workers and bringing on or promoting the ladder climbers (read: AI sycophants who do the least work) to manager or department head roles, saying that us grunts can “10x” to fill the gaps, and that all we need is good and creative leadership to direct our AI use. I could go off on things this business is doing to shoot itself in the foot for hours, even without mentioning AI
The trick is to tell them you’ve been using it more than they have and that it’s not as good as chatGPT for task A, but that for task B claude does okay 25% of the time so we’ll need to 4x the timeline in order to get a good claude output based on that expected value.
But not as good as your personal local LLM that you’ve been training on company data. No one else can use it because it’s illegal to clone. (your personal local LLM is your brain)
Telling your boss you trained a personal LLM with company data will lead to nothing but you holding a box full of your stuff about 10 minutes later.
Yeah, and that’s possible even if they take it for a joke that it is (or isn’t)
Given that this person’s boss wants them to dump company data into Claude, I see no functional difference.
Ralph Wiggum loop that shit
Numbers go up, Claude won’t bother you 👍🏻
What software are you writing? I’m struggling to see what any of this does, and for who. We could set all these AI computers on fire and what would it change? We have water, food, electricity, clothes, homes, cars, etc Oh we have AI Becky videos!
At work today we had a little presentation about Claude Cowork. And I learned someone used it to write a C (maybe C++?) compiler in Rust in two weeks at a cost of $20k and it passed 99% of whatever hell test suite they use for evaluating compilers. And I had a few thoughts.
- 99% pass rate? Maybe that’s super impressive because it’s a stress test, but if 1% of my code fails to compile I think I’d be in deep shit.
- 20k in two weeks is a heavy burn. Imagine if what it wrote was… garbage.
- “Write a compiler” is a complete project plan in three words. Find a business project that is that simple and I’ll show you software that is cheaper to buy than build. We are currently working on an authentication broker service at work and we’ve been doing architecture and trying to get everyone to agree on a design for 2 months. There are thousands of words devoted to just the high level stuff, plus complex flow diagrams.
- A compiler might be somewhat unique in the sense that there are literally thousands of test cases available - download a foss project and try to compile it. If it fails, figure out the bug and fix it. Repeat. The ERP that your boss wants you to stand up in a month has zero test coverage and is going to be chock full of bugs — if for no other reason than you haven’t thought through every single edge case and neither has the AI because lots of times those are business questions.
- There is not a single person who knows the code base well enough to troubleshoot any weird bugs and transient errors.
I think this is a cool thing in the abstract. But in reality, they cherry picked the best possible use case in the world and anyone expecting their custom project is going go anything like this will be lighting huge piles of money on fire.
I wanna make sure I got this right. They used $20,000 in fees in 2 weeks to make a compiler? Also, to what end? Like what’s the expected ROI on that?
Well it’s Anthropic, creators of Claude. It’s a way to show off and convince people AI can do it. $20k is what it would cost you or me, but it’s just free for them.
I don’t even hate AI but it’s kinda sickening the way they overstate the capabilities. But let me tell you how excited the top leadership at my company is about this…
$20k is what it would cost you or me, but it’s just free for them.
No it isn’t. This is not regular software where the bulk of the price is the licensing. With slope-as-a-service, the bulk of the price is the data center operation cost - which Anthropic is certainly not getting for free.
I mean there is a cost associated with it, just like there is a cost associated with having free soda in the break room, but it was free for the person doing the project. It’s absorbed into operational costs.
Considering how these companies are losing money because they subsidize these tokens - I doubt that cost is really absorbed.
https://harshanu.space/en/tech/ccc-vs-gcc/ has a good overview how bad it really is
Thank you. Great addition. That was a very interesting read, though I need to be more awake for reading technical writing like that 🥱.
My point about spending $20k to produce garbage, then, was actually realized in this “perfect” use case.
That was interesting to read if not a bit jargon heavy. Thanks for sharing
99% pass rate? Maybe that’s super impressive because it’s a stress test, but if 1% of my code fails to compile I think I’d be in deep shit.
Also - one of the main arguments of vibe coding advocators is that you just need to check the result several times and tell the AI assistant what needs fixing. Isn’t a compiler test suite ideal for such workflow? Why couldn’t they just feed the test failures back to the model and tell it to fix them, iterating again and again until they get it to work 100%?
Maybe they did, that’s how they got to 99%. The remaining issues are so intricate/complex the LLM just can’t solve them no matter how many test cases you give it.
It’s even simpler than that: using an LLM to write a C compiler is the same as downloading an existing open source implementation of a C compiler from the Internet, but with extra steps, as the LLM was actually fed with that code and is just re-assembling it back together but with extra bugs - plagiarism hidden behind an automated text parrot interface.
A human can beat the LLM at that by simply finding and downloading an implementation of that more than solved problem from the Internet, which at worse will take maybe 1h.
The LLM can “solve” simple and well defined problems because its basically plagiarizing existing code that solves those problems.
Hey, so I started this comment to disagree with you and correct some common misunderstandings that I’ve been fighting against for years. Instead, as I was formulating my response, I realized you’re substantially right and I’ve been wrong — or at least my thinking was incomplete. I figured I’d mention because the common perception is arguing with strangers on the internet never accomplishes anything.
LLMs are not fundamentally the plagiarism machines everyone claims they are. If a model reproduces any substantial text verbatim, it’s because the LLM is overtrained on too small of a data set and the solution is, somewhat paradoxically, to feed it more relevant text. That has been the crux of my argument for years.
That being said, Anthropic and OpenAI aren’t just LLM models. They are backed by RAG pipelines which are verbatim text that gets inserted into the context when it is relevant to the task at hand. And that fact had been escaping my consideration until now. Thank you.
Even the LLM part might be considered Plagiarism.
Basically, unlike humans it cannot assemble an output based on logical principles (i.e. assembled a logical model of the flows in a piece of code and then translate it to code), it can only produce text based on an N-space of probabilities derived from the works of others it has “read” (i.e. fed to it during training).
That text assembling could be the machine equivalent of Inspiration (such as how most programmers will include elements they’ve seen from others in their code) but it could also be Plagiarism.
Ultimately it boils down to were the boundary between Inspiration and Plagiarism stands.
As I see it, if for specific tasks there is overwhelming dominance of trained weights from a handful of works (which, one would expect, would probably be the case for a C-compiler coded in Rust), then that’s a lot more towards the Plagiarism side than the Inspiration side.
Granted, it’s not the verbatim copying of an entire codebase that would legally been deemed Plagiarism, but if it’s almost entirely a montage made up of pieces from a handful of codebases, could it not be considered a variant of Plagiarism that is incredibly hard for humans to pull off but not so for an automated system?
Note that obviously the LLM has no “intention to copy”, since it has no will or cognition at all, what I’m saying is that the people who made it have intentionally made an automated system that copies elements of existing works, which normally assembles the results from very small textual elements (same as a person who has learned how letters and words work can create a unique work from letters and words) but with the awareness that in some situations that automated system they created can produce output based on an amount of sources which is very low to the point that even though it’s assembling the output token by token, it’s pretty much just copying whole blocks from those sources same as a human manually copying a text from a document to a different document would.
In summary, IMHO LLMs don’t always plagiarize, but can sometimes do it when the number of sources that ended up creating the volume of the N-dimensional probabilistic space the LLM is following for that output is very low.
I agree with you on a technical level. I still think LLMs are transformative of the original text and if
when the number of sources that’s what ultimately created the volume of the N-dimensional probabilistic space they’re following is very low.
then the solution is to feed it even more relevant data. But I appreciate your perspective. I still disagree, but I respect your point of view.
I’ll give what you’ve written some more thought and maybe respond in greater depth later but I’m getting pulled away. Just wanted to say thanks for the detailed and thorough response.
A C compiler in two weeks is a difficult, but doable, grad school class project (especially if you use
lexandyaccinstead of hand-coding the parser). And I guarantee 80 hours of grad student time costs less than $20k.Frankly, I’m not impressed with the presentation in your anecdote at all.
Here is the original cite that my company pulled that from if you want more details.
I’ve never written a compiler, nor in Rust, so I have no idea the effort involved. I’m just boggling over the price tag. I’ll bet that’s the cost of an entire offshore team.
Yeah, the thing also has limited scope and requires some meddling to point to necessary includes as evidenced by the first issue, afair. And the code produced is subpar I heard
Agree with all points. Additionally, compilers are also incredibly well specified via ISO standards etc, and have multiple open source codebases available, eg GCC which is available in multiple builds and implementations for different versions of C and C++, and DQNEO/cc.go.
So there are many fully-functional and complete sources that Claude Cowork would have pulled routines and code from.
The vibe coded compiler is likely unmaintainable, so it can’t be updated when the spec changes even assuming it did work and was real. So you’d have to redo the entire thing. It’s silly.
Nah bro. Just tell the agent to be a super duper distinguished software developer and write no bugs and keep the code maintainable /s
Updates? You just vibecode a new compiler that follows the new spec
“I want to add a command line option that auto generates helloworld.exe”
“That’ll be $21,000.”
Ah, that’s the problem, we’ve been getting all these chatbots to generate “hellworld.exe”.
Also, software development is already the best possible use case for LLMs: you need to build something abiding by a set of rules (as in a literal language, lmao), and you can immediately test if it works.
In e.g. a legal use case instead, you can jerk off to the confident sounding text you generated, then you get chewed out by the judge for having hallucinated references. Even if you have a set of rules (laws) as a guardrails, you cannot immediately test what the AI generated - and if an expert needs to read and check everything in detail, then why not just do it themselves in the same amount of time.
We can go on to business, where the rules the AI can work inside are much looser, or healthcare, where the cost of failure is extremely high. And we are not even talking about responsibilities, official accountability for decisions.
I just don’t think what is claimed for AI is there. Maybe it will be, but I don’t see it as an organic continuation of the path we’re in. We might have another dot com boom when investors realize this - LLMs will be here to stay (same as the internet did), but they will not become AGI.
I also often get assigned projects where all the tests are written out beforehand and I can look at an existing implementation while I work…
I would be interested in knowing what language it was for sure, as there is a huge difference between a C and a C++ compiler in terms of complexity
I think this is the reported https://github.com/anthropics/claudes-c-compiler.
And here’s a pretty good article about it https://arstechnica.com/ai/2026/02/sixteen-claude-ai-agents-working-together-created-a-new-c-compiler/
I just posted where I found the source in another comment. It would have probably the information you’re interested in.
Don’t forget that there are tons of C compilers in the dataset already
Man, corporate layoffs kill productivity completely for me.
Once you do layoffs >50% of the job becomes performative bullshit to show you’re worth keeping, instead of building things the company actually needs to function and compete.
And the layoffs are random with a side helping of execs saving the people they have face time with.
Who?
The original creator of Twitter and now creator of Bluesky and whatever this thing that’s falling off the rails is.
Basically another billionaire living in his own little bubble and huffing his own farts too much.
He also had a lot to do with Nostr, early on.
Jack Dorsey, has endorsed and financially supported the development of Nostr by donating approximately $250,000 worth of Bitcoin to the developers of the project in 2023,[13][15] as well as a $10 million cash donation to a Nostr development collective in 2025.
What?
Oops I mistread my source. Have updated my comment.
he left Bluesky around 2 years ago
That must be why they are doing okay, haha.
Pretty much lol. If I remember correctly his reason for leaving was them adding moderation tools
Is that thumbnail a scene from 12 monkeys?
Naw. This is clearly just 1 monkey.
pffft, you give him too much credit.
Right before he dies, yeah
Uhhh, Block is the the parent company of Square (formerly known as Square Up). This is actually a huge company, not some little side thing.
Yeah not sure why they think Block is “new” they just renamed because they have a bunch of businesses beyond Square now.
Gonna need a longer beard.
