Posted by poisonfountain 6 hours ago
The fact that the author can articulate _why_ the AI is getting so good is kind of a moat for specialist, right? Imagine a layman prompting without domain expertise:
"There is likely a race condition here + [long-winded explanation and analysis carefully guiding the AI]"
Degenerates to:
"This button is not working, please fix. I don't care about code. Decide yourself"
Degenerates to:
"Claude make me money"
I just want to emphasise a point... Calculators give 100% correct answers and yet we still hire accountants; for the simple fact that we don't want all to be accountants.
People will hire software engineers for the simple fact that they do not want to be software engineers.
Calculators are not a replacement for accountants, online accounting services are in many cases. Which again can be run by an AI if they reach that level of reliability.
Today with LLMs this is still sci-fi, though.
But bread shops are available on every corner. Will software jobs become as common as bread shops? If yes, what happens to the salaries? Something to think about.
If we apply the same argument to software engineering I think it's a good point... just maybe not the one you intended to make.
It's probably impossible for LLMs to learn and apply that wisdom reliably.
Ask me how I know.
I wouldn't say it's particularly brave, in fact LLMs are probably better at identifying mistakes than most tax payers. The % of Americans using a CPA to file taxes is fairly small.
This is just how it is, and has always been in this industry. And it takes about 10 years to realize it.
When I started my career in software, businesses were still writing new code in COBOL. 10 years later those skills were pretty much useless, except for dwindling maintenance roles.
Then there was the client/server era. Then the web era. Then mobile. Then cloud, etc.
All the same functionality, written and re-written time and time again, using the latest popular stacks and methodologies.
I hope to be retiring in a few years and pretty much everything I have learned over nearly 40 years is no longer applicable or is at best losing relevancy to the way sofware is built today. And that's how it's always been.
SQL was first released in 1973. More new SQL is being written today than ever.
C++ (1985) is the de facto standard implementation language for web browsers, JavaScript engines, networking stacks, telecommunications, video games, high speed trading, CAD/CAM, video rendering and editing, audio processing, filesystems, databases, hardware drivers, automotive, aerospace, and robotics, among others.
Is Rust making inroads? Sure, and it's a tiny fraction of C++ still. It's a long ways from being the standard.
Likewise, Python is often cited as the "AI language," but that's on the surface -- CUDA, tensor libraries, inference languages, GPU kernels, compiler stacks, and so on are usually C++.
Then there's C -- introduced in 1972. Still widely used for greenfield in kernels, device drivers, embedded systems and microcontrollers, filesystems, firmware, network stacks, cryptography, databases, compilers.
LaTeX, MATLAB, Erlang, Verilog, PostScript, Lisp (including Scheme and Clojure), shell scripting (and the UNIX paradigm itself)... the list of old tech that still sees new projects in 2026 goes on.
I have no idea how things will play out, but so far I am not worried because the amount of software continues to increase, and AI only accelerates that trend. This will require the same mental modeling, first principles thinking, and relentless curiosity that already formed the foundation of the software engineer skillset.
Right now non-tech people just think AI will do anything they want and are the one in charge of hiring/firing, managing, etc. It's horrible to be a software dev right now, you've to deal with AI and lunatics.
Of course Domain Knowledge is important but, right now it's very hard to have reasonable conversation because... you know... AI this, AI that. I had a customer showing me a Claude vibe coded atrocity trying to convince me it's was a great app, now ask yourself: How are devs even supposed to collaborate with this without going insane? Simple, you can't.
There is a massive number of software engineers that are closer to plumbers than computer scientists and for them the progressing AI models are going to be a problem.
Yes, yes, 1000x yes.
As a bit of an aside, I have been toying with the idea of adding some sort of second pass/security auditing/scaling offering to my consultancy for people vibe coding projects which wind up generating interest. (Not sure what the fuck else I'm going to do!) I have a few non-technical friends who have found themselves in this situation and there's a real need for it.
The aspects of it which I find daunting are the ones you've referenced, though. I imagine many people -- especially the ones who've built mobile apps for $300 in tokens -- are going to balk at the costs I'd have to charge for such a service. We're also now living in an era where everyone is an "expert" (lunatic) ... with just a little help from Claude/Gemini/Grok/whatever. I can already foresee people second guessing every suggestion, decision, line item, etc. I'd also be taking on a liability that'd be tricky to completely work around via legal language for any bugs or security issues which might/would inevitably slip through review. Ironic because nobody blinks when LLMs excrete those things.
But, anyways, circling back around. Yeah, trying to find work in this market has been a new exercise in frustration. AI is all anyone wants to talk about, it's driven hourly rates through the floor and most of the open gigs revolve around model training and carry an implicit expiration window for the trainer. It sucks and I really don't know what I'm going to do to keep my consultancy open going forward. (As signs of how desperate I'm getting, I recently signed up for Task Rabbit and am seriously considering applying for a job at Tractor Supply.)
There might be a need for it but as a consultant your daily rate should be way above what a small vibe coder is willing to pay.
> As signs of how desperate I'm getting, I recently signed up for Task Rabbit and am seriously considering applying for a job at Tractor Supply.
I hope you'll find a way to keep going. Signing up for gig work is a race to the bottom though and not something I'd recommend. May I ask how you've arrived at this point?
Anything that can replace a deeply experienced s/ware engineer can replace anyone in the employment stack, meaning that only the owners of capital will be left, and they too will soon fade as the economy falls off a cliff and money has no value, because the only value that money has is the value of a human backing that, with thought, with ideas, with human output.
Whether you like it or not, "Economic output" is just a different phrase for "Human output that is valuable". When all human output is valued at the fractions of a penny per month of work, there is no future.
The people and companies leaning into fully autonomous agents are high on their own supply, in my opinion. They're kicking back in their beach chairs as their pipelines spit out Stüssy S after Stüssy S, with massive architectural flaws and attack surfaces, just lighting gobs of money on fire.
Yesterday, I created a fully functional POC for something really cool, it took me all day as I reshaped the agent's rough boilerplate ideas into usable components, and I never once hit a session limit on my $100 month Claude sub. I spent the majority of my time thinking about how I needed to prompt to turn what was in my head into working and secure code. You can't just give the agent a vague idea and expect anything less than a dumpster app.
It's probably enough to fool C Suite people into believing the AI apocalypse is coming, which is the crux of the problem, and what is fueling what is certainly a gigantic bubble — but when it comes down to it, shitty software is shitty software. To fix it, you have to know what good software looks and feels like under the hood, and why it's shitty or feels bad to use. It might start up, but mutability, staying up, remaining stable, and remaining secure are very different stories.
The clock is ticking on everything that has been developed by a LLM with a novice user behind the prompts.
Well, except for roles where being human is an inherent part of the value for customers: bartender, prostitute, certain kinds of boutique sales, professional athlete, stage actor, etc. And for roles that have to be human for legal reasons.
Of course such roles make up a small part of the entire job market.
Just because LLMs are good at translating English to code, doesn’t imply they are good at many other jobs.
Coding isn’t that hard, it’s just not enjoyable to most people. The enjoyment has always been the barrier to entry.
hard agree on the last statement. programming is language. if you're literate you can code.
Who said that?
More to the point, how many plumbers does society need?
Direct quote
> Direct quote
And, in your (and GP's mind), that means the same thing as "LLMs can replace plumbers"?
After all, I said:
>>>> When all human output is valued at the fractions of a penny per month of work, there is no future.
I mean, I know it's fashionable to not read the article, but are we all really responding without even reading the comments? Are two paragraphs well beyond the attention span of the readers here?
Okay, lets go with that asinine comeback: What do you think happens when the only work left for humans to do involves 100% physical labour and 0% thought?
How many plumbers does a society need? Electricians? Even in construction, you can automate almost everything away with cranes and similar.
Now imagine that all the doctors, all the office workers, all the warehouse workers, all the bankers, lawyers, teachers, ... basically any job that requires thought ... all those people are now joining the legions of plumbers.
That sort of 1000x increase in supply will drive prices to pennies.
The LLM doesn't need to replace plumbers directly; all it needs to do is replace everyone else, and the value of plumbers approach zero anyway.
I have zero doubt that half of humanity can all have jobs continuously expanding the mansions of the other half who don't do any work but receive all the benefits.
Software engineering was a nice target because inputs and outputs are just data and you don't need to figure out robotics. But idk, 3 years ago it seemed illusory (at least for me) that LLMs could take over software engineering, but now here we are. They are still not 100% there yet (software engineers still have jobs), but we are getting ever closer.
Companies are in the process of figuring out robotics, and even if it's not figured out, then we might introduce a gig-ified blue collar economy where an unskilled, underpaid gig worker implements instructions by AI. Plus a lot of blue collar work already today involves robots (cranes, excavators, trucks, etc).
Seems some on HN haven't been keeping up with progress in physical robotics. Unique physical work is lagging behind a bit, but not by much. Expect to see robots doing simple plumbing jobs within a few years, not a few decades.
Nope, just knowledge workers. We’re decades away from automating many manual labor professions, even “unskilled” ones.
Turns out brains just aren’t as special as we thought.
How do you figure? We’ve already automated away way more manual labor jobs than we currently have.
Nope, just a specific kind. Those who developed and cultivated only a very specific skill set at the expense of all others.
I used to think being a generalist, and having persued technical roles with a people facing element was to my detriment, but it’s turned out to be the best decision I ever made.
Being a generalist was very useful to me 5 years ago. Now AI models have made everyone a generalist. That wide but not terribly deep skillset was immediately devalued by the AI models.
You can argue that the models fuck up 20 percent of the time, or that they make poor code but there is a massive part for the industry that is totally fine with that and I think people ignore it to their detriment.
AI is fundamentally an equivalent to slave economy. Cheap, plentiful workforce. This time ethically neutral. You either get Greece or Rome. I’d prefer Greece but it will probably be Rome. From the past we can predict the future.
I’m starting to be more sensitive to the argument that without god, people are unable to have a strong moral foundation. Not for the people expressing creativity in how they fuck, but as a check on those in power.
In my own experience such people are often far from objectively moral or good people themselves, and overcompensate some deep issues.
It is very true in my experience. It is also very not true in my experience.
FWIW I’m an atheist. Curious what you mean by issues-riddled mind. What issues? What’s the unhealthy place? There is no one person I’d accuse of lacking morality through godlessness, but I do see a trend. Most particularly in the people and communities who would have previously chosen godliness and replaced it with nothing, not those who previously would have chosen godlessness.
If this were true, why did the medieval peasant have less rights and autonomy in society than we do now?
Also, I’m “starting to be more sensitive to” I’m not fully bought in.
I like to think that one of the symptoms is politics becoming really absolutist, idealistic and cultish. You do not debate followers of a different religion. But many topics really becoming kind of a mini religions.
I don’t know for sure though, there are arguments against it too and other factors.
I think substantial amount of people really need some kind of subjective spiritual experience to their life and maybe ignoring that need breeds some maladaptive tendencies
maybe thats a reason that god was deleted from the western cultural lexicon, so that broken communities could be capitalized upon? no way, surely god is merely a deprecated irrelevant vestige. it's not like a fractured social fabric is a ripe substrate of raw suffering to mine profit from. surely a few hundred generations were enough for our morals to have been encoded into genetics, we don't have to bother consciously practicing morality any more. that's for the narrow minded.
<alt version of above paragraphs from ludicrous perspective of individual experiencing theocracy and its own form of propaganda>
..... this isn't intended to be aimed at anyone except those who delete god to make money, and those who use god to make money. there's plenty of negative aspects to religion. the argument is intended to focus on the sheer idiocy of expecting morality to spontaneously manifest in the absence of external motivation or any teaching of lessons already collectively learned the hard way.
Concepts like "checking your privilege" or being "canceled" closely parallel religious ideas of original sin and repentance, where individuals must acknowledge their unearned moral failings to become "good".
Actions like using specific pronouns, displaying yard signs, or performing land acknowledgments function similarly to reciting a catechism; they signal allegiance to a shared belief system and reassure the in-group
Protests and social movements often evoke the communal, revival-like atmosphere of religious gatherings, providing participants with a sense of purpose and belonging.
But what’s most convincing is that many times it is hypocritical in the same way religions are. There is no room for questioning or doubt and yet the actions do not align with the performance. Which means it isn’t driven by dry results but fulfills a deeper human need.
It's the exact same story that we've heard countless times by now. Hosted on a blog with just a single post. Named in a way that suggests that said blog was created for this very single post.
What is there to learn from this other than LLMs seem to be bad for some people's psyches and that AI companies need these very stories to not get their funding shut down?
Would you put a "Hey i'm feeling a little useless" post on your main blog / linkedin?
It might be easier to adapt to this new tech when you're 19 compared to when you're 59.
But honestly, this discussion _also_ has happened ad-nauseam by now. Everything that was worth saying has been said. And then some.
People don't actually want to talk about LLMs. They want a hug. And that's fine, human and all.
But could you please just start asking for hugs instead of encoding that into vaguely profound sounding takes on AI? I'm tired of this play pretend.
How is that true? I've been using Opus on an industry scale over last 6 months and this is just not real.
It has consistently with a certain percentage of chance each time (and no claude.md and skills do not stop it fully):
* Suggested to remove tests to allow for things to pass
* Suggested remove an error so that things can be "unblocked"
* Suggested to use a second path when the original path ran into problem instead of making the original path accomodate for that possibility.
* Suggested or silently added "features" or "guardrail" that I don't want.
* Can be left unsupervised only if given a goal that it can verify against itself. Without such clear goal (e.g. this test in the integration environment must be fixed), it flounders.
I'm not using just the native harness (e.g. CC) either, with additional, customized harness, the behavior improves somewhat but are still fundamentally constrained and cannot really be trusted without verification.
See my methodology (100% handwritten): https://aperocky.com/blog/post.html?slug=agentic-development....
Being a heavy user I think I've ran into every single hallucination that the model can do over development release and operations. I am still a heavy user but there are a lot of value in recognizing where exactly LLM's limit is and work around that.
This reads like someone is trying to convince me, that ai is just this good, and that the author is telling me to use more ai.
To me this sounds like: Trust me, it’s really bad, i know what I’m talking about. Just lean into it, or change profession.
(Whether any one reading this, myself included, survives in the industry long enough to reach the other side of that transition is a different question.)
[EDIT] The reason I use books as an example is that 4.2 million books were published in 2025 (https://ideas.bkconnection.com/10-awful-truths-about-publish...); 3.5m self published (with most likely LLM assisted or wholly generated) and the remainder traditionally published. (That's ~9,600 new self-published books a day.) Who actually still sells enough copies to make money in this paradigm and why offers hints as to where the software industry is likely headed.