The problem is a management pattern:
removing people and organizational slack because they don’t generate immediate profit,
and then expecting the knowledge to still be there when it’s needed.
Short-term cost cutting leads to less junior hiring,
and removes the slack that experienced engineers need in order to teach.
As a result, tacit knowledge stops being transferred.
What remains is documentation and automation.
But documentation is not the same as field experience.
Automation is not the same as judgment.
Without people who have actually worked with the system,
you end up with a loss of tacit knowledge—and eventually, declining productivity.
AI is following the same pattern.
What AI is being sold as right now is not really productivity.
In many domains, productivity is already sufficient.
What’s being sold is workforce reduction.
The West has seen this before, especially in the case of General Electric.
GE pursued aggressive short-term financial optimization,
cutting costs, focusing on quarterly results, and maximizing shareholder returns.
In the process, it hollowed out its own long-term capabilities.
It effectively traded its future for short-term gains.
The same mindset is visible today.
The core problem is that decision-makers—often far removed from actual engineering work—
believe that tacit knowledge can be replaced with documentation, tools, and processes.ti cannot.
Tacit knowledge comes from direct experience with real systems over time.
If you remove the people and the learning pipeline,
that knowledge does not stay in the organization. It disappears.
You are spot on w.r.t every assertion you've made. When bean-counters took over the ecosystem they optimised immediate profitability over everything else. Which in turn means, in their mind, every part of the system needs to be firing at 100% all the time. There's no room for experimentation, repair, or anything else.
I've commented about lack of slack on several times here on HN because when I notice a broken system now a days, 90% of it is due to lack of slack in the system to absorb short term shocks.
I’ll note at the end of the last century I worked at IBM research which had a budget of 6 Billion dollars. Management was trying very hard to get better return on that investment. Even today IBM though often ridiculed in the tech space (sometimes they do deserve it) spends a lot on R&D.
IBM employees have garnered six Nobel Prizes, seven Turing Awards,
20 inductees into the U.S. National Inventors Hall of Fame, 19 National Medals of Technology,
five National Medals of Science and three Kavli Prizes. As of 2018,
the company had generated more patents than any other business in each of 25 consecutive years.
> the company had generated more patents than any other business in each of 25 consecutive years.
A couple things about those patents, from a former IBMer who has quite a few in his time there.
First, not all patents are created equal. Most of those IBM patents are software-related, and for pretty trivial stuff.
Second, most of those patents are generated by the rank and file employees, not research scientists. The IBM patent process is a well-oiled machine but they ain't exactly patenting transistor-level breakthroughs thousands of times a year.
>. In many domains, productivity is already sufficient. What’s being sold is workforce reduction.
This is a blindspot to many. People working on entrepreneurial projects need to build a lot. They start with nothing. They need (for example) features. There's a lot to do.
Most firms are not that. Visa, Salesforce, LinkedIn or whatnot. They have a product. They have features. They have been at it for a while. They also have resources. They are very often in a position of finding nails for a "write more software" hammer.
It's unintuitive because they all have big wishlist and to do lists and and a/b testing system for pouring software into but...
If there were known "make more software, make more money" opportunities available, they would have already done them.
Actual growth and new demand needs to come from arenas outside of this. Eg companies that suck at software(either making or acquiring) might be able to get the job done.
The Problem, bringing this back to the article, is fungibility. A lot of this "human capital" stuff cannot be easily repackaged. It's a "living" thing. Talent and skills pipelines can be cut off, and vanish.
A danger in Ai coding (and other fields) is that it leverages preexisting human capital and doesn't generate any for later.
> If there were known "make more software, make more money" opportunities available, they would have already done them.
Sometimes they're available, but not palatable, when the opportunity could threaten their existing comfortable way of doing things. Either by "self-cannibalism" or by changing the ecology so that the main product isn't so profitable.
Then the opportunities are ignored, or actively worked-against via lobbying, embrace-extend-extinguish, etc.
This sounds all true to me, but I think there is more. It is not just decisions by management, it is also the wider economic context. Low interest rates and, for the US, having the world reserve currency as your own currency both seem to make many of these changes attractive or even inevitable. Low interest rates lead to 'innovation' which I put in scare quotes because besides real innovation it can also mean something that passes as innovation but in the end just turns out to be a bubble of stuff that was not valuable enough. The 'innovation' then crowds out investments in more boring sectors like manufacturing. This is also not good for the population in general because fewer jobs are left for people who are not suited for working in highly 'innovative' sectors.
Why would anyone have a sight longer than a quarter? I mean how does long term thinking help the execs get their compensation this quarter? Sheesh..worst case scenario is that the work done now will benefit someone else when they've already left.
Also when companies grow big enough "business" becomes the main business of the company. By that I mean everything unrelated to the actual original domain, such as playing in the financial markets, doing stock buybacks, lobbying, cheating etc. When your CEO is an MBA and your real market is Wall Street any actual product RD and support is a real annoying cost that just cuts into the profits and thus into the exec compensation.
> ...any actual product RD and support is a real annoying cost that just cuts into the profits...
Worse, it might not generate a return. If you have enough profits, you just buy anyone who successfully produced something innovative. Let them take the risks. As Cisco used to say, "Silicon Valley is our R&D lab."
In the case of the military I'd say the real reason is political. After the fall of the Berlin wall, Europe collectively agreed (knowingly or not) that war is now a thing of the past and the goal should be the complete dismantling of militaries worldwide, starting with Europe. Lead by example, etc.
They agreed that war was a thing of the past, but still continued to push for NATO to allow new members anyway, ironically causing Russia (and China and everyone who is NOT in NATO) to suspect that war was NOT a thing of the past and therefore never quite abandoning their military completely. Unpopular opinion: the West should either NEVER have abandoned its military production (so as to maintain NATO actual preparedness for war, given that's the only reason for its existence) OR it should just have dismantled NATO and announced to the world that it strongly believes war is a thing of the past, and that other countries are advised to follow suit. But we actually chose the easy, halfway path: keep NATO, keep our militaries "looking strong" (which gives the signal our rivals should also do the same, obviously), but not actually be ready for any sort of major war and as the article points out, even lose actual capacity to become ready for war within any realistic timeframe. The worst possible outcome :(.
And the next level of this is, that even companies that realize this, mostly go ahead acting like this anyway, because they think someone else can train the juniors. Some other company will appear to do that, but nimby!
Over time the lack of good judgement will lead to a decline in their products' quality, which will be difficult to recover from.
Seems to me that - optimistically - this would shift the job of a software engineer into a more formal engineering role, and that the actual implementation is done by AI. In the same way in other areas, engineering and implementation differ and implementation can be (and is) automated.
No idea how this should take form, though, and if it’s even realistic. But it seems like due to AI, formal specs and all kinds of “old school” techniques are having a renaissance while we figure out how to distribute load between people and AI.
That sounds right, but it can be superbly wrong because that presupposes that you can debug what the AI gets very confidently wrong.
There are three legs to the stool: specification, implementation, and verification. Implementation and verification both take low-level knowledge and sophisticated knowledge of how things break.
Indeed, even if were possible for someone to create any program most of the time just by directing a team of AI agents, when something does not work one needs the ability to zoom in through the abstraction levels and understand exactly the program that is executed, so only knowing to generate prompts becomes insufficient.
This is the same with compilers. Most of the time a programmer needs to know only the high-level language that is used for writing the program. Nevertheless, when there is a subtle bug or just the desired performance cannot be reached, a programmer who also understands the machine language of the processor has a great advantage by being able to solve the bug or the performance problem, which without such knowledge would be solved in much more time or never.
> The problem is a management pattern: removing people and organizational slack because they don’t generate immediate profit, and then expecting the knowledge to still be there when it’s needed.
It's always seemed to me that the problem is corporate profit and personal profit above all. 'Management' is a subset of this, and so is pretty much everything else, including the current drive for AI.
It's the Western, perhaps American, approach to business and emphasived by MBAs and the media. Lowering costs, driving share price, dividends and corporate profit.
This race over the few decades has hollowed out most Western companies.
Listen to any entrepreneur podcast, or read any website, and it's all about 'how quickly can I get to exit', i.e. personal profit.
Capitalism is the worst form of economic system, apart from all the rest.
> What AI is being sold as right now is not really productivity. In many domains, productivity is already sufficient. What’s being sold is workforce reduction.
And workforce reduction is a nobel goal. In fact, I think it's one of the most important things humanity should focus on. We should strive for a workforce of zero. Humans currently was an enormous amount of their life working instead of more worthwhile pursuits.
I despise the rhetoric around this, we didn't "lose jobs" over AI, we saved ourselves a lot of work. What it does do is highlight a problem in our current society: the link between labour and the access to resources (e.g. money).
I don't think that AI is the ultimate answer to the problem of work, but it can contribute to it.
> The problem is a management pattern: removing people and organizational slack because they don’t generate immediate profit, and then expecting the knowledge to still be there when it’s needed.
I think that's still a symptom. The real problem is ideology: the monomaniacal focus on profit-making business, which infects our political leaders, down to capitalists and business leaders, down to the indoctrinated rank-and-file. Towards the end of the cold war, the last constraint on it were abolished, the the victory over the Soviet Union made it unquestioned.
The Chinese don't have that ideological problem. Their government appears to not give a shit about how much profit individual business make, they care about building out supply chains and a capabilities. They will bury the West, so long as the West remains in the thrall of libertarian business ideology.
The US is stuck in this weird irony where they recognize that Soviet-style central planning is a disaster but can't recognize that it's what megacorps do when they're insulated from competition. Internal politics, perverse incentives and a system that can sustain massive inefficiencies right up until the point that it doesn't.
In general productive economic activity generates a surplus and that surplus allows for slack. Human beings intuitively understand this. Hobbies are frequently de facto training for things that aren't currently happening but might later. Family-owned and operated businesses are much less likely to try to outsource their core competency for the sake of quarterly profits.
But regulatory capture and market consolidation causes the surplus to go to the corporate bureaucracies capturing the regulators instead of human beings with self-determination and goals other than number go up, and then the system optimizes for capturing the government rather than satisfying the people. "When you legislate buying and selling the first things to be bought and sold are the legislators." You throw away the competitive market and subject yourselves to the unaccountable bureaucracy, and then try to pretend it's not the same thing because this time the central planners are wearing business suits.
> megacorps do when they're insulated from competition. Internal politics, perverse incentives and a system that can sustain massive inefficiencies right up until the point that it doesn't.
That's the end stage. The bigger problem is the companies rotting from the inside even though they're still alive, because they use their resources to suppress your alternatives to them while they're slowly dying on top of you.
Yes, many Americans and other Westerners believe that the so-called "socialist" economies, like those of the Soviet Union and of Eastern Europe were non-capitalist.
This is only an illusion created by the fact that the communists were careful to rename all important things, to fool the weaker minds that the renamed things are something else than what they really are.
In reality, the "socialist" economies were more capitalist than the capitalist economies of USA and Western Europe. They behaved exactly like the final stage of capitalism, where monopolies control every market and there is no longer any competition.
Unfortunately, after a huge sequence of mergers and acquisitions started in the late nineties of the last century, the economies of USA and of the EU states resemble more and more every year the former socialist economies, instead of resembling the US and W. European economies of a few decades ago.
Everyone wants to tag the evil with their opposition's name. The evil is concentration of power. But no one wants to call it that because then they can't pretend that it's something different when they're doing it themselves.
Witness the people who keep proposing to solve market consolidation with higher taxes. Higher taxes go to the government, and therefore the interests that have captured the government. Are we going to solve it by taking money from Warren Buffet and giving it to Larry Ellison? Do we benefit from increased funding for Palantir? No, you have to break up the consolidated markets through some combination of antitrust enforcement and peeling back the regulatory capture that prevents new competitors from entering the market.
Yes - ultimately it's the same system. Far from being daring and innovatory, it's backward-looking, unimaginative, and bureaucratic.
Vision for the future is limited to grandiose fantasies straight out of 1950s pulps and the "heroic" creation of narcissistic corporations that are cynically extractive and treat employees and customers with equal contempt.
The differences which used to provide a convincing cover story - no single Great Leader, a functional consumer economy, votes that appear to make a difference - are being dismantled now.
What's left are the same mechanisms of total monitoring (updated with modern tech) and reality-denying totalitarian oppression, run for the exclusive benefit of a tiny oligarchy which self-selects the very worst people in the system.
I don't agree with "the west forgot how to make things", it moved supply chains for cheap consumer goods to asia, but in the B2B space a lot of things are manufactured in Europe: companies like Bosch, Volkswagen, ASML, Alstom and Airbus are cranking out extremely complicated machines that last many years in demanding environments. It's just a different level of value-add vs. low cost electronics (for instance).
I still code daily without any coding assistance mostly because I believe this is the way to not forget how things are done, even trivial things.
My main point against using AI is that I do not want to depend basically on anything when I'm in front of the screen (obviously not including, documentation, books, SO and alike).
I closely see people that are 100% dependent on AI for literally everything, even the most trivial daily tasks and I find that truly scarly because it means that brain effort drops drammatically to a minimum level. To be stolen mental effort is not a minor thing.
Giving away that at least for me means to become a dependent zombie. Knowledge comes basically from manual trial/error almost daily.
Technology being technology if anything has shown us that we can be pushed and manipulated in every single conceivable way. And in my opinion depending on AI is the ultimate way for companies to penetrate and manipulate a very delicate ability of a human being: to think and wonder about things.
I beg to differ, insofar as my own experience has been the exact opposite. I enjoy fixing other people's mistakes. And I especially enjoy outsmarting the LLMs. I find that I can obsessively breathe down the neck of an LLM for far longer than I could ever stay in the traditional flow state.
> I find that I can obsessively breathe down the neck of an LLM for far longer than I could ever stay in the traditional flow state.
I can do that too. Most programmers can.
That's because it requires less skill! Critiquing something is always easier than doing it.
I can literally keep an LLM fixing things forever by just saying things like "This is not scalable", or "this is not maintainable", or "this is not flexible" or "this is not robust", ... etc ad nausem.
That doesn't take skill at the level to actually write the software. For the market which is hoping to switch to mostly LLM coding, the prize they are eyeing is skill devaluation and not just, as many think, productivity gains.
They have no reason to double output, but they'd sure love to first halve the people employed, and then halve the salaries of those people (supply/demand + a glut of programmers in the market), and then halve salaries again because almost no skill necessary...
That's because it requires less skill! Critiquing something is always easier than doing it.
No, it was always the other way around. Mediocre programmers always wanted to rewrite everything because reading and understanding an existing codebase was always harder than writing some greenfield thing with a “modern language” or “modern libraries” or “modern idioms.” So they’d go and do that and end up with 100x the bugs.
I think I might enjoy it for a little bit and then become very depressed at the idea that it will never end, a future of fixing things that should never have been broken in the first place and which won't stay fixed.
Who really truly enjoys that and doesn't see it as a chore?
I find the real way to review other people's code is to program with it and then I start seeing where the problems are much more clearly. I would do a review and spot nothing important then start working on my own follow-on change and immediately run into issues.
The problem is the LLMs completely change the equation. Before LLMs, beyond very junior (needs serious coaching) levels, reviewing was typically faster than writing the code that was reviewed. With LLMs, writing code is orders of magnitude faster than reviewing it. We already see open source projects getting buried in LLM slop and you have to find the real human or at least carefully curated contributions among the slop.
I would not be surprised if many open source projects will outright stop taking PRs. I have had the same feeling several times - if I'm communicating with an LLM through the GitHub PR interface, I'd rather just directly talk to an LLM myself.
But ending PRs is going to be painful for acquiring new contributors and training more junior people. Hopefully the tooling will evolve. E.g. I'd love have a system where someone has to open an issue with a plan first and by approving you could give them a 'ticket' to open a single PR for that issue. Though I would be surprised if GitHub and others would create features that are essentially there to rein in Copilot etc.
I highly question the ability of companies to gauge the level of experience of any dev.
The distinction between junior, mid, senior, lead is a facade. It is a soft gradient that spans multiple areas, but is tainted and skewed by the technology du jour.
Technically you don't have to be an employed developer to become a senior developer. It boils down to your personal willingness to learn and invest time building.
What companies seek these days are people having the experience with (dysfunctional) organizational structure and working around the shortcomings of the organizations communication and funding patterns, nothing more.
Does that really make you senior or just politically versed?
The pattern shows up the most whenever failing software pokes holes in perception.
> Technically you don't have to be an employed developer to become a senior developer.
That's incredibly unlikely. Do you need to be an employed surgeon to become a senior (or whatever they call it) surgeon??
I very much doubt you can be senior without having actually spent years doing it professionally. The experience is everything, no book will give you the sort of understanding you need. That's unfortunately human nature, we are not capable to learn and internalize things simply from reading or watching others do it, we absolutely need to do it ourselves to truly learn. Didactic books always have exercises for this reason.
You can learn facts and techniques from books, obviously. But just because you've read a book about Michelin restaurants that you can now be a Michelin Chef.
Maybe they mean you can be not employed and build products yourself? Technically true, but that's like running your own surgeries or something, you're still doing surgery.
Analogies to other professions give your argument an air of legitimacy, with none.
There’s plenty of people in this world who are expert programmers without following any traditional path.
“Oh yeah, like who”, you say.
Con Kolivas, anaesthetist, work on kernel schedulers including the Staircase Deadline (RSDL) scheduler which was a precursor to the Completely Fair Scheduler in Linux and the Brain Fuck Scheduler and the ck Patchset.
>I read the Fogbank story and recognized it immediately. Not the nuclear material. The pattern. Build capability over decades. Find a cheaper substitute. Let the human pipeline atrophy. Enjoy the savings. Then watch it all collapse when a crisis demands what you optimized away.
>In defense, the substitute was the peace dividend. In software, it’s AI.
Before it was AI, the cheaper alternative was remote contract dev teams in Eastern Europe, right?
Not sure why that was ever the plan, as there are clearly not enough people.
Also over here, east of 15°E we were fired all the same.
I believe the plan is to quite simply "do less overall unless it's about AI", but everyone was waiting for others to start layoffs first.
I spent six months working part time and the decision makers made it clear that this is preferable for them long term. Beats getting fired, but I couldn't sustain this lifestyle - I'm frugal but not that frugal.
It had to be H1B Indians and outsourcing to India. As a European, I have seen some "Eastern European devs" around, sure. But they were not present at every company I worked with. Indians were. Quality-wise, it was always the same story, but I'm not going to elaborate. Everyone who is ready to accept it, knows what I would be saying anyway.
I saw academic rigor fall of a cliff in exchange for 'better job alignment' between end 80s when I had my first class after finishing highschool called 'Formal verification in software' on to beginning of the 2000s when I left giving the first class to new students 'Programming in Java'. All the 'teaching how to think' was replaced with 'how to get a well paying job'.
I remember same complaints about junior engineers copy pasting snippets of code from StackOverflow without understanding. And without curiosity to understand, without code review and mentorship from senior engineers they never grew to the senior level. But that is only some of them, others used StackOverflow to learn, did not use the snippets without understanding them first and properly adapting to their context, and they got good coaching in their teams and now have reached senior level from there. I see the same dynamic with LLMs, just more opportunities for both juniors to learn more by following up, and for seniors to to create tooling to enforce better architectur, test coverage and fault resiliency.
I think you're missing the point. Nobody removed people thanks to their SO copy-paste skills. If anything, more folks were hired to troubleshoot and sort out any copy pasta blunders (since you actually need working software, at the end of the day).
With LLMs this is no longer true - the thing can vibe a great deal before anyone notices that they have 100.000 lines of code doing what a focused, human reviewed and tested 10.000 lines can do. And as this goes on, it becomes increasingly more difficult for anyone to actually dig into and fix things in the 100.000 without the help of LLMs (thus adding even more slop on the pile).
There's a certain irony in that the article itself is quite clearly assisted by AI. Not a criticism per se as I don't have a problem with AI assistance, but food for thought given the material being commented on.
The tropes that AI introduces into articles are very noticeable, quite annoying, and very unnatural -- they unfortunately don't write well. It seems people use them to "polish" up their writing but in reality it would have read better if they hadn't.
My current pet peave is using period instead of comma, as in:
> My people lived the other side of this equation. Not the factory floor. The receiving end.
Ostensibly this is supposed to add gravitas, but it's very often done in places where that gravitas isn't needed, and it comes off as if I'm reading the script for an action movie trailer.
> The tropes that AI introduces into articles are very noticeable, quite annoying, and very unnatural -- they unfortunately don't write well.
Quite paradoxical: when its 8n purpose native language we can spot it a mile away but theres no shortage of engineers who claim how good the code output is.
Whatever the reason for the default tone of AI in english, it's still there when generating code. It makes me think that the senior engineers who claim that it produces awesome output just don't understand the specific programming language as a someone who thinks in it.
I am not sure if it is necessarily copied. A lot of influencer-style people used some of these patterns (periods, not X but Y). So I'm not sure who is copying who?
Made me stop reading a few paragraphs in. I don't have a "problem" in the ethical sense either, but as the sibling comment notes, the way LLMs write is rather grating. To make matters worse, a) people seem to use them to add pointless volume / "filler" to their texts, so now I have to wade through pages and pages of this stuff, and b) I have no easy way to distinguish between an article at least based on novel human insights vs entirely LLM-generated from a "write me something about X topic" prompt. I don't think it's a stretch to say that the latter just isn't worth reading given the state of the art.
I don't have a problem with AI assistance either, but this undermines the point the article is making. For me it is like a priest preaching gay sex is wrong and then being caught in bed with a male prostitute (snorting cocaine optional). Leaves bad taste in the mouth.
The text has few of the obvious AI tells. The only thing that, to me, looks characteristic of LLM-generated text is the short and terse sentence structure, but this has been a "prestigious" way to write in English since Hemingway.
Sort of a taste receptor I’m sure many have developed now.
The most obvious patterns here are: antithesis constructions, words choices and distribution, attempt at profundity in every paragraph but instead are runs of text that doing say anything, and even the perfect use of compound hyphenation. I think and can appreciate that there is definitely an attempt at personalization and guidance to make it less LLM-y and not just a default prompt, but it’s still kind of obvious. You could use a detector tool too of course.
People are not perfect. I went to Ukraine just days before the invasion. Travel and Hotels in Kiev had become extremely cheap. You asked the Ukrainians about the possible invasion. "Not going to happen" everybody said."Russia talks always aggressively, but never does anything".
They did not properly prepare and as a result lost 20% of its territory in days.
Days after that I was back is Austria and could not stop thinking about some of the people I spoke with being dead.
Since that I have also been in Dubai and Saudi Arabia as an entrepreneur and engineer. "What are you going to do when drones are used against your infrastructure?" If you followed the Russian war and first Iranian strike it was obvious that drones were going to be used against them. "not going to happen" again.
The have lost tens of billions for lacking proper preparation. They could have been protected spending just hundreds of millions of dollars over years.
I'd say that Ukraine were very prepared for the invasion, though? They managed to survive for the first 2 weeks, leading to a long-term war. The Donbas war had already been going on for 8 years, and I don't think Ukrainians were under some illusion that those weren't Russians.
On the flip side, all around the world you have "leaders" talking about imaginary conflicts with foreign countries that we must spend billions (they have a friend who really should get the contract) to prepare for and if the other side (tm) gets in your whole family will be killed instantly.
In hindsight, it's easy to be smart. You picked two examples where somebody said "never gonna happen" and then it happened. How about the countless examples where somebody said the same and then the thing actually didn't happen?
Take millions playing the lottery. To each of them, I can confidently say "you won't win, not gonna happen". For almost all of them I'll be right. There will be one who wins, were I was wrong, and they will say "see, told you so". That doesn't mean my prediction was wrong. It means you are having a reporting bias.
GP also probably had a sampling bias. The ones who were actually concerned about the impending Russian invasion presumably fled out of the country (or at least, away from the major cities to rural areas that probably see less fighting)
> They did not properly prepare and as a result lost 20% of its territory in days.
They did though. While nobody actually believed Putin would be dumb enough, the Ukrainian army was still, just in case, extremely busy on preparing defences, organising stockpiles, preparing defensive tactics.
Excellent post. Two stand-out points are deskilling through abolition of apprenticeship (or equivalent progression through the rank and responsibility), and loss of institutional knowledge, especially tacit knowledge stored in individual people. These are people problems more than they are technology problems. Without continuity of process and practice stuff gets lost. Sometimes change really is progress, for example software safety and security practices have progressed over the past 50 years, but other times change is just churn, or choices driven by misaligned incentives which will bite later, as the article describes.
The article makes no sense, and stars with a very wrong perspective on things.
This kind of forgetting is normal. It's how things work when time and resources are finite. The only problem here is the belief that you can keep capacity to do something without actively exercising it, and thus the expectation that you can "just" resume doing things after a long break, without paying up a cold-start cost.
But you can't, and there's no reason to be surprised. I bet the Pentagon and the EU weren't. They didn't need those Stingers and shells for decades, didn't expect to need them soon - but they knew they could get them if they really needed them, but it's gonna be costly.
I don't get why people think this is unusual or surprising, or somehow outrageous and proves something about society or "mindsets of elites" - other than positive aspects like adaptability and resilience.
This is true at all scales. Your body and brain optimizes aggressively, too. An individual saying "I need to warm up" or "I need to hit the gym a few times and then I'll be able", or "yes, I can, but I haven't done it for years so I need an hour with a book/documentation..." - all that is exactly the same as EU going "yes we can make artillery shells... though we haven't in a while so we need some time and some millions of EUR to get our supply chain sorted out first".
For that matter, a lot of human civilization has been about identifying things that were normal and making them rare. "Normal" infant mortality of 40%, famines, floods, history being lost, etc.
Anyway, when it comes to "this is normal" I think we should take care to distinguish between interpretations of:
1. "This specific case should not have taken certain people by surprise."
2. "This is a manifestation of a broader phenomenon."
3. "This is natural and therefore cannot or should not be solved." [Naturalistic fallacy.]
First of all this is clearly AI-assist writing (being charitable here).
And the premise makes no sense anyway. The only risk of forgetting how to make shells is when other countries are making shells more efficiently. Non-western countries are not going to reject AI-coding, nor are they going to make software more efficiently by hand.
You could say COBOL has had this "problem" for 40 years also. That's why we need to constantly be inventing new ways of making things. The old ways are always forgotten over time.
If you REALLY need something long-forgotten, then you have lazy-load it back into being at significant cost. That's the price of constant progress.
The point of the article is that sometimes the "old ways" really means "not particularly profitable or necessary in the short term" but the bill comes due in a crisis. The reason US/EU manufacturing was "the old ways" is that people could make easier money with financial engineering, an insight that extended all the way to Raytheon.
COBOL is a bad example, but higher-level languages vs. assembly is not. If you write a lot of C you really don't need to know assembly.... until you stumble across a weird gcc bug and have no clue where to look. If you write a lot of C# you don't really need to know anything about C... until your app is unusably slow because you were fuzzy on the whole stack / heap concept. Likewise with high-level SSGs and design frameworks when you don't know HTML/CSS fundamentals.
As the author says maybe AI is different. But with manufacturing we were absolutely confusing "comfortable development" with "progress." In Ukraine the bill came due, and the EU was not actually able to manufacture weapons on schedule. So people really should have read to the end of "building a C compiler with a team of Claudes":
The resulting compiler has nearly reached the limits of Opus’s abilities. I tried (hard!) to fix several of the above limitations but wasn’t fully successful. New features and bugfixes frequently broke existing functionality.
But these are hard IT things a human programmer really struggles with as well. What % of software written is that? Very very low. Most software is dull and requires business vagueness to be translated into deterministic logic and interfaces; LLMs are pretty great at that as it is. If humans use their old ways to fix complex problems and llms do the rest, we still only need a handful of those humans. For now.
"For now" is sort of the entire point of the article :)
Even in the Before Times, it was much cognitively cheaper to write code than it is to read someone else's code closely, or manage lots of independent code across a team, or to make a serious change to existing code. It's so much easier to just let everyone slap some slop on the pile and check off their user stories. I think it will take years to figure out exactly what the impact of LLMS on software is. But my hunch is that it'll do a lot of damage for incremental benefit.
With the sole exception of "LLMs are good at identifying C footguns," I have yet to see AI solve any real problems I've personally identified with the long-term development and maintenance of software. I only see them making things far worse in exchange for convenience. And I am not even slightly reassured by how often I've seen a GitHub project advertise thousands of test cases, then I read a sample of those test cases and 98% of them are either redundant or useless. Or the studies which suggest software engineers consistently overestimate the productivity benefits of AI, and psychologically are increasingly unable to handle manual programming. Or the chardet maintainer seemingly vibe-benchmarking his vibe-coded 7.0 rewrite when it was in reality a lot slower than the 6.0, and he's still digging through regression bugs. It feels like dozens of alarms are going off.
These are good point and I am not overestimating; we are simply seeing the productivity boost in our company and the rise in profitability. We practice TDD, but only at integration level, so we have tests upfront for api and frontend and the AI writes until it works. SOTA models are simply good enough not to do;
function add(a,b) = c // adds two numbers
test: add(1,2)=3
to implement
function add(a,b) return 3
So when you have enough tests (and we do), it will deliver quality. Having AI write the tests is mostly useless. But me writing the code is not necessarily better and certainly not faster for most cases our clients bring us.
Yes. Just like globalization created companies like TSMC, AI will do the same. Software engineers who don't rely on LLM code generators will have a moat because they can do it cheaply and sustainably.
Another reason is that LLMs train on the existing code we already know, don't expect new programming languages or frameworks this means that the software engineering skills that exist today will be relevant for a long time.
I am not so much convinced by your last point, that point of new languages and frameworks. I think the cutoff date is closing in on our current now. If models cannot easily become bigger, they will likely advertise using "up-to-date-ness". Maybe they will be merely a few days behind. Or bigger models will make use of smaller but more up-to-date models.
I think engineering skills will still remain relevant due to taste and proper judgement. A model trained on everything and the kitchen sink has probably not the fitting bias for given specific problems in my project. Accepting too much AI generated code without steering the ship will result in some drift of taste and ultimately make some mediocre project like done by people without good domain knowledge and without good taste. It might even be short term a business, but it lacks the long term excellence, that sets projects with good judgement apart from the common rabble.
> I think the cutoff date is closing in on our current now. If models cannot easily become bigger, they will likely advertise using "up-to-date-ness". Maybe they will be merely a few days behind. Or bigger models will make use of smaller but more up-to-date models
But they will still rely on assembly, C, Rust, Linux, HTML, TCP/IP... Doesn't matter how up to date they are, they rely on existing code they have been trained on, they can't just create new languages without the training data.
To be charitable to TFA, there are a dearth of accurate and well-understood labels for the kind of X versus Y they want to make between national economies.
Even "First/Third world" has been fraying at the edges for decades since it was originally about political alignment.
The West started to forget how to code a long time before AI. At first, it was the work visas to bring programmers in, then it was outsourcing. At this point, I'm not even sure if AI is doing more harm than good in this department as it might be able to bring some jobs back to the "West", if it turns out that it's cheaper than outsourcing.
The outsourcing was shedding more of the trivial jobs, while trying to keep key positions at home, but increasingly, it also started to lose the key positions too. It's possible that AI can make it so that the key positions will be harder to justify to outsource... but, who knows... maybe not.
I disagree with the premise - interesting but I interpret the same fact pattern differently.
The history of technology is the replacement of manual processes with automated ones.
Consider a very basic process: checkout of a restaurant.
Writing the price of each item on a sheet of paper, manually adding them and writing the total was replaced with typing in the prices and eventually with just pushing the button for the item. Paper still exists for jotting down your order but within seconds of leaving the table it’s transitioned to computer.
This has enabled lots of desirable advances- speed, accuracy, new payment rails, and increasingly, elimination of the server in checkout- you tap a credit card on a tabletop device.
Did we “forget” how to do checkout? No. We purposely changed it.
But if the internet connection goes down or the backend server powering the cash register app goes down, there is an atrophied and not-regularly exercised skill set (maybe not even trained, IDK) that has to be implemented on-the-fly and it’s slow and frustrating for everyone.
Businesses don’t exercise (or perhaps even train) this process because it’s just not needed enough to warrant the cost.
Military procurement of weapons systems is hardly the place to point to as a technological tradition. There are lots of cases where no one pays the money to keep a production process in place; the reasons are all related to shortsighted “cost savings” or failing to anticipate changing needs.
With coding today, we are seeing the same kind of shift in priorities as my restaurant example. Having humans write code in the 2020 (pre-GPT) tradition was extremely inefficient in terms of time-from-idea-to-implementation.
We’ve found a new way to do the mundane part of that task (the mechanics of translating spec to implementation).
We are figuring out how to do that while preserving quality (and a lot of it is learning how to specify appropriately).
Will we “forget” how to “build” code?
No, but the skills to generate source code by hand will atrophy just as the skills to draw blueprints by hand atrophied with the advent of CAD.
Will we find examples where someone prematurely optimized away knowledge of a skill or process, incorrectly thinking it was no longer needed? Of course.
But the productivity gains we get will be so great on average that no one will go back to doing things the old way.
There will be old-timers and hobbyists who will preserve some of that knowledge; for most it will just be a curiosity.
Everyone is taught at a young age how to do basic addition and multiplication. That's all check out requires. People are not taught at a young age how Rust lifetimes work or how to write human maintainable code.
I agree, as with everything in 2026, the reality lands somewhere in the middle of the discourse online. But pretending this is in practice anything like the check out example is wrong.
The point you seem to be missing is that focusing only on optimization makes us all fragile to system shocks.
> Businesses don’t exercise (or perhaps even train) this process because it’s just not needed enough to warrant the cost.
Until a crisis hits. Covid and supply chain failures. Iran war and straight of Hormuz. Prolonged War in Europe with no production pipeline available. Banks collapsing after unsustainable overleveraging in supposedly "safe" mortgages.
For every optimization and cost-saving measure that is deployed, there should be a backup plan in place. MBA types and "technologists" keep missing this. What is the backup plan for the case where most of the economy activity is built on software produced by business who overleveraged on LLM for code generation?
Though I do believe you are making them in good faith, I find those comparisons do not hold.
CAD still requires you know what to do, and without CAD you can still draw blueprints by hand because you know what the result should be. Checkout is basic arithmetic you can do on a paper or even your personal phone. In both cases it is clear what the process is and what the output should be, and it doesn’t replace knowledge and training and certification.
With coding, none of that is true. By and large, there is a trend of people who don’t know what they’re doing shitting out software, or people who should know better not verifying the very flawed output they get. That is already having negative consequences in people’s lives.
Speak for yourself. I now dare to code much harder problems and learning is bliss. No more having to sit down to dig needle-in-haystack through horrible documentation or random Stack Overflow posts.
LLMs are a magnificent tool if you use them correctly. They enable deep work like nothing before.
The problem is the education system focused on passivity (obeyance), memorization, and standardized testing. And worst of all, aiming for the lowest common denominator. So most people are mentally lazy and go for the easy win, almost cheating. You get school and interview cheating and vivecoders.
But it's not the only way to use LLMs.
Similarly, in Wikipedia you can spend hours reading banal pop-slop content or instead spend that time reading amazing articles about history, literature, arts, and science.
Even if you are the absolute unicorn who gets paid to "code much harder problems" and "learning", the rest of the industry exists to deliver actual products and services.
So unless you nurture some type of https://xkcd.com/208/ fantasy, this is not just about you. The industry as a whole needs to find a way to work with LLMs without automating programming away entirely, and the industry as a whole needs to find a way to ensure that newcomers are able to be productive even if code-generation tools are taken away from them.
There was a time when companies had terrible development practices and could forget how to build, test, and deploy software, but is anyone seeing that now? We have much better development practices nowadays.
It doesn’t seem much like defense industry problems.
>The combination of technical skill and the judgment to know when the AI is wrong barely exists in the market anymore.
I see a talent pipeline collapse in next 5 years. "Software engineering is over coding is a solved problem" as being chanted by semi literate media and the AI grifter's marketing departments would further scare away the allocation of human capital to software engineering easily commanding 3x rise in salaries due to resource shortage.
Yes that is one key that resonated with me. The author did a great job of putting these recurring concepts into their own words
The other that really resonated was something that I read before along the lines of… we think that once humanity learns something, that knowledge stays and we build on it. But it’s not true, knowledge is lost all the time. We need to actively work to keep knowledge alive
That’s why libraries and the internet archive are so important. Wikipedia, too
> Leadership qualities. Our last hiring round tells you how rare that is: 2,253 candidates, 2,069 disqualified, 4 hired. A 0.18% conversion rate.
It's minor but this is just wrong. If you're going to hire 4 candidates, there could be 2,253 perfectly qualified candidates even if only 0.18% get hired. The conversion rate is meaningless; it just tells us how many jobs were on offer. There is no way that the skills this fellow wanted were so rare and difficult that only 1/500 candidates could possibly handle the job. Humans even in the 1/20 mark are pretty competent if you're willing to train them and legitimate geniuses crop up at around 1/200.
This is some convoluted BS built on the premise that wars need to make sense, economically or otherwise. No, wars do not need to make sense. If a person, a dictator or a president, unilaterally starts a war that forfeits the lives of both the dictator's (possibly fabricated) enemies and its own people, that person is knowingly committing murder. Logically, such a person should be handled with at least as much prejudice as a lone wolf that opens fire on a crowd. So we need to fix our legal systems to be better at preventing wars, not our economic systems to be better at fighting them.
This will end with the way of COBOL with a few people that still have the expert-level understanding of refactoring old code without causing outages or service disruption.
We’ll see, but right now I now see developers 24/7 hooked onto their agents and in the future we will experience a de-skilling problem which clean code, best practices, security and avoiding NIH syndrome will be all flushed down the toilet.
While the Fogbank story is a funny anecdote, I don't see it as a fitting example for atrophied skills. It's like writing a clean implementation of some software and it just doesn't match the legacy version until you realize that the legacy version had an unnoticed bug that made it behave the way it does.
Isn't that is the point of technological civilization development? People for example forgot how to weave on the handloom, or all the parts production and the maintenance for the watermills. And wooden sailships - top mastery of handling and engineering developed for millennia, gone.
As it was said - the future is here, it just distributed non-uniformly, so somebody is still and will be for some time sailing, manufacturing things and writing code.
We have both forgotten how to make things and also decided we can make more profit letting someone else make everything for every market. We have moved to a generation fixated on maximizing profit. However there is logic there as the cost to access the ability to make things is prohibitively expensive. As someone who makes open hardware with a nod to the environment and reusability, you can not justify or even find more locally sourced options than China.
Coding is different though, coding doesn't have a cost barrier, it has a ability barrier. I think we will loose a lot of people who never were passionate about programming and perhaps go back to a happy equilibrium. AI is only production ready if you have someone who understands software development. AI will improve speed to market if you have the right team, it doesn't remove the need for some to learn to code. You will of course end up with startups using exclusively AI but they will be those who end up with major security breaches or simply cannot scale as the AI goes in the wrong direction for the future. Tbh that's probably a positive as it weeds out the start ups that are focused on buzzwords for funding and not product.
No matter what happens to the viability of software development as a career, I will always care about the craft as I have done the past twenty years and change. The imperatives to adopt LLMs in situations where they do not benefit me nor my work is what is driving me away. I have to agree with latexr; the people who seem to benefit the most from the current moment are those who see software as a means to an end without much concern for quality, longevity, nor customer experience.
Why is speed-to-market such an important metric? I do not understand the need to mimic the largest players in the industry, nor do I see any particularly profound long term benefits to first mover advantage.
> I think we will loose a lot of people who never were passionate about programming
Anecdotally, what I’m seeing right now is the opposite. People who don’t care about programming are joining, while those who do care are getting tired of the bullshit and leaving. The good programmers are the ones leaving, the hacks are extremely happy to use LLMs.
When shit hits the fan, there won’t be many people left to clean it.
Those examples are nonsensical. None of those are necessary to get working code. The VSCode example is particularly baffling. Firstly, I’m sure you understand there are other editors people use for code; secondly, I know even people who don’t code who have picked up VSCode for text editing and are fine with it.
I can't not write the tired comment of how ridiculous it is to criticize AI and then use AI to write your article. It's tired, but so is this writing style.
For the actual problem, I fear this can't be solved by warning people, the pain will need to be felt. The system we live in, basically free market capitalism, cannot do anything else except local optimization. Maybe it's for the best, I don't know. The alternative of top down planning wouldn't have this problem, but it would have other problems. I work for a mid size somewhat luxury brand, and the major goal right now is cost cutting and AI for efficiency everywhere instead of using it to create better products or better ways to reach out customers. When I think about who will buy our luxury products if all jobs were optimized out of existence, I don't have an answer, but again I think the pain will need to be felt to change course.
> After spending an additional $69 million and years of reverse engineering, they finally produced viable Fogbank. Then discovered the new batch was too pure. The original had contained an unintentional impurity that was critical to its function.
Same thing that happened to the unfortunate Dr. Jekyll!
when you offshore or automate away the hands-on knowledge, you don't just lose the workers, you lose the entire institutional memory, and no amount of money can buy that back overnight
The US manufacturing sector is the biggest it has ever been. Exports are at all time record highs. The only thing that declined about manufacturing is the jobs. We build way more than we ever did but with far fewer people.
What we did do is decide that basic items aren't worth it. Our capacity is limited, our labor pool is limited, expenses are high, it doesn't make sense to make trinkets when we can make complex high precision parts and devices.
But no, we did not forget how to make things. We chose to use our capacity in a smarter way.
> I run engineering teams in Ukraine. My people lived the other side of this equation. Not the factory floor. The receiving end.
With all due respect, but many european taxpayers help pay for Ukraine. I am not disagreeing on the premise of the West killing itself via systematic recessions - Trump invading Iran leading to inflation as an example - so a lot of things are going on that show a ton of incompetency both in the USA and the EU, but at the same time I also get question marks in my eyes when this criticism comes from a country that receives money from others. That money could instead go to make EU countries more competitive, for instance. I am not saying this should necessarily be the case, mind you; I fully understand the nature of Putin's imperialism. But we need to really consider all factors when it comes to strategic mistakes with regards to production - and that includes taking up debts all the time. There are always a few who benefit in war, just as they benefit from subsidies from taxpayers (inside and outside as well).
Ukraine is "receiving money from others"? We are benefactors of the Ukrainians' bravery and sacrifices. How much money could we have not spent if Hitler had been stopped in Czechoslovakia?
You are completely ignoring the argument of your parent comment. They are saying that money is being spent to the benefit and best interest of the spenders, that it’s not a handout.
You are, of course, free to disagree and make your point, but ignoring the argument does not advance the discussion.
There are some pretty substantial differences. Russia is on the strategic back foot here trying to figure out a way to stop NATO's advance. They've only turned to violence after long attempts at resolving the tension diplomatically and the US has been implacable. Putin's actually been pretty hesitant in his escalations so far; he's 70 and has a long history of trying to avoid war.
Hitler was more about wanting more land and resources for Germany, and he saw war as being a legitimate tool for achieving his aims that he deployed early and enthusiastically.
> There are some pretty substantial differences. Russia is on the strategic back foot here trying to figure out a way to stop NATO's advance.
His rationale for invading Ukraine was to "demilitarise and denazify" it. The NATO point seems largely be invented by people who dislike NATO in the west.
> They've only turned to violence after long attempts at resolving the tension diplomatically and the US has been implacable.
I hope the "tension" you are referring to was not the little green men taking over Crimea and the Donbas in 2014.
> Putin's actually been pretty hesitant in his escalations so far; he's 70 and has a long history of trying to avoid war.
This is a totally unseriousness statement. Can you remind me what Putin was doing in Syria again?
There's an english transcript [0] of his speech from when they went in up on the Kremin website. He opened with something like
> I will begin with what I said in my address on February 21, 2022. I spoke about our biggest concerns and worries, and about the fundamental threats which irresponsible Western politicians created for Russia consistently, rudely and unceremoniously from year to year. I am referring to the eastward expansion of NATO, which is moving its military infrastructure ever closer to the Russian border.
They're claiming the NATO thing is relevant. Opening paragraph justification.
> Russia is on the strategic back foot here trying to figure out a way to stop NATO's advance. They've only turned to violence after long attempts at resolving the tension diplomatically and the US has been implacable. Putin's actually been pretty hesitant in his escalations so far; he's 70 and has a long history of trying to avoid war.
Is that why Russians rejected negotiations when Ukraine offered to never join NATO and Russians insist on keeping invaded territories?
Rather bad premise in the article.
1.) Germany, Italy and Eastern Europe are very industrial regions. The author forgets defence is not only the industry.
2.) The author doesn't show any source that Chinese developers don't use AI
As an anecdotal evidence I code way more now with agents because i have an entity who has vast amount of knowledge about pretty much everything and I have the creativity to use that well.
But you already knew how to code before LLM coding agents, juniors will jump straight into using agents without learning to code by hand, hence the premise of the article.
The real issue, in my view, is not AI itself.
The problem is a management pattern: removing people and organizational slack because they don’t generate immediate profit, and then expecting the knowledge to still be there when it’s needed.
Short-term cost cutting leads to less junior hiring, and removes the slack that experienced engineers need in order to teach. As a result, tacit knowledge stops being transferred.
What remains is documentation and automation.
But documentation is not the same as field experience. Automation is not the same as judgment. Without people who have actually worked with the system, you end up with a loss of tacit knowledge—and eventually, declining productivity.
AI is following the same pattern.
What AI is being sold as right now is not really productivity. In many domains, productivity is already sufficient. What’s being sold is workforce reduction.
The West has seen this before, especially in the case of General Electric.
GE pursued aggressive short-term financial optimization, cutting costs, focusing on quarterly results, and maximizing shareholder returns. In the process, it hollowed out its own long-term capabilities. It effectively traded its future for short-term gains.
The same mindset is visible today.
The core problem is that decision-makers—often far removed from actual engineering work— believe that tacit knowledge can be replaced with documentation, tools, and processes.ti cannot.
Tacit knowledge comes from direct experience with real systems over time. If you remove the people and the learning pipeline, that knowledge does not stay in the organization. It disappears.
> removing people and organizational slack
You are spot on w.r.t every assertion you've made. When bean-counters took over the ecosystem they optimised immediate profitability over everything else. Which in turn means, in their mind, every part of the system needs to be firing at 100% all the time. There's no room for experimentation, repair, or anything else.
I've commented about lack of slack on several times here on HN because when I notice a broken system now a days, 90% of it is due to lack of slack in the system to absorb short term shocks.
I’ll note at the end of the last century I worked at IBM research which had a budget of 6 Billion dollars. Management was trying very hard to get better return on that investment. Even today IBM though often ridiculed in the tech space (sometimes they do deserve it) spends a lot on R&D.
Lucent at the same time went through the same issue: how to monetise Bell Labs.
Bell Labs greatest work came out when AT&T was a monopoly. Once they were broken up (1984?) they started feeling the pain.
When the Lucent spinoff took place, the new entities had no Monopoly money to fund unconstrained research while management's behaviour never changed.
I don't know how BL fared under Alcatel and now Nokia, but haven't heard of anything interesting for years.
Did anything come out from those billions?
> Did anything come out from those billions?
Per wikipedia:
> the company had generated more patents than any other business in each of 25 consecutive years.
A couple things about those patents, from a former IBMer who has quite a few in his time there.
First, not all patents are created equal. Most of those IBM patents are software-related, and for pretty trivial stuff.
Second, most of those patents are generated by the rank and file employees, not research scientists. The IBM patent process is a well-oiled machine but they ain't exactly patenting transistor-level breakthroughs thousands of times a year.
> When bean-counters took over the ecosystem [...] in their mind, every part of the system needs to be firing at 100% all the time.
This is only fair, because they themselves are firing at 100% all the time IYKWIM ;)
>. In many domains, productivity is already sufficient. What’s being sold is workforce reduction.
This is a blindspot to many. People working on entrepreneurial projects need to build a lot. They start with nothing. They need (for example) features. There's a lot to do.
Most firms are not that. Visa, Salesforce, LinkedIn or whatnot. They have a product. They have features. They have been at it for a while. They also have resources. They are very often in a position of finding nails for a "write more software" hammer.
It's unintuitive because they all have big wishlist and to do lists and and a/b testing system for pouring software into but...
If there were known "make more software, make more money" opportunities available, they would have already done them.
Actual growth and new demand needs to come from arenas outside of this. Eg companies that suck at software(either making or acquiring) might be able to get the job done.
The Problem, bringing this back to the article, is fungibility. A lot of this "human capital" stuff cannot be easily repackaged. It's a "living" thing. Talent and skills pipelines can be cut off, and vanish.
A danger in Ai coding (and other fields) is that it leverages preexisting human capital and doesn't generate any for later.
> If there were known "make more software, make more money" opportunities available, they would have already done them.
Sometimes they're available, but not palatable, when the opportunity could threaten their existing comfortable way of doing things. Either by "self-cannibalism" or by changing the ecology so that the main product isn't so profitable.
Then the opportunities are ignored, or actively worked-against via lobbying, embrace-extend-extinguish, etc.
This sounds all true to me, but I think there is more. It is not just decisions by management, it is also the wider economic context. Low interest rates and, for the US, having the world reserve currency as your own currency both seem to make many of these changes attractive or even inevitable. Low interest rates lead to 'innovation' which I put in scare quotes because besides real innovation it can also mean something that passes as innovation but in the end just turns out to be a bubble of stuff that was not valuable enough. The 'innovation' then crowds out investments in more boring sectors like manufacturing. This is also not good for the population in general because fewer jobs are left for people who are not suited for working in highly 'innovative' sectors.
Why would anyone have a sight longer than a quarter? I mean how does long term thinking help the execs get their compensation this quarter? Sheesh..worst case scenario is that the work done now will benefit someone else when they've already left.
Also when companies grow big enough "business" becomes the main business of the company. By that I mean everything unrelated to the actual original domain, such as playing in the financial markets, doing stock buybacks, lobbying, cheating etc. When your CEO is an MBA and your real market is Wall Street any actual product RD and support is a real annoying cost that just cuts into the profits and thus into the exec compensation.
> Why would anyone have a sight longer than a quarter? I mean how does long term thinking help the execs get their compensation this quarter?
Vesting schedules, conditional grants, contractual equity ownership requirements
> ...any actual product RD and support is a real annoying cost that just cuts into the profits...
Worse, it might not generate a return. If you have enough profits, you just buy anyone who successfully produced something innovative. Let them take the risks. As Cisco used to say, "Silicon Valley is our R&D lab."
It is a very difficult mindset to argue against.
In the case of the military I'd say the real reason is political. After the fall of the Berlin wall, Europe collectively agreed (knowingly or not) that war is now a thing of the past and the goal should be the complete dismantling of militaries worldwide, starting with Europe. Lead by example, etc.
It's subtler than that. Europe was just constantly reminded by its big brother not to duplicate NATO structures, which are dependent on the US.
They agreed that war was a thing of the past, but still continued to push for NATO to allow new members anyway, ironically causing Russia (and China and everyone who is NOT in NATO) to suspect that war was NOT a thing of the past and therefore never quite abandoning their military completely. Unpopular opinion: the West should either NEVER have abandoned its military production (so as to maintain NATO actual preparedness for war, given that's the only reason for its existence) OR it should just have dismantled NATO and announced to the world that it strongly believes war is a thing of the past, and that other countries are advised to follow suit. But we actually chose the easy, halfway path: keep NATO, keep our militaries "looking strong" (which gives the signal our rivals should also do the same, obviously), but not actually be ready for any sort of major war and as the article points out, even lose actual capacity to become ready for war within any realistic timeframe. The worst possible outcome :(.
And the next level of this is, that even companies that realize this, mostly go ahead acting like this anyway, because they think someone else can train the juniors. Some other company will appear to do that, but nimby! Over time the lack of good judgement will lead to a decline in their products' quality, which will be difficult to recover from.
Seems to me that - optimistically - this would shift the job of a software engineer into a more formal engineering role, and that the actual implementation is done by AI. In the same way in other areas, engineering and implementation differ and implementation can be (and is) automated.
No idea how this should take form, though, and if it’s even realistic. But it seems like due to AI, formal specs and all kinds of “old school” techniques are having a renaissance while we figure out how to distribute load between people and AI.
That sounds right, but it can be superbly wrong because that presupposes that you can debug what the AI gets very confidently wrong.
There are three legs to the stool: specification, implementation, and verification. Implementation and verification both take low-level knowledge and sophisticated knowledge of how things break.
Indeed, even if were possible for someone to create any program most of the time just by directing a team of AI agents, when something does not work one needs the ability to zoom in through the abstraction levels and understand exactly the program that is executed, so only knowing to generate prompts becomes insufficient.
This is the same with compilers. Most of the time a programmer needs to know only the high-level language that is used for writing the program. Nevertheless, when there is a subtle bug or just the desired performance cannot be reached, a programmer who also understands the machine language of the processor has a great advantage by being able to solve the bug or the performance problem, which without such knowledge would be solved in much more time or never.
> The problem is a management pattern: removing people and organizational slack because they don’t generate immediate profit, and then expecting the knowledge to still be there when it’s needed.
It's always seemed to me that the problem is corporate profit and personal profit above all. 'Management' is a subset of this, and so is pretty much everything else, including the current drive for AI.
It's the Western, perhaps American, approach to business and emphasived by MBAs and the media. Lowering costs, driving share price, dividends and corporate profit.
This race over the few decades has hollowed out most Western companies.
Listen to any entrepreneur podcast, or read any website, and it's all about 'how quickly can I get to exit', i.e. personal profit.
Capitalism is the worst form of economic system, apart from all the rest.
See also https://georgzoeller.com/blog/posts/the-tech-job-apocalypse-...
You sound convincing, but it also reads very AI generated. A lot of people will stop reading half way.
> What AI is being sold as right now is not really productivity. In many domains, productivity is already sufficient. What’s being sold is workforce reduction.
And workforce reduction is a nobel goal. In fact, I think it's one of the most important things humanity should focus on. We should strive for a workforce of zero. Humans currently was an enormous amount of their life working instead of more worthwhile pursuits.
I despise the rhetoric around this, we didn't "lose jobs" over AI, we saved ourselves a lot of work. What it does do is highlight a problem in our current society: the link between labour and the access to resources (e.g. money).
I don't think that AI is the ultimate answer to the problem of work, but it can contribute to it.
> The problem is a management pattern: removing people and organizational slack because they don’t generate immediate profit, and then expecting the knowledge to still be there when it’s needed.
I think that's still a symptom. The real problem is ideology: the monomaniacal focus on profit-making business, which infects our political leaders, down to capitalists and business leaders, down to the indoctrinated rank-and-file. Towards the end of the cold war, the last constraint on it were abolished, the the victory over the Soviet Union made it unquestioned.
The Chinese don't have that ideological problem. Their government appears to not give a shit about how much profit individual business make, they care about building out supply chains and a capabilities. They will bury the West, so long as the West remains in the thrall of libertarian business ideology.
The US is stuck in this weird irony where they recognize that Soviet-style central planning is a disaster but can't recognize that it's what megacorps do when they're insulated from competition. Internal politics, perverse incentives and a system that can sustain massive inefficiencies right up until the point that it doesn't.
In general productive economic activity generates a surplus and that surplus allows for slack. Human beings intuitively understand this. Hobbies are frequently de facto training for things that aren't currently happening but might later. Family-owned and operated businesses are much less likely to try to outsource their core competency for the sake of quarterly profits.
But regulatory capture and market consolidation causes the surplus to go to the corporate bureaucracies capturing the regulators instead of human beings with self-determination and goals other than number go up, and then the system optimizes for capturing the government rather than satisfying the people. "When you legislate buying and selling the first things to be bought and sold are the legislators." You throw away the competitive market and subject yourselves to the unaccountable bureaucracy, and then try to pretend it's not the same thing because this time the central planners are wearing business suits.
> megacorps do when they're insulated from competition. Internal politics, perverse incentives and a system that can sustain massive inefficiencies right up until the point that it doesn't.
You just described Lucent.
That's the end stage. The bigger problem is the companies rotting from the inside even though they're still alive, because they use their resources to suppress your alternatives to them while they're slowly dying on top of you.
Yes, many Americans and other Westerners believe that the so-called "socialist" economies, like those of the Soviet Union and of Eastern Europe were non-capitalist.
This is only an illusion created by the fact that the communists were careful to rename all important things, to fool the weaker minds that the renamed things are something else than what they really are.
In reality, the "socialist" economies were more capitalist than the capitalist economies of USA and Western Europe. They behaved exactly like the final stage of capitalism, where monopolies control every market and there is no longer any competition.
Unfortunately, after a huge sequence of mergers and acquisitions started in the late nineties of the last century, the economies of USA and of the EU states resemble more and more every year the former socialist economies, instead of resembling the US and W. European economies of a few decades ago.
Everyone wants to tag the evil with their opposition's name. The evil is concentration of power. But no one wants to call it that because then they can't pretend that it's something different when they're doing it themselves.
Witness the people who keep proposing to solve market consolidation with higher taxes. Higher taxes go to the government, and therefore the interests that have captured the government. Are we going to solve it by taking money from Warren Buffet and giving it to Larry Ellison? Do we benefit from increased funding for Palantir? No, you have to break up the consolidated markets through some combination of antitrust enforcement and peeling back the regulatory capture that prevents new competitors from entering the market.
Yes - ultimately it's the same system. Far from being daring and innovatory, it's backward-looking, unimaginative, and bureaucratic.
Vision for the future is limited to grandiose fantasies straight out of 1950s pulps and the "heroic" creation of narcissistic corporations that are cynically extractive and treat employees and customers with equal contempt.
The differences which used to provide a convincing cover story - no single Great Leader, a functional consumer economy, votes that appear to make a difference - are being dismantled now.
What's left are the same mechanisms of total monitoring (updated with modern tech) and reality-denying totalitarian oppression, run for the exclusive benefit of a tiny oligarchy which self-selects the very worst people in the system.
West: We need profits and then we’ll try to build something useful.
China: We need to build this useful thing and then later let’s try to make profits, too.
> The real issue, in my view, is not AI itself
in shootings technically the guns are not the issue since they dont fire on their own.. they do enable the ability to shoot though
I don't agree with "the west forgot how to make things", it moved supply chains for cheap consumer goods to asia, but in the B2B space a lot of things are manufactured in Europe: companies like Bosch, Volkswagen, ASML, Alstom and Airbus are cranking out extremely complicated machines that last many years in demanding environments. It's just a different level of value-add vs. low cost electronics (for instance).
I still code daily without any coding assistance mostly because I believe this is the way to not forget how things are done, even trivial things.
My main point against using AI is that I do not want to depend basically on anything when I'm in front of the screen (obviously not including, documentation, books, SO and alike).
I closely see people that are 100% dependent on AI for literally everything, even the most trivial daily tasks and I find that truly scarly because it means that brain effort drops drammatically to a minimum level. To be stolen mental effort is not a minor thing.
Giving away that at least for me means to become a dependent zombie. Knowledge comes basically from manual trial/error almost daily.
Technology being technology if anything has shown us that we can be pushed and manipulated in every single conceivable way. And in my opinion depending on AI is the ultimate way for companies to penetrate and manipulate a very delicate ability of a human being: to think and wonder about things.
> They can’t tell you what the AI got wrong.
AI code generators are trolls. They confidently plausible content which is partly wrong. Then humans try to find their errors.
This is not fun. It has no flow.
I beg to differ, insofar as my own experience has been the exact opposite. I enjoy fixing other people's mistakes. And I especially enjoy outsmarting the LLMs. I find that I can obsessively breathe down the neck of an LLM for far longer than I could ever stay in the traditional flow state.
> I find that I can obsessively breathe down the neck of an LLM for far longer than I could ever stay in the traditional flow state.
I can do that too. Most programmers can.
That's because it requires less skill! Critiquing something is always easier than doing it.
I can literally keep an LLM fixing things forever by just saying things like "This is not scalable", or "this is not maintainable", or "this is not flexible" or "this is not robust", ... etc ad nausem.
That doesn't take skill at the level to actually write the software. For the market which is hoping to switch to mostly LLM coding, the prize they are eyeing is skill devaluation and not just, as many think, productivity gains.
They have no reason to double output, but they'd sure love to first halve the people employed, and then halve the salaries of those people (supply/demand + a glut of programmers in the market), and then halve salaries again because almost no skill necessary...
That's because it requires less skill! Critiquing something is always easier than doing it.
No, it was always the other way around. Mediocre programmers always wanted to rewrite everything because reading and understanding an existing codebase was always harder than writing some greenfield thing with a “modern language” or “modern libraries” or “modern idioms.” So they’d go and do that and end up with 100x the bugs.
I think I might enjoy it for a little bit and then become very depressed at the idea that it will never end, a future of fixing things that should never have been broken in the first place and which won't stay fixed.
Perhaps you have the psychological make up to thrive in this new environment. Glad it is working for you.
It should have the same flow as reviewing PRs from humans.
Who really truly enjoys that and doesn't see it as a chore?
I find the real way to review other people's code is to program with it and then I start seeing where the problems are much more clearly. I would do a review and spot nothing important then start working on my own follow-on change and immediately run into issues.
The problem is the LLMs completely change the equation. Before LLMs, beyond very junior (needs serious coaching) levels, reviewing was typically faster than writing the code that was reviewed. With LLMs, writing code is orders of magnitude faster than reviewing it. We already see open source projects getting buried in LLM slop and you have to find the real human or at least carefully curated contributions among the slop.
I would not be surprised if many open source projects will outright stop taking PRs. I have had the same feeling several times - if I'm communicating with an LLM through the GitHub PR interface, I'd rather just directly talk to an LLM myself.
But ending PRs is going to be painful for acquiring new contributors and training more junior people. Hopefully the tooling will evolve. E.g. I'd love have a system where someone has to open an issue with a plan first and by approving you could give them a 'ticket' to open a single PR for that issue. Though I would be surprised if GitHub and others would create features that are essentially there to rein in Copilot etc.
I highly question the ability of companies to gauge the level of experience of any dev.
The distinction between junior, mid, senior, lead is a facade. It is a soft gradient that spans multiple areas, but is tainted and skewed by the technology du jour.
Technically you don't have to be an employed developer to become a senior developer. It boils down to your personal willingness to learn and invest time building.
What companies seek these days are people having the experience with (dysfunctional) organizational structure and working around the shortcomings of the organizations communication and funding patterns, nothing more.
Does that really make you senior or just politically versed?
The pattern shows up the most whenever failing software pokes holes in perception.
> Technically you don't have to be an employed developer to become a senior developer.
That's incredibly unlikely. Do you need to be an employed surgeon to become a senior (or whatever they call it) surgeon??
I very much doubt you can be senior without having actually spent years doing it professionally. The experience is everything, no book will give you the sort of understanding you need. That's unfortunately human nature, we are not capable to learn and internalize things simply from reading or watching others do it, we absolutely need to do it ourselves to truly learn. Didactic books always have exercises for this reason.
You can learn facts and techniques from books, obviously. But just because you've read a book about Michelin restaurants that you can now be a Michelin Chef.
Maybe they mean you can be not employed and build products yourself? Technically true, but that's like running your own surgeries or something, you're still doing surgery.
Analogies to other professions give your argument an air of legitimacy, with none.
There’s plenty of people in this world who are expert programmers without following any traditional path.
“Oh yeah, like who”, you say.
Con Kolivas, anaesthetist, work on kernel schedulers including the Staircase Deadline (RSDL) scheduler which was a precursor to the Completely Fair Scheduler in Linux and the Brain Fuck Scheduler and the ck Patchset.
>I read the Fogbank story and recognized it immediately. Not the nuclear material. The pattern. Build capability over decades. Find a cheaper substitute. Let the human pipeline atrophy. Enjoy the savings. Then watch it all collapse when a crisis demands what you optimized away.
>In defense, the substitute was the peace dividend. In software, it’s AI.
Before it was AI, the cheaper alternative was remote contract dev teams in Eastern Europe, right?
Not sure why that was ever the plan, as there are clearly not enough people.
Also over here, east of 15°E we were fired all the same.
I believe the plan is to quite simply "do less overall unless it's about AI", but everyone was waiting for others to start layoffs first.
I spent six months working part time and the decision makers made it clear that this is preferable for them long term. Beats getting fired, but I couldn't sustain this lifestyle - I'm frugal but not that frugal.
Happy to help and eventually take over.
It had to be H1B Indians and outsourcing to India. As a European, I have seen some "Eastern European devs" around, sure. But they were not present at every company I worked with. Indians were. Quality-wise, it was always the same story, but I'm not going to elaborate. Everyone who is ready to accept it, knows what I would be saying anyway.
India for the most part.
I saw academic rigor fall of a cliff in exchange for 'better job alignment' between end 80s when I had my first class after finishing highschool called 'Formal verification in software' on to beginning of the 2000s when I left giving the first class to new students 'Programming in Java'. All the 'teaching how to think' was replaced with 'how to get a well paying job'.
I remember same complaints about junior engineers copy pasting snippets of code from StackOverflow without understanding. And without curiosity to understand, without code review and mentorship from senior engineers they never grew to the senior level. But that is only some of them, others used StackOverflow to learn, did not use the snippets without understanding them first and properly adapting to their context, and they got good coaching in their teams and now have reached senior level from there. I see the same dynamic with LLMs, just more opportunities for both juniors to learn more by following up, and for seniors to to create tooling to enforce better architectur, test coverage and fault resiliency.
I think you're missing the point. Nobody removed people thanks to their SO copy-paste skills. If anything, more folks were hired to troubleshoot and sort out any copy pasta blunders (since you actually need working software, at the end of the day).
With LLMs this is no longer true - the thing can vibe a great deal before anyone notices that they have 100.000 lines of code doing what a focused, human reviewed and tested 10.000 lines can do. And as this goes on, it becomes increasingly more difficult for anyone to actually dig into and fix things in the 100.000 without the help of LLMs (thus adding even more slop on the pile).
There's a certain irony in that the article itself is quite clearly assisted by AI. Not a criticism per se as I don't have a problem with AI assistance, but food for thought given the material being commented on.
The tropes that AI introduces into articles are very noticeable, quite annoying, and very unnatural -- they unfortunately don't write well. It seems people use them to "polish" up their writing but in reality it would have read better if they hadn't.
My current pet peave is using period instead of comma, as in:
> My people lived the other side of this equation. Not the factory floor. The receiving end.
Ostensibly this is supposed to add gravitas, but it's very often done in places where that gravitas isn't needed, and it comes off as if I'm reading the script for an action movie trailer.
> The tropes that AI introduces into articles are very noticeable, quite annoying, and very unnatural -- they unfortunately don't write well.
Quite paradoxical: when its 8n purpose native language we can spot it a mile away but theres no shortage of engineers who claim how good the code output is.
Whatever the reason for the default tone of AI in english, it's still there when generating code. It makes me think that the senior engineers who claim that it produces awesome output just don't understand the specific programming language as a someone who thinks in it.
People have also started copying the AI tropes, especially your period/comma example.
I am not sure if it is necessarily copied. A lot of influencer-style people used some of these patterns (periods, not X but Y). So I'm not sure who is copying who?
Made me stop reading a few paragraphs in. I don't have a "problem" in the ethical sense either, but as the sibling comment notes, the way LLMs write is rather grating. To make matters worse, a) people seem to use them to add pointless volume / "filler" to their texts, so now I have to wade through pages and pages of this stuff, and b) I have no easy way to distinguish between an article at least based on novel human insights vs entirely LLM-generated from a "write me something about X topic" prompt. I don't think it's a stretch to say that the latter just isn't worth reading given the state of the art.
I don't have a problem with AI assistance either, but this undermines the point the article is making. For me it is like a priest preaching gay sex is wrong and then being caught in bed with a male prostitute (snorting cocaine optional). Leaves bad taste in the mouth.
Out of curiosity, what are you basing this on?
The text has few of the obvious AI tells. The only thing that, to me, looks characteristic of LLM-generated text is the short and terse sentence structure, but this has been a "prestigious" way to write in English since Hemingway.
Sort of a taste receptor I’m sure many have developed now.
The most obvious patterns here are: antithesis constructions, words choices and distribution, attempt at profundity in every paragraph but instead are runs of text that doing say anything, and even the perfect use of compound hyphenation. I think and can appreciate that there is definitely an attempt at personalization and guidance to make it less LLM-y and not just a default prompt, but it’s still kind of obvious. You could use a detector tool too of course.
Blog posts aren't typically written like Hemingway.
Find some pre 2020 that are, and you'd have a point.
People are not perfect. I went to Ukraine just days before the invasion. Travel and Hotels in Kiev had become extremely cheap. You asked the Ukrainians about the possible invasion. "Not going to happen" everybody said."Russia talks always aggressively, but never does anything".
They did not properly prepare and as a result lost 20% of its territory in days.
Days after that I was back is Austria and could not stop thinking about some of the people I spoke with being dead.
Since that I have also been in Dubai and Saudi Arabia as an entrepreneur and engineer. "What are you going to do when drones are used against your infrastructure?" If you followed the Russian war and first Iranian strike it was obvious that drones were going to be used against them. "not going to happen" again.
The have lost tens of billions for lacking proper preparation. They could have been protected spending just hundreds of millions of dollars over years.
It is about humans, not AI.
> They did not properly prepare and as a result lost 20% of its territory in days.
Ukraine has been preparing since 2014. Without preparation there would be a Russian talking head right now in Kyiv.
I'd say that Ukraine were very prepared for the invasion, though? They managed to survive for the first 2 weeks, leading to a long-term war. The Donbas war had already been going on for 8 years, and I don't think Ukrainians were under some illusion that those weren't Russians.
On the flip side, all around the world you have "leaders" talking about imaginary conflicts with foreign countries that we must spend billions (they have a friend who really should get the contract) to prepare for and if the other side (tm) gets in your whole family will be killed instantly.
Killing of families is what happened in Ukraine in the Russia controlled territories.
In hindsight, it's easy to be smart. You picked two examples where somebody said "never gonna happen" and then it happened. How about the countless examples where somebody said the same and then the thing actually didn't happen?
Take millions playing the lottery. To each of them, I can confidently say "you won't win, not gonna happen". For almost all of them I'll be right. There will be one who wins, were I was wrong, and they will say "see, told you so". That doesn't mean my prediction was wrong. It means you are having a reporting bias.
GP also probably had a sampling bias. The ones who were actually concerned about the impending Russian invasion presumably fled out of the country (or at least, away from the major cities to rural areas that probably see less fighting)
> They did not properly prepare and as a result lost 20% of its territory in days.
They did though. While nobody actually believed Putin would be dumb enough, the Ukrainian army was still, just in case, extremely busy on preparing defences, organising stockpiles, preparing defensive tactics.
> Since that I have also been in Dubai and Saudi Arabia as an entrepreneur and engineer.
Why would we listen to anything related to right or wrong from you then if you don't care?
Excellent post. Two stand-out points are deskilling through abolition of apprenticeship (or equivalent progression through the rank and responsibility), and loss of institutional knowledge, especially tacit knowledge stored in individual people. These are people problems more than they are technology problems. Without continuity of process and practice stuff gets lost. Sometimes change really is progress, for example software safety and security practices have progressed over the past 50 years, but other times change is just churn, or choices driven by misaligned incentives which will bite later, as the article describes.
What comes to mind is how the cure for scurvy was simply… forgotten, causing it to come back.
Every day Peter Naur’s paper programming as theory building gets more relevant
Link: https://gwern.net/doc/cs/algorithm/1985-naur.pdf
The article makes no sense, and stars with a very wrong perspective on things.
This kind of forgetting is normal. It's how things work when time and resources are finite. The only problem here is the belief that you can keep capacity to do something without actively exercising it, and thus the expectation that you can "just" resume doing things after a long break, without paying up a cold-start cost.
But you can't, and there's no reason to be surprised. I bet the Pentagon and the EU weren't. They didn't need those Stingers and shells for decades, didn't expect to need them soon - but they knew they could get them if they really needed them, but it's gonna be costly.
I don't get why people think this is unusual or surprising, or somehow outrageous and proves something about society or "mindsets of elites" - other than positive aspects like adaptability and resilience.
This is true at all scales. Your body and brain optimizes aggressively, too. An individual saying "I need to warm up" or "I need to hit the gym a few times and then I'll be able", or "yes, I can, but I haven't done it for years so I need an hour with a book/documentation..." - all that is exactly the same as EU going "yes we can make artillery shells... though we haven't in a while so we need some time and some millions of EUR to get our supply chain sorted out first".
> This kind of forgetting is normal
Just as shift in power and the rise and fall of nations is normal.
For that matter, a lot of human civilization has been about identifying things that were normal and making them rare. "Normal" infant mortality of 40%, famines, floods, history being lost, etc.
Anyway, when it comes to "this is normal" I think we should take care to distinguish between interpretations of:
1. "This specific case should not have taken certain people by surprise."
2. "This is a manifestation of a broader phenomenon."
3. "This is natural and therefore cannot or should not be solved." [Naturalistic fallacy.]
First of all this is clearly AI-assist writing (being charitable here).
And the premise makes no sense anyway. The only risk of forgetting how to make shells is when other countries are making shells more efficiently. Non-western countries are not going to reject AI-coding, nor are they going to make software more efficiently by hand.
Programmers in non-western countries may not be able to afford $100 per month on vibe coding.
They may keep taking the longer and harder route of a mixture of AI and hand coding.
> The combination of technical skill and the judgment to know when the AI is wrong barely exists in the market anymore.
Well then train them, instead of selecting 0.18% of applicants and calling it a day.
It's not some innate, immutable property - people can be taught even in adulthood.
Also it's not like they'll work for a year and switch jobs - not in the current market.
You could say COBOL has had this "problem" for 40 years also. That's why we need to constantly be inventing new ways of making things. The old ways are always forgotten over time.
If you REALLY need something long-forgotten, then you have lazy-load it back into being at significant cost. That's the price of constant progress.
The point of the article is that sometimes the "old ways" really means "not particularly profitable or necessary in the short term" but the bill comes due in a crisis. The reason US/EU manufacturing was "the old ways" is that people could make easier money with financial engineering, an insight that extended all the way to Raytheon.
COBOL is a bad example, but higher-level languages vs. assembly is not. If you write a lot of C you really don't need to know assembly.... until you stumble across a weird gcc bug and have no clue where to look. If you write a lot of C# you don't really need to know anything about C... until your app is unusably slow because you were fuzzy on the whole stack / heap concept. Likewise with high-level SSGs and design frameworks when you don't know HTML/CSS fundamentals.
As the author says maybe AI is different. But with manufacturing we were absolutely confusing "comfortable development" with "progress." In Ukraine the bill came due, and the EU was not actually able to manufacture weapons on schedule. So people really should have read to the end of "building a C compiler with a team of Claudes":
At least with Opus 4.6, a human cannot give up "the old ways" and embrace agentic development. The bill comes due. https://www.anthropic.com/engineering/building-c-compilerBut these are hard IT things a human programmer really struggles with as well. What % of software written is that? Very very low. Most software is dull and requires business vagueness to be translated into deterministic logic and interfaces; LLMs are pretty great at that as it is. If humans use their old ways to fix complex problems and llms do the rest, we still only need a handful of those humans. For now.
"For now" is sort of the entire point of the article :)
Even in the Before Times, it was much cognitively cheaper to write code than it is to read someone else's code closely, or manage lots of independent code across a team, or to make a serious change to existing code. It's so much easier to just let everyone slap some slop on the pile and check off their user stories. I think it will take years to figure out exactly what the impact of LLMS on software is. But my hunch is that it'll do a lot of damage for incremental benefit.
With the sole exception of "LLMs are good at identifying C footguns," I have yet to see AI solve any real problems I've personally identified with the long-term development and maintenance of software. I only see them making things far worse in exchange for convenience. And I am not even slightly reassured by how often I've seen a GitHub project advertise thousands of test cases, then I read a sample of those test cases and 98% of them are either redundant or useless. Or the studies which suggest software engineers consistently overestimate the productivity benefits of AI, and psychologically are increasingly unable to handle manual programming. Or the chardet maintainer seemingly vibe-benchmarking his vibe-coded 7.0 rewrite when it was in reality a lot slower than the 6.0, and he's still digging through regression bugs. It feels like dozens of alarms are going off.
https://en.wikipedia.org/wiki/The_Mythical_Man-Month
These are good point and I am not overestimating; we are simply seeing the productivity boost in our company and the rise in profitability. We practice TDD, but only at integration level, so we have tests upfront for api and frontend and the AI writes until it works. SOTA models are simply good enough not to do;
function add(a,b) = c // adds two numbers
test: add(1,2)=3
to implement
function add(a,b) return 3
So when you have enough tests (and we do), it will deliver quality. Having AI write the tests is mostly useless. But me writing the code is not necessarily better and certainly not faster for most cases our clients bring us.
This is why a comprehensive computer science degree is necessary. Seeing and working only with the trees leads to destroying some forests, eventually.
Yes. Just like globalization created companies like TSMC, AI will do the same. Software engineers who don't rely on LLM code generators will have a moat because they can do it cheaply and sustainably.
Another reason is that LLMs train on the existing code we already know, don't expect new programming languages or frameworks this means that the software engineering skills that exist today will be relevant for a long time.
I am not so much convinced by your last point, that point of new languages and frameworks. I think the cutoff date is closing in on our current now. If models cannot easily become bigger, they will likely advertise using "up-to-date-ness". Maybe they will be merely a few days behind. Or bigger models will make use of smaller but more up-to-date models.
I think engineering skills will still remain relevant due to taste and proper judgement. A model trained on everything and the kitchen sink has probably not the fitting bias for given specific problems in my project. Accepting too much AI generated code without steering the ship will result in some drift of taste and ultimately make some mediocre project like done by people without good domain knowledge and without good taste. It might even be short term a business, but it lacks the long term excellence, that sets projects with good judgement apart from the common rabble.
> I think the cutoff date is closing in on our current now. If models cannot easily become bigger, they will likely advertise using "up-to-date-ness". Maybe they will be merely a few days behind. Or bigger models will make use of smaller but more up-to-date models
But they will still rely on assembly, C, Rust, Linux, HTML, TCP/IP... Doesn't matter how up to date they are, they rely on existing code they have been trained on, they can't just create new languages without the training data.
When you've run out of ideas just portray "the west" as some monolithic portrait in some decline-porn fan fiction as clickbait.
Reminds me of this post:
https://berthub.eu/articles/posts/how-tech-loses-out/
"the west" ?
You mean the world?
Deepseek was being glazed here, Im sure chinese programmers use it like CC
To be charitable to TFA, there are a dearth of accurate and well-understood labels for the kind of X versus Y they want to make between national economies.
Even "First/Third world" has been fraying at the edges for decades since it was originally about political alignment.
The West started to forget how to code a long time before AI. At first, it was the work visas to bring programmers in, then it was outsourcing. At this point, I'm not even sure if AI is doing more harm than good in this department as it might be able to bring some jobs back to the "West", if it turns out that it's cheaper than outsourcing.
The outsourcing was shedding more of the trivial jobs, while trying to keep key positions at home, but increasingly, it also started to lose the key positions too. It's possible that AI can make it so that the key positions will be harder to justify to outsource... but, who knows... maybe not.
I disagree with the premise - interesting but I interpret the same fact pattern differently.
The history of technology is the replacement of manual processes with automated ones.
Consider a very basic process: checkout of a restaurant.
Writing the price of each item on a sheet of paper, manually adding them and writing the total was replaced with typing in the prices and eventually with just pushing the button for the item. Paper still exists for jotting down your order but within seconds of leaving the table it’s transitioned to computer.
This has enabled lots of desirable advances- speed, accuracy, new payment rails, and increasingly, elimination of the server in checkout- you tap a credit card on a tabletop device.
Did we “forget” how to do checkout? No. We purposely changed it.
But if the internet connection goes down or the backend server powering the cash register app goes down, there is an atrophied and not-regularly exercised skill set (maybe not even trained, IDK) that has to be implemented on-the-fly and it’s slow and frustrating for everyone.
Businesses don’t exercise (or perhaps even train) this process because it’s just not needed enough to warrant the cost.
Military procurement of weapons systems is hardly the place to point to as a technological tradition. There are lots of cases where no one pays the money to keep a production process in place; the reasons are all related to shortsighted “cost savings” or failing to anticipate changing needs.
With coding today, we are seeing the same kind of shift in priorities as my restaurant example. Having humans write code in the 2020 (pre-GPT) tradition was extremely inefficient in terms of time-from-idea-to-implementation.
We’ve found a new way to do the mundane part of that task (the mechanics of translating spec to implementation).
We are figuring out how to do that while preserving quality (and a lot of it is learning how to specify appropriately).
Will we “forget” how to “build” code?
No, but the skills to generate source code by hand will atrophy just as the skills to draw blueprints by hand atrophied with the advent of CAD.
Will we find examples where someone prematurely optimized away knowledge of a skill or process, incorrectly thinking it was no longer needed? Of course.
But the productivity gains we get will be so great on average that no one will go back to doing things the old way.
There will be old-timers and hobbyists who will preserve some of that knowledge; for most it will just be a curiosity.
Everyone is taught at a young age how to do basic addition and multiplication. That's all check out requires. People are not taught at a young age how Rust lifetimes work or how to write human maintainable code.
I agree, as with everything in 2026, the reality lands somewhere in the middle of the discourse online. But pretending this is in practice anything like the check out example is wrong.
The point you seem to be missing is that focusing only on optimization makes us all fragile to system shocks.
> Businesses don’t exercise (or perhaps even train) this process because it’s just not needed enough to warrant the cost.
Until a crisis hits. Covid and supply chain failures. Iran war and straight of Hormuz. Prolonged War in Europe with no production pipeline available. Banks collapsing after unsustainable overleveraging in supposedly "safe" mortgages.
For every optimization and cost-saving measure that is deployed, there should be a backup plan in place. MBA types and "technologists" keep missing this. What is the backup plan for the case where most of the economy activity is built on software produced by business who overleveraged on LLM for code generation?
Though I do believe you are making them in good faith, I find those comparisons do not hold.
CAD still requires you know what to do, and without CAD you can still draw blueprints by hand because you know what the result should be. Checkout is basic arithmetic you can do on a paper or even your personal phone. In both cases it is clear what the process is and what the output should be, and it doesn’t replace knowledge and training and certification.
With coding, none of that is true. By and large, there is a trend of people who don’t know what they’re doing shitting out software, or people who should know better not verifying the very flawed output they get. That is already having negative consequences in people’s lives.
Speak for yourself. I now dare to code much harder problems and learning is bliss. No more having to sit down to dig needle-in-haystack through horrible documentation or random Stack Overflow posts.
LLMs are a magnificent tool if you use them correctly. They enable deep work like nothing before.
The problem is the education system focused on passivity (obeyance), memorization, and standardized testing. And worst of all, aiming for the lowest common denominator. So most people are mentally lazy and go for the easy win, almost cheating. You get school and interview cheating and vivecoders.
But it's not the only way to use LLMs.
Similarly, in Wikipedia you can spend hours reading banal pop-slop content or instead spend that time reading amazing articles about history, literature, arts, and science.
> Speak for yourself.
Even if you are the absolute unicorn who gets paid to "code much harder problems" and "learning", the rest of the industry exists to deliver actual products and services.
So unless you nurture some type of https://xkcd.com/208/ fantasy, this is not just about you. The industry as a whole needs to find a way to work with LLMs without automating programming away entirely, and the industry as a whole needs to find a way to ensure that newcomers are able to be productive even if code-generation tools are taken away from them.
There was a time when companies had terrible development practices and could forget how to build, test, and deploy software, but is anyone seeing that now? We have much better development practices nowadays.
It doesn’t seem much like defense industry problems.
This still happens. Lots of my career has been figuring out what code is actually running in prod, and determining if it even works.
How do you become a senior engineer if no one hires you as a junior anymore.
Talk confidently in your interview with non-technical managers when the last senior has left and there's nobody there to check your work.
So the same as it is now, be a good salesperson.
>The combination of technical skill and the judgment to know when the AI is wrong barely exists in the market anymore.
I see a talent pipeline collapse in next 5 years. "Software engineering is over coding is a solved problem" as being chanted by semi literate media and the AI grifter's marketing departments would further scare away the allocation of human capital to software engineering easily commanding 3x rise in salaries due to resource shortage.
First it was “learn to code” and bazillion videos of TikTok schmucks showing off slacking at work, now everything is solved. The puzzle is complete.
Do you think this is a tooling problem or more about incentives and how engineers are trained now?
I think the article is making the point that it is a cultural problem about cost cutting and short term thinking.
Yeah, agreed - short-term incentives seem to drive a lot of this. Do you think tools can help, or is it mostly cultural?
Is this written by a real person though?
> Optimized for minimum cost with zero margin for surge. On paper, efficient. In practice, one bad day away from collapse.
I'm going to steal that one and add it to Stross': "Efficiency is the reciprocal of resilience."
Yes that is one key that resonated with me. The author did a great job of putting these recurring concepts into their own words
The other that really resonated was something that I read before along the lines of… we think that once humanity learns something, that knowledge stays and we build on it. But it’s not true, knowledge is lost all the time. We need to actively work to keep knowledge alive
That’s why libraries and the internet archive are so important. Wikipedia, too
> Leadership qualities. Our last hiring round tells you how rare that is: 2,253 candidates, 2,069 disqualified, 4 hired. A 0.18% conversion rate.
It's minor but this is just wrong. If you're going to hire 4 candidates, there could be 2,253 perfectly qualified candidates even if only 0.18% get hired. The conversion rate is meaningless; it just tells us how many jobs were on offer. There is no way that the skills this fellow wanted were so rare and difficult that only 1/500 candidates could possibly handle the job. Humans even in the 1/20 mark are pretty competent if you're willing to train them and legitimate geniuses crop up at around 1/200.
He writes 2,253 candidates and 2,069 were disqualified. 184 were qualified, so 1 in 12 was considered competent.
This is some convoluted BS built on the premise that wars need to make sense, economically or otherwise. No, wars do not need to make sense. If a person, a dictator or a president, unilaterally starts a war that forfeits the lives of both the dictator's (possibly fabricated) enemies and its own people, that person is knowingly committing murder. Logically, such a person should be handled with at least as much prejudice as a lone wolf that opens fire on a crowd. So we need to fix our legal systems to be better at preventing wars, not our economic systems to be better at fighting them.
This will end with the way of COBOL with a few people that still have the expert-level understanding of refactoring old code without causing outages or service disruption.
We’ll see, but right now I now see developers 24/7 hooked onto their agents and in the future we will experience a de-skilling problem which clean code, best practices, security and avoiding NIH syndrome will be all flushed down the toilet.
Hope we dont forget humanity one day!
While the Fogbank story is a funny anecdote, I don't see it as a fitting example for atrophied skills. It's like writing a clean implementation of some software and it just doesn't match the legacy version until you realize that the legacy version had an unnoticed bug that made it behave the way it does.
Isn't that is the point of technological civilization development? People for example forgot how to weave on the handloom, or all the parts production and the maintenance for the watermills. And wooden sailships - top mastery of handling and engineering developed for millennia, gone.
As it was said - the future is here, it just distributed non-uniformly, so somebody is still and will be for some time sailing, manufacturing things and writing code.
We have both forgotten how to make things and also decided we can make more profit letting someone else make everything for every market. We have moved to a generation fixated on maximizing profit. However there is logic there as the cost to access the ability to make things is prohibitively expensive. As someone who makes open hardware with a nod to the environment and reusability, you can not justify or even find more locally sourced options than China.
Coding is different though, coding doesn't have a cost barrier, it has a ability barrier. I think we will loose a lot of people who never were passionate about programming and perhaps go back to a happy equilibrium. AI is only production ready if you have someone who understands software development. AI will improve speed to market if you have the right team, it doesn't remove the need for some to learn to code. You will of course end up with startups using exclusively AI but they will be those who end up with major security breaches or simply cannot scale as the AI goes in the wrong direction for the future. Tbh that's probably a positive as it weeds out the start ups that are focused on buzzwords for funding and not product.
No matter what happens to the viability of software development as a career, I will always care about the craft as I have done the past twenty years and change. The imperatives to adopt LLMs in situations where they do not benefit me nor my work is what is driving me away. I have to agree with latexr; the people who seem to benefit the most from the current moment are those who see software as a means to an end without much concern for quality, longevity, nor customer experience.
Why is speed-to-market such an important metric? I do not understand the need to mimic the largest players in the industry, nor do I see any particularly profound long term benefits to first mover advantage.
> I think we will loose a lot of people who never were passionate about programming
Anecdotally, what I’m seeing right now is the opposite. People who don’t care about programming are joining, while those who do care are getting tired of the bullshit and leaving. The good programmers are the ones leaving, the hacks are extremely happy to use LLMs.
When shit hits the fan, there won’t be many people left to clean it.
So you see people who don't care about programming, joining and getting comfortable with vscode and claude code and devops?
Because it seems to me like there's a lot of coding-adjacent things they still need to be able to do even if they never look at a line of code.
Those examples are nonsensical. None of those are necessary to get working code. The VSCode example is particularly baffling. Firstly, I’m sure you understand there are other editors people use for code; secondly, I know even people who don’t code who have picked up VSCode for text editing and are fine with it.
> The defense industry thought peace would last forever, too.
Not really since they are always pushing for more wars.
exactly, as they say everyone has to learn to code.
I can't not write the tired comment of how ridiculous it is to criticize AI and then use AI to write your article. It's tired, but so is this writing style.
For the actual problem, I fear this can't be solved by warning people, the pain will need to be felt. The system we live in, basically free market capitalism, cannot do anything else except local optimization. Maybe it's for the best, I don't know. The alternative of top down planning wouldn't have this problem, but it would have other problems. I work for a mid size somewhat luxury brand, and the major goal right now is cost cutting and AI for efficiency everywhere instead of using it to create better products or better ways to reach out customers. When I think about who will buy our luxury products if all jobs were optimized out of existence, I don't have an answer, but again I think the pain will need to be felt to change course.
> After spending an additional $69 million and years of reverse engineering, they finally produced viable Fogbank. Then discovered the new batch was too pure. The original had contained an unintentional impurity that was critical to its function.
Same thing that happened to the unfortunate Dr. Jekyll!
when you offshore or automate away the hands-on knowledge, you don't just lose the workers, you lose the entire institutional memory, and no amount of money can buy that back overnight
> The West Forgot How to Make Things. Now It's Forgetting How to Code
Can we stop repeating this nonsense headline please? We did not stop manufacturing things.
Manufacturing is a huge industry in the West. https://en.wikipedia.org/wiki/Manufacturing_in_the_United_St...
The US manufacturing sector is the biggest it has ever been. Exports are at all time record highs. The only thing that declined about manufacturing is the jobs. We build way more than we ever did but with far fewer people.
What we did do is decide that basic items aren't worth it. Our capacity is limited, our labor pool is limited, expenses are high, it doesn't make sense to make trinkets when we can make complex high precision parts and devices.
But no, we did not forget how to make things. We chose to use our capacity in a smarter way.
>Denis Stetskov
？
Putin's propagandist, or just useful idiot.
> I run engineering teams in Ukraine. My people lived the other side of this equation. Not the factory floor. The receiving end.
With all due respect, but many european taxpayers help pay for Ukraine. I am not disagreeing on the premise of the West killing itself via systematic recessions - Trump invading Iran leading to inflation as an example - so a lot of things are going on that show a ton of incompetency both in the USA and the EU, but at the same time I also get question marks in my eyes when this criticism comes from a country that receives money from others. That money could instead go to make EU countries more competitive, for instance. I am not saying this should necessarily be the case, mind you; I fully understand the nature of Putin's imperialism. But we need to really consider all factors when it comes to strategic mistakes with regards to production - and that includes taking up debts all the time. There are always a few who benefit in war, just as they benefit from subsidies from taxpayers (inside and outside as well).
Ukraine is "receiving money from others"? We are benefactors of the Ukrainians' bravery and sacrifices. How much money could we have not spent if Hitler had been stopped in Czechoslovakia?
> Ukraine is "receiving money from others"?
Yes. https://www.eeas.europa.eu/delegations/united-states-america...
You are completely ignoring the argument of your parent comment. They are saying that money is being spent to the benefit and best interest of the spenders, that it’s not a handout.
You are, of course, free to disagree and make your point, but ignoring the argument does not advance the discussion.
> Ukraine is "receiving money from others"?
Factually correct.
> We are benefactors of the Ukrainians' bravery and sacrifices.
Who's we?
> How much money could we have not spent if Hitler had been stopped in Czechoslovakia?
Very different situation, in all aspects.
You see zero similarities between Hitler invading Poland and Putin invading Ukraine?
There are some pretty substantial differences. Russia is on the strategic back foot here trying to figure out a way to stop NATO's advance. They've only turned to violence after long attempts at resolving the tension diplomatically and the US has been implacable. Putin's actually been pretty hesitant in his escalations so far; he's 70 and has a long history of trying to avoid war.
Hitler was more about wanting more land and resources for Germany, and he saw war as being a legitimate tool for achieving his aims that he deployed early and enthusiastically.
NATO has advanced into which part of Russia?
Not Russia, Eastern Europe - https://en.wikipedia.org/wiki/Enlargement_of_NATO
So, NATO hasn't advanced into Russia then?
Just Russia advancing into the Ukraine (after promising not to if the USSR nukes were given to Russia)?
Gotcha.
> There are some pretty substantial differences. Russia is on the strategic back foot here trying to figure out a way to stop NATO's advance.
His rationale for invading Ukraine was to "demilitarise and denazify" it. The NATO point seems largely be invented by people who dislike NATO in the west.
> They've only turned to violence after long attempts at resolving the tension diplomatically and the US has been implacable.
I hope the "tension" you are referring to was not the little green men taking over Crimea and the Donbas in 2014.
> Putin's actually been pretty hesitant in his escalations so far; he's 70 and has a long history of trying to avoid war.
This is a totally unseriousness statement. Can you remind me what Putin was doing in Syria again?
There's an english transcript [0] of his speech from when they went in up on the Kremin website. He opened with something like
> I will begin with what I said in my address on February 21, 2022. I spoke about our biggest concerns and worries, and about the fundamental threats which irresponsible Western politicians created for Russia consistently, rudely and unceremoniously from year to year. I am referring to the eastward expansion of NATO, which is moving its military infrastructure ever closer to the Russian border.
They're claiming the NATO thing is relevant. Opening paragraph justification.
[0] http://en.kremlin.ru/events/president/transcripts/67843
> Russia is on the strategic back foot here trying to figure out a way to stop NATO's advance. They've only turned to violence after long attempts at resolving the tension diplomatically and the US has been implacable. Putin's actually been pretty hesitant in his escalations so far; he's 70 and has a long history of trying to avoid war.
Is that why Russians rejected negotiations when Ukraine offered to never join NATO and Russians insist on keeping invaded territories?
Rather bad premise in the article. 1.) Germany, Italy and Eastern Europe are very industrial regions. The author forgets defence is not only the industry. 2.) The author doesn't show any source that Chinese developers don't use AI
I don’t know, but the evidence shows that software engineering is not that deep of an art.
People come and go at rates that would not be sustainable in any manufacturing business.
Yes, businesses tend to believe that.
No, every time people switch knowledge gets lost and code quality degrades.
In part I blame accounting rules justifying investments is easier than maintenance.
As an anecdotal evidence I code way more now with agents because i have an entity who has vast amount of knowledge about pretty much everything and I have the creativity to use that well.
But you already knew how to code before LLM coding agents, juniors will jump straight into using agents without learning to code by hand, hence the premise of the article.