HN Reader

How does one explain the drop starting January 2023 (esp for things like Customer Service Rep, which is an NLP-heavy task) when most corporations didnt even start LLM/NLP pilots until mid/late 2023? I skimmed thru the 100+ page paper but didnt see an explanation for this strange leading effect.

SWE figures dropped mid-2022 (almost magically in line with interest rate hikes) and LLM-copilots werent introduced for another year. The paper notes they did an adjustment for the end of ZIRP. I dont know enough econometrics to understand whether this adjustment was sufficient, but the chart doesnt make sense since the labor efforts seem to be leading the actual technology by over a year or more. From informal surveys, LLM-copilot usage didnt become widespread until late 2023 to mid 2024, certainly not widespread enough to cause macro labor effects in mid-2022.

I think not hiring juniors is a tragedy of the commons situation. It started before the AI boom, during COVID. It's not tax-related as people claim here, since this phenomenon is not US-only.

The ZIRP era made companies hire people as if there was no tomorrow, and companies started "poaching" engineers from others, including juniors. I saw some interns with 2 years of experience getting offers as seniors. I had friends being paid to attend boot camp.

Then everyone realized they were training junior engineers who would quickly get offers from other companies as “Senior" and leave. So companies stopped hiring them.

I made a stupid simple model where hiring in all age brackets rose slowly until 2021 and then fell slowly. That produces very similar looking graphs, because the many engineers that were hired at the peak move up the demographic curve over time. Normalizing the graph to 2022 levels, as the paper seems to do, hides the fact that the actual hiring ratios didn't change at all.

https://docs.google.com/spreadsheets/d/1z0l0rNebCTVWLk77_7HA...

We need to reframe it. At this point, what we call "AI" is not a technology, but a subscription company.

A technology is a tool you can adopt in your toolchain to perform at task, even if in this case it's outsourcing cognitive load. For a subscription company, well, as long as the subscription is active, you get to outsource some of the cognitive load. When Anthropic's CEO says that white color jobs will disappear, he means that he is selling Enterprise subscriptions, and that companies will inevitably buy it.

Lots of apparently unexplored alternative explanations. In times of uncertainty, you don't hire unless you need to. Another junior developer or customer service agent can be delayed, and youngsters in the labor force are most exposed to this. But if you need a home health aide, you probably don't have a lot of choice: somebody has to change grandma's diapers. Tariffs top my list of uncertainties for businesses, but interest rates are a close second.

Any economic data from between 2020 and 2025 should be tossed in the garbage. We will have no idea what affect AI has or hasn't had until AI has been available outside of the extremely confounded current circumstances. Tell me how employment looks after the next recession when the after effects of the pandemic, rapid inflation, interest rate unpredictability, and tariff whiplash are hopefully all behind us.

Trends take time to work out.

"AI" dives in and disrupts and then it turns out that AI isn't too I. The disrupt phase where HR dumps staff based on dubious promises and directions from above takes a few months. The gradual re-hiring takes way longer than the dumping phase and will not trigger thresholds.

I've spent quite a while with "AI". LLMs do have a use but dumping staff is not one of the best ideas I've seen. I get that a management team are looking for trimmings but AI isn't the I they are looking for.

In my opinion (MD of a small IT focused company) LLMs are a better slide rule. I have several slide rules and calculators and obviously a shit load of computers. Mind you my slide rules can't access the internet, on the other hand my slide rules always work, without internets or power.

Software engineering is in correction mode since 2022, right after the Covid highs. AI is just the facade for job cuts. Zuck has been doing “the year of efficiency” for years now.

I started university (in Australia) in 2004, not long after the dot com crash. CS enrolment rates were low, kids were getting scared off due to perceived lack of jobs. As a result, there was a shortage of grad talent (companies were already ramping up hiring again by 2004). I got a grad job just fine, in 2008, and I've never been short of work since.

So my advice to high school kids of 2025: right now is the perfect time to enrol in CS. 5 years from now, the AI hype will be over, and employers will be short on grads.

There are other factors that drive employment of young workforce, which don't have much to do with knowledge or skills. These are sort of blue-collar jobs within IT.

There is a lot of low-level drudgery work which is currently outsourced or assigned to non-employees (mostly young) in most companies. For example, the IT support and maintenance is mostly done outside of the West. This works requires a lot of back and forth. Don't think AI taking over this area.

Also, some work is assigned to young workers to spread the accountability and risk ownership. Not sure if you can hold AI as accountable as humans.

Also, young workers are generally preferred for their agility, flexibility and ability to work hard. There were easier to exploit and if they were unmarried, they won't mind giving all of their time and attention to work for less pay. I used to work in teams that stayed overnight at office to complete projects. Young people are also very cohesive and they team up well.

So, looking at the situation purely from a knowledge perspective may not give the full picture.

Discussed earlier in https://news.ycombinator.com/item?id=45052423

Several teams around me stopped hiring juniors over the past couple of years. It’s not that the newcomers aren’t good, it’s just that no one has time to train them. AI showed up at just the right moment to offer a convenient excuse, and companies are happy to save on the cost of mentoring. But long term, this feels like borrowing from the future. Without someone to train, there’s no one ready to step up later.

I don't get how interest rates is given at best a cursory phrase when ZIRP regime ending is one of the biggest macro events of the past several decades. Seems like it would deserve more of a spotlight.

Skimming this, I'm not sure why it couldn't be explained by the layoffs we had a couple years ago, which were primarily at tech companies (which are indeed more exposed to LLMs) and probably hit junior devs more.

I doubt the whole narrative is true, it may just be hope that we do not need to hire because we will be able to do it with AI. But reality outside of software and tech is very different. I work in an organisation that heavily pushing AI and even with that push, a typical employee is still not utilising it fully, unprepared and expecting training from the employer. There is also a disconnect over connecting org data to these tools and between developers and cyber teams.

Looking at the "overall" graph they provide, it's clear that jobs as a whole have flatlined since 22, and even non-exposed jobs have had a growth rate hugely reduced. If we disregard the existence of AI, it would be easy to assume this is just the regular economic slowdown impacting younger people more than older ones. I don't see any correlation work done to isolate the source, especially as a lot of the "ai-exposed" work is also very much exposed to how much money is flowing in.

My experience: we're hiring for an AI engineer position and for a frontend developer position (and yes, we posted our positions here on Hackernews).

We have a stream of cookie-cutter candidates. As if they are clones of each other, it's uncanny. They typically have a BS degree in some foreign university, then a CS Masters' in the US, experience with robotics, then several years of experience in large companies.

And they completely fold during in-person coding tasks. Like, not being able to explain the difference between DFS and BFS (depth/breadth-first search). Or being able to write a simple custom metric and train a network in Pytorch.

And a similar story for the frontend developer position.

We now literally have to add more filters to not get inundated by underqualified candidates. These filters will make it harder for beginners to even _get_ to the resume review stage.

No conclusions from me, but something's been broken in the CS jobs market for a while.

I'd be interested to see more discussion on Section 174, how engineer salaries suddenly triggered a higher tax bill in 2022. https://qz.com/tech-layoffs-tax-code-trump-section-174-micro...

Did Section 174 changes accelerate tech job losses? It's unclear from the paper because they only look at the data excluding tech firms, not only tech firms.

I want to see cause-and-effect: the cause being tax changes, the effect being tech firm layoffs, but they don't analyze the effect.

Short term thinking. Yes, maybe those jobs are replaceable. But how do you train the next gen of seniors?

Meh... just rehashing what he said before. The paper itself is fundamentally flawed, examining only a minuscule portion of the job market. If we step back and look at Europe's struggling economies over recent decades, we see that economic downturns disproportionately affect young people. Greece serves as the poster child, followed by Spain and Italy. In Germany alone, we've lost 50,000 jobs in manual labor heavy industries (mainly automotive) this past year. We're also seeing a 60% decline in apprenticeships for labor intensive roles at DAX companies that aren't even AI affected yet. AI has become a convenient scapegoat for a faltering economy driven by geopolitical tensions, protectionism and unqualified leadership in the world's largest economies. Roaring 20s indeed.

Young people from where? I’m in South American and the young people here are earning much more than their parents earn but 1/3 what software engineers earn in San Francisco writing software.

For a startup, I now can solo it with the help of AI far beyond what was possible a few years ago.

This does not mean I otherwise would have hired a few juniors. More likely I would not have taken that business beyond the idea stage.

But direct massacres I can personally observe are in translation, copywriting and illustration. Yes, a good person can do better, but for a majority of cases AI has already become "good enough".

Then there is an onslought from just non-AI automation, in e.g. retail and finsncial services. This had been building for over a decade, but the pandemic lockdows exponentially accelerated it.

Part of the job downturn may be due to AI, not in the obvious way of LLMs doing peoples jobs but due to the money allocation:

>Companies will spend $375 billion globally in 2025 on A.I. infrastructure, the investment bank UBS estimates. (nyt)

which means they have less money to hire and train recruits.

By what mechanism can you distinguish a decline from one cause compared to another?

I could understand if, given enough time for all of the factors involved to occilate that you could pick out a signal, but that's not the data that exists right now. Surely the only way to identify the first time instance of a cause of decline could only by accumulating a count of clear instances where the cause occurred and measuring those as a proportion of other cases.

I'd wager that vast resources being spent in one very niche subfield is what's hurting job creation from young people.

Aside from this there simply doesn't exist another mechanism in which AI can meaningfully affect employment in the long term.

It's not AI replacing jobs. There is a hiring downturn due to economic conditions (reaction to covid over-hiring and subsequent interest increases) and the people on the bottom of the ladder get bumped off.

Researchers need to project the pre-Covid hiring rates and omit the subsequent overhiring boom to credibly report on any believable loss in jobs. Dropping personnel who should never have been hired to begin with is a return to normal, not a new concerning trend. Start by showing the data from at least 2015 onward rather than lying with statistics using charts starting in 2021 for maximum drama.

It's striking that the decline in hiring started when ChatGPT was first demoed, but wasn't really good enough to do much useful.

If it wasn't working now, though, we would see those graphs going back up.

A lot of CEOs are saying do more with less. Basically they are saying, “If there are any requests to hire additional people the first thing I ask them is if any AI tools have been tried first.”

Then every 2-bit president, leader, manager, CEO out there regurgitates the same thing.

So yes, companies want to save money and do more with leas. It certainly won’t help job seekers or the economy.

Interest rates have way more impact.

Tax code changes & interest rate changes have been unfavorable. Secondly “learn to code” and vastly lowered CS standards have churned out a lot of young developers that are unskilled, much like in the .com bust. Lastly AI which at best has completely shut the unskilled out.

Seems to me like the jobs disappeared in reaction to the extremely obvious: a certain president causing a crisis to get himself a tax cut, the latest in a very long series of government decisions to make labor extremely unattractive for the private sector. AI is an excuse, nothing more. And I mean both Macron and Trump.

AI is not even remotely close to the point where it can replace a human working for you, that's just not it. Everybody keeps screaming "NOT YET", and perhaps that's true, but either case, it's not currently true.

However, economic conditions are so that companies now want to not hire, because price rises due to tariffs mean much less spending happening, with the outlook steeply downward. It will very mean widely spread layoffs soon. The first to suffer are new entrants to the job market and the close-to-pension near-the-exit people as well, but let's not kid ourselves: they will hardly be the only ones.

The irony is that to me, this seems to be the reverse of what people think they see: AI is a clear investment target at the moment. It's being accused of killing jobs, but reality? It's one of the few clear investment targets. It is a portion of the economy that is entirely opposing the general trend. AI is not just not causing of firings, but actually is putting a break on the firings (by providing jobs funded by investment). It's a huge positive for the economy at the moment, and I bet it's a huge positive for jobs at the moment. Like most investments, the gains it promises have barely materialized right now: it's promising to replace people but it isn't. Frankly, the only other large investment target I see is weapons. And, while AI can be used in weapons, it's still leagues better than making actual weapons in my book. Nobody is literally blowing off anyone's head by firing an AI model at them.

I find this an obvious insight once you look at what companies are actually doing: they're not killing jobs and providing the same services. They're killing jobs and providing less services, for less money, to try to survive. If AI was replacing people, they'd be providing more service for less money. Money is the problem here. Money is artificial, entirely a decision of society. Where is the money going? Look at the stock market: it certainly isn't going to investors (look at an EU stock market, outside of weapons manufacturers)

And there's the key, fundamentally, businesses' desire to replace people is not a desire of business. It's a government decision. A tax decision. A very large portion of government income is income tax of various kind. People-based taxes. This means: you generate revenue from the means of production, capital (ie. "a factory"), input goods, and people's labor? It's people's labor that is taxed. Everything else is actually tax-exempt. Capital is free (you can spend 100% of your investor's money on buildings, no tax, assuming you follow standard practice). Input goods are free (VAT/Sales tax exempt, you don't even have to pay it initially). You want to keep more of your company's revenue? I don't think enough people realize this: but you lose a LOT more on the people than on anything else. In fact you only lose on people in some sense (because capital and input goods can at least theoretically be sold or more likely subleased for similar value as you got them for, they're "free", or only historically low interest is required). You almost only pay the government for the privilege of getting labor from someone else.

In France, the actual tax is over 70%, if you measure it fully. By that I mean: let's say I pay you $100. Any tax, patronal (employer tax), income tax, insurance, pension (which are all government), import tax, VAT, ... is all to be paid from the $100. You like the product my company makes. You buy all you can from it, you spend all of that $100 on my product. How much do I see on my bank account? $28.7, by my last count. That amount, by the way, is then taxed at 40% before it goes into my bank account, so it's comfortably under $20. The $28.7 is what I could reinvest into my company (and need to buy the input goods in the first place). Even this is not counting many forms of tax that are effectively mandatory for everyone, such as rental tax, garbage tax, park tax ...

In other words, if I didn't have to pay any tax at all, I wouldn't bat an eye to have your work done by 3 or even 4 people, as that would be a much better deal. I'd be looking for extra services people would pay for and provide them.

Now I get that government is necessary, but we're paying for government services in the west by artificially increasing the cost of labor 500%. That is where tax comes from. There's other taxes, but they're perhaps not rounding errors, but obviously it's not what I'm worried about. And you are complaining that CEO's, or AI, or greed, or ... is costing jobs?

No. The fact that in at least one view something like 80% of all efforts (all "economic value", mostly labor) in France is just keeping the government running, nothing more, that is making it critical for the private sector to avoid using labor, to avoid providing jobs like the plague. Oh, and of course, then the government allows foreign (read: US) companies to come in and "use contractors", in other words: to not pay labor tax (Uber, Deliveroo, ...), which they of course don't allow for French companies.

And then, because the government is never happy with a big disaster unless it grows into a nice huge well-fed catastrophe, government makes it really hard to stop spending on people, so over hiring is a huge, extremely expensive, mistake. 100-person companies go bankrupt due to 5 unnecessary hires at the wrong time. I get that this is people's lives we're talking about, but given that the government gets 80%+ of the value of people's labor, I feel like it's perfectly reasonable to ask the government to take care of these people. Of course, they don't, not well.

Of course, it's "companies' greed" that gets blamed for all the effects. Truth is this just isn't true, frankly not even at huge companies like Total. Or, to put it differently, if party A's greed is not even 20% of the total, and B demands over 80%, I feel like a healthy amount of blame needs to go to party B.

what I hate about it the most is, all AI company CEO constantly deny the fact that at least 50% of the jobs will be gone

> For automative occupations, a lot of it is software engineering, auditing, and accounting, where there are well-defined workflows and LLMs are good at doing one-off tasks without a lot of feedback. For augmentative cases, you’re looking at more complex or managerial roles.

Nope coding is definitely augmentive for Sep 2025 and before. Maybe more so than managing.

Yeah we the people at silicon valley really care about young people job prospects of course, yeah, who would ever even doubt that.

Maybe its the secret recession?

If we don't solve this by 2030, This and including the projected 10% increase in unemployment in the same year will give us another crash.

That is what true "AGI" is.

What are those "augmentative" jobs that fresh CS grads can transition to?

AI taking jobs is the biggest bullshit and hype the world has ever seen that been pushed by the likes of Salesforce etc to tell their customers: Look out AI product is so good that we used it ourselves to cut our own workforce. And inside the company walls they whisper to each other : our AI product is such crap that it is really useless and the only way to sell it is to credit our AI for tha layoff and hiring freeze we are doing anyway.

Demand will shift to AI-literate knowledge workers whose productivity will be increased.

I don't think it's AI primarily, but rather AI is being used as a justification and a blame, and what's a better blame than this new trend that will last years and nothing can be done about it? I think it's just companies' greed, another excuse to lower the wages more and/or use ones who are ok with lower wages. It also has to do with procurement that's already using AI to process applications, with all the negative impacts this will have. I also did notice something when it comes to the job market: mostly those who are suffering are men, not women. I initially thought maybe because of the nature of the jobs, but that wasn't the case. Looking around on LinkedIn, a lot of women are doing product/project management jobs, analytics, business-related roles, even HR and procurement aren't impacted, and I have yet to see a single woman personally or online that was let go because of AI, even jobs in legal work, administration work -none of them are impacted that much. Meanwhile, I personally know several engineers (not software engineers) who are jobless with years of experience, and a LOT of software programmers too! I was discussing this with a friend and he also brought up a new term, "boomers' harems," as in, managers who are mostly old men prefer to hire young women now, and he did support his claim with many examples of fresh graduate girls that are working as engineers in aerospace or similar industries already, and their male classmates are still looking around.

So in conclusion, while AI does have an impact, personally if I would quantify it, it probably won't exceed 25% of the issue. Other factors make up the rest, but everyone is distracted by AI for all sorts of reasons.

AI has destroyed salaries and it is destroying jobs. The problem AI actually solves is wages. Why? Because perfection is the enemy of the good. Anyone who still believes AI is not destroying jobs is delusional.

Does the ADP data include international employees in lower cost geos?

This article continues to propagate the misconception that young is the same thing as inexperienced.

Yes no shit.

As individuals less work is a good thing obviously. But as a society less work is a bad thing.

Perverse.

In Canada, mass immigration is creating high unemployment for young people. Don't be so quick to blame AI when government policies are to blame. https://www.kingstrust.ca/wp-content/uploads/2024/11/TKTC-De...

Programming in a structured high-level language like C is much more productive than programming in assembly. Programming in a managed language with GC can be more productive still.

The creation of package managers and the widespread availability of open-source repos means developers don't need to write as much from scratch.

The creation of search engines and Stack Overflow did (and still do) much of the useful things that people use AI for (boilerplate, debugging obscure error messages).

Machines have gotten exponentially faster for the last several decades. This means devs need to spend less time optimizing code. And the time to compile and run speeds up, meaning you can prototype things faster.

Why is it, that somehow none of these inarguable improvements to the speed and efficiency of development haven't lead to a a massive decrease in the number of developers? If we take it as axiomatic that AI significantly improves productivity, why, for the first time in history, does that not result in more programming jobs?

There's a sizable overlap between the anti-AI movement (e.g. https://www.reddit.com/r/antiai/) and self-identified socialists.

One thing that confuses me is that the anti-AI movement has adopted several talking points saying that nobody wants AI. (e.g. nobody wants it, the AI companies are pushing it on us, management is pushing it on us, it just creates low quality slop, the demand is fake, etc).

But if there's no demand, then there's no threat to jobs. On the other hand if there is a threat to jobs then there must be demand (since it competes with human labor in the job market). I'm not sure why they're taking this particular tactic, but it's going to lead to strange comment sections where people won't know how to stay on message as it becomes clearer that AI is impacting the labor force.

A lot of people intuitively imagine that to compete with humans there's some sort of 1:1 exchange ratio. But initially the calculus will be something like a 10 team with plus AI performs about on par with a 12 person team (or whatever the details work out to). So we will see AI impacting job numbers well before AI can do your entire job.

I would expect something more convincing for such a (seemingly) popular article of this title. Seems like a relatively straightforward difference-in-differences or synthetic control could be applied to show something convinving, but this is just some dude's opinions as 'evidence'.

Well, that’s the makings of revolution.

Don't underestimate the effect of Elon nuking the Twitter employee base when he acquired the company. All the predictions were that Twitter would die. It hasn't. That sent a powerful message to the other leaders of tech companies: "You can do just as much with far fewer people" / "Most of your employees aren't doing needle moving work". Shortly thereafter, Zuck announced layoffs at Meta. These new attitudes towards headcount have cascaded down from the top-tier tech companies down through the ecosystem of companies.

Even if we take the empirical claims from highly partisan sources at face value, obvious axiomatic problems remain. AI employment doom, like other forms of neoluddism is predicated on a zero-sum view of economics.

As an example, consider the claim, "Steam shovel adoption is destroying entry level jobs in ditch digging"

Yes, we want more productive technologies to be used where applicable. We want workers to specialize and become more productive with less hours worked. This isn't destructive at all. Framing it as such is either deliberately misleading or demonstrates a misunderstanding of basic econ.

[flagged]

My challenge here is that every time I see “The Atlantic” or “The New York Times”, I can’t shake the feeling that it’s got to be a paid advertisement or some influence piece by a special interest group. I am not sure what to make of the articles that appear in those places anymore.