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.
advael · 1h ago
The 2022 drop for SWE is easy for me to explain, and it's not on these analysts' list of factors (though I'm not an economic quant, I don't know how you could really control for it): In 2017, a tax bill was passed that cut a particular tax incentive in 2022 in an effort to be counted as "revenue neutral" despite being otherwise a massive tax cut overall. The incentive in question was a writeoff for "Research and development". This means that in 2022, it got effectively much more expensive to hire anyone who falls under that category, including developers not directly necessary for the day-to-day function of a business (hell, one might argue they would have counted anyway) and scientists of most kinds. That this hit big firms, which have a higher relative amount of R&D efforts going at a given time, first makes a lot of sense.
For customer service, my explanation is that companies literally do not care about customer service. Automated phone trees, outsourced call centers whose reps have no real power to help a customer, and poorly-made websites have been frustrating people for decades, but businesses never seem to try to compete on doing better at it. It's a cheap win with investors who want to hear about AI initiatives to lay off yet even more of this department, because it doesn't matter if the quality of service declines, there are no market or regulatory forces that are punishing this well enough to ever expect firms to stop breaking it, let alone fix it
TuringNYC · 52m ago
Love this note. For those interested, this is the Tax Cuts and Jobs Act (TCJA) of 2017 Section 179.
For a software engineering business, the Tax Cuts and Jobs Act (TCJA) of 2017 significantly impacted how software costs can be expensed under Section 179. While Section 179 previously allowed for the immediate expensing of many software purchases, TCJA reforms restricted this deduction primarily to "off-the-shelf" software. Custom-developed software and internal development costs are no longer eligible for Section 179 expensing and must now be capitalized and amortized.
Under the TCJA, Section 179 cannot be used for software that a company develops for itself. This includes the direct costs for the engineers, programmers, and other personnel involved in the development process.
The report not addressing this elephant in the room is a disappointing.
kevindamm · 42m ago
I think it may have been a one-two punch of §174 and the end of ZIRP.
Of note, the OBBB reinstated the ability to deduct R&D, so businesses are no longer required to capitalize and amortize R&D expenses (including software development).
Section 174 was the big one that affected how SE salaries could be deducted, that many blamed for the start of layoffs.
However, that's back for tax year 2025, so why aren't we seeing the jobs come back? Maybe it really was 174 then, but AI now?
MontyCarloHall · 30m ago
>why aren't we seeing the jobs come back?
It was only just reinstated, so it's probably too early to see the effects.
I also expect that despite the restoration of Section 174, companies realized that they not only overhired during ZIRP, but also that they don't actually need that headcount, given the outcome of Musk's Twitter layoffs. There were so many prognostications that Twitter would imminently implode after downsizing from ~8k to ~1.5k employees, and when these claims never came to pass, it was a wake-up call to the rest of the industry [0].
Even in that thread, a lot of people were saying "it's only been three months, give it a bit more time."
yieldcrv · 5m ago
Recruiters have been blowing my inbox up since the day Trump signed the OBBB
although I think entry level is still in shambles, for now
tru3_power · 23m ago
What was the reasoning behind this change? Isn’t R&D something we’d encourage to be a tax write off since it reduces the cost of well.. R&D? Or is R&D not as important as I’m thinking?
tart-lemonade · 8m ago
It was done to try and make the bill somewhat pencil out and make the national debt increase less egregious. Everyone just assumed it would be delayed forever or reversed before it could take effect, but those negotiations failed, triggering massive waves of layoffs.
The story that I’ve heard (probably in here) is that the administration did it to make the tax bill look more balanced over the long term by phasing out that tax write-off, while giving them (or the next administration) the time to reverse it before it really impacted anything. But, nobody reversed it, until this year.
rcpt · 21m ago
That administration really didn't like tech
le-mark · 39m ago
IMO there exists a concerted effort to push this “ai takin our jerbs”. I just wonder by who and why?
advael · 35m ago
People heavily invested in AI, obviously. We have an economy that's heavily driven by investor perception, and while "AI is taking jobs" sounds bad for workers or humanity as a whole, it can still be a sales pitch for investing in AI
ivewonyoung · 1m ago
I've seen three major pushers of this narrative. Well 3.5.
1) AI CEOs, like the Anthropic CEO quoted in the article, Sam Altman(till recently atleast), Perplexity CEO, Microsoft CEO etc. It brings them new VC money, investment, and customers who "don't want to miss out on this big trend".
1.5) Media heavily pushing the above CEOs quotes, probably just for clicks and engagement, brings them money
2) BlueSky appears to really hate on anything AI to the point of personal attacks, putting AI tweeters on blocklists and bans etc.
3) Reddit, especially the antiwork side
samtho · 5m ago
From the perspective of the publications, it’s fear-mongering which makes clicks.
Out of control AI is a common sci-fi trope because it’s a convenient allegory for the uncaring systems that determine the quality or continuation of human life.
Of course, investors are loving this because there is no such thing as bad press.
CoastalCoder · 24m ago
Please don't parody regional accents to imply stupidity.
ivewonyoung · 38m ago
Those changes got permanently reversed in the recent passage of the Big Beautiful Bill by a one-in-a-year reconciliation process that was able to pass with only 51 votes in the Senate with zero votes from the opposition.
Note that the reversal only applies to American software jobs, not offshore ones. So maybe tech hiring is going to pick up again soon. Those changes should've been reversed before they took effect in 2022 by the govt at the time.
awesome_dude · 23m ago
> It's a cheap win with investors who want to hear about AI initiatives to lay off yet even more of this department, because it doesn't matter if the quality of service declines, there are no market or regulatory forces that are punishing this well enough to ever expect firms to stop breaking it, let alone fix it
There's also some argument that, if people cannot get customer service to "help" they stop asking for help - driving that cost down.
And not having to remedy issues in the product = no repair/replace cost
And people are then left with only a few options, one of which... buy a replacement... which in a restricted market is a WIN because more money coming in...
blindriver · 32m ago
You absolutely misunderstand Section 174 and you are spreading misinformation.
The only companies this affected are those right at the margins of becoming profitable. It doesn't affect new startups and it doesn't affect established businesses. And if you are at the margins of becoming profitable you have likely accumulated more than enough tax credits for all your losses.
The changes to Section 174 is not the explanation of why software engineering jobs were lost in 2022. They were lost because every company overhired from 2020-2022 and they have to absorb it given the drop in activity once the Pandemic was over.
choilive · 1h ago
I had the same thoughts, there are clearly indicators that the weakness in the labor market started happening before LLMs and AI took over popular discourse.
All the more reason to believe that while correlated, LLMs are certainly not the largest contributor, or even the cause of the job market weakness for young people. The more likely and simple explanation is that there are cracks forming in the economy not just in the US but globally; youth employment is struggling virtually everywhere. Can only speculate on the reasons, but delayed effects from questionable monetary and fiscal policy choices, increasing wealth gaps, tariffs, geopolitics, etc. have certainly not helped.
MontyCarloHall · 17m ago
>The paper notes they did an adjustment for the end of ZIRP. I dont know enough econometrics to understand whether this adjustment was sufficient,
They did it by regressing the number of jobs y_{c,q,t} at company c, time t, and "AI exposure quantile" q, with separate parameters jointly controlling for company/quantile (a), company/time (b) and quantile/time (g). In the paper [0], this is in Equation 4.1, page 15, which I have simplified here:
log(y_{c,q,t}) ~ a_{c,q} + b_{c,t} + g_{q,t}
Any time-dependent effects that would affect all jobs at the the company equally irrespective of how much AI exposure they have (e.g. end of ZIRP/Section 174) should be absorbed into b_{c,t}. (A potential flaw of this method is that ZIRP/Section 174 may have disproportionately affected junior positions with high AI exposure, e.g. software engineers.)
They normalized g_{q,t} with respect to October 2022 and quantile 1 (least AI exposure), and plotted the results for each age group and quantile (Figure 9, page 20).
The decline for quantiles 3, 4, and 5 for the youngest age group is pronounced, and only happens in mid-2024. The plots shown in the article are misleading, and are likely a reflection of ZIRP as you say. The real meat of the paper is Figure 9.
Yeah my company started stepping up outsourcing in 2023. We also started some AI projects. The AI projects haven't made much progress but the outsourcing is at an extremely advanced stage.
deelowe · 1h ago
I personally sat in meetings in 2022 where we adjusted staffing projections in anticipation of AI efficiency. Sure some of it was "overhiring," but the reality was that those staffing goals were pre-ai. Once they were updated, that's when the layoffs started because management didn't want anyone who didn't have an AI or big data background.
MontyCarloHall · 1h ago
Gen-AI was still extremely niche in 2022; ChatGPT didn't come out until the end of the year, on 30 November, and it was pretty much just a toy curiosity until mid-2023 when GPT-4 came out. I am very surprised that leadership at your company was seriously discussing the business impact of AI that early on.
deelowe · 57m ago
Chat gpt wasn't considered a toy... Not sure where you got that. We were interested in what openai was going from very early given the founders' history.
MontyCarloHall · 50m ago
It absolutely was considered a toy by most people when it debuted in late 2022. This was the era when memes abounded about how ChatGPT would dutifully answer the query "what's the world record for crossing the English Channel on foot?" [0] or "what weighs more, a pound of feathers or a kilogram of bricks?" [1]
Most people didn't start taking ChatGPT/gen-AI seriously until mid-2023, when GPT-4 became widely used.
I mean very good job staying on-top of new tech but you're not actually trying to imply your not the anomaly in that regard right?
raincole · 41m ago
Huge difference between people with forward thinking and without, right.
danans · 48m ago
> SWE figures dropped mid-2022 (almost magically in line with interest rate hikes) and LLM-copilots werent introduced for another year
It was pretty clear by late 2022 that AI assisted coding was going to transform how software development was done. I remember having conversations with colleagues at that time about how SWE might transform into an architecture and systems design role, with transformer models filling in implementations.
If it was clear to workers like us, it was pretty clear to the c-suite. Not that it was the only reason for mass layoffs, but it was a strong contributor to the rationale.
Many large companies were placing a bet that there were turbulent times ahead, and were lightening their load preemptively.
giantg2 · 48m ago
My company had multiple call center modernizing projects going on starting around 2021, including many NLP based routing and task upgrades.
kraig911 · 45m ago
in addition to this detail I might add I can't remember the last time I had a customer service call that took place with someone stateside. It's easy to point to AI when offshoring for favorable interest rates is really the reason.
atleastoptimal · 36m ago
It is possible that multiple trends are coalescing
1. layoffs after web3 hiring spree
2. End of Zirp
However I think now, in 2025 is it impossible to reasonably claim AI isn't making an impact in hiring. Those who disagree on here seem to be insistent on some notion that AI has no benefits whatsoever, thus could never cause job loss.
I sense some conflation of causation/correlation at hand.
klik99 · 47m ago
Exactly this - I've said it before and will say it again - new technologies emerge in response to trends, often to accelerate existing trends and does not create them.
I see a few explanations for what you're saying, and those might be true, but I strongly believe part of it is investment (particularly VC, less so PE) has hit diminishing returns in tech and which means less subsidized "disruption", which means less money to hire people. AI becoming hugely popular right when this was happening is not a coincidence. And it's not just startups, less investment in startups also mean less clients for AWS and Azure. A16Z / Sand Hill switching to AI is not them just chasing the latest trend, it's a bid to reduce cost on people, which is the most expensive part of a tech company, as the only way to extend their unicorn-focused investment strategy.
carabiner · 1h ago
2 things can be true. I was applying for jobs in 2022 and we all knew that the market was crap then because of overhiring during pandemic.
com2kid · 33s ago
[delayed]
fuzzfactor · 1h ago
The economy is destroying the jobs and AI is just the raven on the shoulder of a stumbling bull . . .
safety-space · 53m ago
yes, this is it
blindriver · 37m ago
Why can't it be both?
cyanydeez · 39m ago
I think it's lipstick on a pig. We've seen tech companies collude before, and I'm guessing they're doing it again, trying to drive down the price of talent and make their employees less demanding.
thrawa8387336 · 1h ago
Coordination, plain and simple.
TuringNYC · 1h ago
Could you please explain more. Very interested in anything that explains the massive hole in the timeline.
thrawa8387336 · 1h ago
This is a second hand anecdote but someone commented here or on X, that basically he was on a cruise and overheard two heads of HR of 2 big Co's talking to each other about shenanigans.
Would not be the craziest considering that AI has to make a ROI. Even if it's not up there yet to do so organically. If you annihilate the entry labor market, then after some time, you have no choice but to use AI because there is no one remaining with the skills. AI is lower than entry level -> No one is hiring new grads -> There is no new talent being developed -> use AI for everything!
cactusplant7374 · 31m ago
> This is a second hand anecdote but someone commented here or on X, that basically he was on a cruise and overheard two heads of HR of 2 big Co's talking to each other about shenanigans.
I have no idea what this means.
tejohnso · 16m ago
It's describing the setting for a conspiracy theory. Multiple (in this case 2) people (in this case powerful ones) getting together and deciding that a certain outcome would be mutually beneficial.
And the second paragraph details the conspiracy is to work together to remove a certain type of employee in large numbers, so that AI tools have to be used in order to make up for that loss.
johnsmith1840 · 1h ago
Openai also invented time travel to coordinate with companies.
Nothing to do with thr mass exodus and offshoring of US jobs.
The BPO industry is GROWING the opposite of standard AI understanding ideas.
Also call center is a good one I was doing research myself and call center jobs overseas have GROWN pretty rapidly over time these jobs are moving not vanishing.
reliabilityguy · 1h ago
Coordination of…?
farai89 · 4m ago
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.
gerdesj · 15m ago
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.
whatever1 · 11m ago
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.
jampa · 42m ago
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.
causal · 4m ago
Also AI hype is sucking all capital out of traditional hiring
throwmeaway222 · 58m ago
My guess:
25% musk (getting rid of 80% of twitter showing every CEO that at least half the staff is sleeping)
25% ending of zirp
20% CEOs BETTING on AI
10% Hiring Andrei instead of Andrew
10% Ending DEI
10% Actually AI
daxfohl · 2m ago
Betting on AI would be "hire some AI engineers and see what they can do", not "get rid of some engineers and see what happens". I think what you mean is "scapegoating AI".
I think there's also a pinch of "we've run out of ideas / high margin projects" or "we're tired of funding 'platform 2.0' projects that end up creating more problems than they solve".
But generally I agree with your assessment. Especially the Musk effect I think gets underplayed.
tasty_freeze · 48m ago
Do you really think that companies were hiring that many unqualified people for DEI reasons? Just about every time the subject came up there were people saying it was just lip service and theater -- put a DEI statement on your company's public-facing webpage and pretend they actually did anything.
strken · 4m ago
If you rephrase "ending DEI" to "reducing the risk of getting sued when laying off x% of staff", does it make any more sense?
Nobody knows how to actually hire competent staff because it's a constantly changing bar: if you give people leetcode, they start cramming leetcode; if you review their GitHub profile, they start spending disproportionate amounts of time on projects; if you give them take-homes, they spend 5x the recommended time; if you give them real-world problems in a timed interview, that's probably harder to game, but some candidates will send a completely different person along. On top of that, some people just interview really well but aren't good 9 to 5. At a big enough company, you've always got a list of people who you incorrectly hired and want to get rid of.
DEI is a minor barrier to doing that for some cohorts. It's not that you hired incompetent people in XYZ groups to bump up your diversity numbers, it's that you hired incompetent people in every group and now you're unable to get rid of some of the ones in XYZ.
Also, let's not forget that some people are just genuinely sexist and/or racist and/or whateverist, either consciously or unconsciously. What happens when those people aren't held back by HR as strongly?
leetrout · 42m ago
Maybe. I had two different hiring managers in two different companies explicitly tell me and others to hire women or PoC. Yes it was illegal. No one cares.
IX-103 · 7m ago
[delayed]
ajsnigrutin · 7m ago
> No one cares.
Those same people then wonder why more and more young men are turning towards "right wing" parties, how trump won, why AfD is on the rise, etc.
NothingAboutAny · 20m ago
Only anecdotally I can tell you when I was at a medium sized (~200 employee) fintech business in Australia, I was told by my engineering manager to hire any woman or PoC that applied.
but in my 2 year tenure I think only 1 of either applied, both were hired immediately.
pzo · 17m ago
also (affected mostly startups):
- collapse of Silicon Valley Bank
- section 173
phyzome · 59m ago
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.
iJohnDoe · 40m ago
Not wrong. Also want to point out the continued layoffs at Microsoft, Intel, etc.
wewewedxfgdf · 8m ago
Interest rates have way more impact.
yowlingcat · 1m ago
[delayed]
foxfired · 40m ago
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.
resters · 46m ago
Demand will shift to AI-literate knowledge workers whose productivity will be increased.
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.
causal · 7m ago
I also have yet to see anyone meaningfully differentiate between "AI is taking jobs" and "AI hype is causing stupid executive decisions" and even "AI hype is sucking up all the capital that would normally go towards hiring".
dr_kiszonka · 27m ago
What are those "augmentative" jobs that fresh CS grads can transition to?
redwood · 54m ago
Does the ADP data include international employees in lower cost geos?
g42gregory · 37m ago
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.
t0lo · 18m ago
They are the mouthpiece of the empire- but they aren't promoting products- they're promoting the dissolution of social and communal values and trying to erase the idea of intellectualism- the nyt casually, the atlantic explicitly, and the new yorker very covertly- it shows the powers that really run america have a deep ambivalence towards it and will likely hop to the next centre of prosperity when they get bored. The american experiment is over because they decided it to be. How weak and sad your country has become.
ipnon · 30m ago
They are the party mouthpieces of the Democrats. It’s not useful as an empirical analysis or philosophical exercise. But it is useful as a declaration of belief for a certain cohort of American elites. So the revealed news is really that the Democrats will add “we must weaken or break up Big Token to increase youth employment” to their midterm and presidential platform. Whether this is sound macroeconomic policy is beside the point.
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.
iJohnDoe · 43m ago
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.
charcircuit · 1h ago
This article continues to propagate the misconception that young is the same thing as inexperienced.
dataexec · 56m ago
We can assume the two are so strongly correlated that it's not worth it to differentiate. Even though many exceptions exist.
charcircuit · 35m ago
I don't think it's true at all. Considering it only takes 2 to 3 years to rack up the 10,000 hours to "master" something, young people can get very good at a lot of things. The biggest barrier in my opinion are child labor laws that get in the way of people getting experience.
iammrpayments · 11m ago
If you do something 80 hours per week for 2 years it totals 8320 hours? This is far from average behavior even if you have nothing else to do except coding.
programjames · 58m ago
I think you are referring to
> The strategic thinking that goes into longer-horizon tasks may be something LLMs aren’t as good at, which aligns with why entry-level workers are more affected than experienced workers.
I think the article is talking in generalities, so on average entry-level software engineers have less experience with long-horizon tasks (e.g. months-long development), though there are definitely the exceptions that prove this rule.
swayvil · 53m ago
As individuals less work is a good thing obviously. But as a society less work is a bad thing.
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.
For customer service, my explanation is that companies literally do not care about customer service. Automated phone trees, outsourced call centers whose reps have no real power to help a customer, and poorly-made websites have been frustrating people for decades, but businesses never seem to try to compete on doing better at it. It's a cheap win with investors who want to hear about AI initiatives to lay off yet even more of this department, because it doesn't matter if the quality of service declines, there are no market or regulatory forces that are punishing this well enough to ever expect firms to stop breaking it, let alone fix it
For a software engineering business, the Tax Cuts and Jobs Act (TCJA) of 2017 significantly impacted how software costs can be expensed under Section 179. While Section 179 previously allowed for the immediate expensing of many software purchases, TCJA reforms restricted this deduction primarily to "off-the-shelf" software. Custom-developed software and internal development costs are no longer eligible for Section 179 expensing and must now be capitalized and amortized.
Under the TCJA, Section 179 cannot be used for software that a company develops for itself. This includes the direct costs for the engineers, programmers, and other personnel involved in the development process.
The report not addressing this elephant in the room is a disappointing.
Of note, the OBBB reinstated the ability to deduct R&D, so businesses are no longer required to capitalize and amortize R&D expenses (including software development).
https://warrenaverett.com/insights/one-big-beautiful-bill-se...
No comments yet
However, that's back for tax year 2025, so why aren't we seeing the jobs come back? Maybe it really was 174 then, but AI now?
It was only just reinstated, so it's probably too early to see the effects.
I also expect that despite the restoration of Section 174, companies realized that they not only overhired during ZIRP, but also that they don't actually need that headcount, given the outcome of Musk's Twitter layoffs. There were so many prognostications that Twitter would imminently implode after downsizing from ~8k to ~1.5k employees, and when these claims never came to pass, it was a wake-up call to the rest of the industry [0].
[0] https://www.livemint.com/companies/news/elon-musk-fired-80-p...
But that "everything app"? It hasn't happened. The money transfer app ("Twitter Payment Platform")? Still MIA.
Oh, they sure did: https://news.ycombinator.com/item?id=34617964
Even in that thread, a lot of people were saying "it's only been three months, give it a bit more time."
although I think entry level is still in shambles, for now
https://blog.pragmaticengineer.com/section-174/
1) AI CEOs, like the Anthropic CEO quoted in the article, Sam Altman(till recently atleast), Perplexity CEO, Microsoft CEO etc. It brings them new VC money, investment, and customers who "don't want to miss out on this big trend".
1.5) Media heavily pushing the above CEOs quotes, probably just for clicks and engagement, brings them money
2) BlueSky appears to really hate on anything AI to the point of personal attacks, putting AI tweeters on blocklists and bans etc.
3) Reddit, especially the antiwork side
Out of control AI is a common sci-fi trope because it’s a convenient allegory for the uncaring systems that determine the quality or continuation of human life.
Of course, investors are loving this because there is no such thing as bad press.
Note that the reversal only applies to American software jobs, not offshore ones. So maybe tech hiring is going to pick up again soon. Those changes should've been reversed before they took effect in 2022 by the govt at the time.
There's also some argument that, if people cannot get customer service to "help" they stop asking for help - driving that cost down.
And not having to remedy issues in the product = no repair/replace cost
And people are then left with only a few options, one of which... buy a replacement... which in a restricted market is a WIN because more money coming in...
The only companies this affected are those right at the margins of becoming profitable. It doesn't affect new startups and it doesn't affect established businesses. And if you are at the margins of becoming profitable you have likely accumulated more than enough tax credits for all your losses.
The changes to Section 174 is not the explanation of why software engineering jobs were lost in 2022. They were lost because every company overhired from 2020-2022 and they have to absorb it given the drop in activity once the Pandemic was over.
All the more reason to believe that while correlated, LLMs are certainly not the largest contributor, or even the cause of the job market weakness for young people. The more likely and simple explanation is that there are cracks forming in the economy not just in the US but globally; youth employment is struggling virtually everywhere. Can only speculate on the reasons, but delayed effects from questionable monetary and fiscal policy choices, increasing wealth gaps, tariffs, geopolitics, etc. have certainly not helped.
They did it by regressing the number of jobs y_{c,q,t} at company c, time t, and "AI exposure quantile" q, with separate parameters jointly controlling for company/quantile (a), company/time (b) and quantile/time (g). In the paper [0], this is in Equation 4.1, page 15, which I have simplified here:
log(y_{c,q,t}) ~ a_{c,q} + b_{c,t} + g_{q,t}
Any time-dependent effects that would affect all jobs at the the company equally irrespective of how much AI exposure they have (e.g. end of ZIRP/Section 174) should be absorbed into b_{c,t}. (A potential flaw of this method is that ZIRP/Section 174 may have disproportionately affected junior positions with high AI exposure, e.g. software engineers.)
They normalized g_{q,t} with respect to October 2022 and quantile 1 (least AI exposure), and plotted the results for each age group and quantile (Figure 9, page 20).
The decline for quantiles 3, 4, and 5 for the youngest age group is pronounced, and only happens in mid-2024. The plots shown in the article are misleading, and are likely a reflection of ZIRP as you say. The real meat of the paper is Figure 9.
[0] https://digitaleconomy.stanford.edu/wp-content/uploads/2025/...
Most people didn't start taking ChatGPT/gen-AI seriously until mid-2023, when GPT-4 became widely used.
[0] https://miro.medium.com/v2/resize:fit:1400/format:webp/1*yJs...
[1] https://pbs.twimg.com/media/Fpl09fAakAE1cFW?format=jpg&name=...
It was pretty clear by late 2022 that AI assisted coding was going to transform how software development was done. I remember having conversations with colleagues at that time about how SWE might transform into an architecture and systems design role, with transformer models filling in implementations.
If it was clear to workers like us, it was pretty clear to the c-suite. Not that it was the only reason for mass layoffs, but it was a strong contributor to the rationale.
Many large companies were placing a bet that there were turbulent times ahead, and were lightening their load preemptively.
1. layoffs after web3 hiring spree
2. End of Zirp
However I think now, in 2025 is it impossible to reasonably claim AI isn't making an impact in hiring. Those who disagree on here seem to be insistent on some notion that AI has no benefits whatsoever, thus could never cause job loss.
I sense some conflation of causation/correlation at hand.
I see a few explanations for what you're saying, and those might be true, but I strongly believe part of it is investment (particularly VC, less so PE) has hit diminishing returns in tech and which means less subsidized "disruption", which means less money to hire people. AI becoming hugely popular right when this was happening is not a coincidence. And it's not just startups, less investment in startups also mean less clients for AWS and Azure. A16Z / Sand Hill switching to AI is not them just chasing the latest trend, it's a bid to reduce cost on people, which is the most expensive part of a tech company, as the only way to extend their unicorn-focused investment strategy.
Would not be the craziest considering that AI has to make a ROI. Even if it's not up there yet to do so organically. If you annihilate the entry labor market, then after some time, you have no choice but to use AI because there is no one remaining with the skills. AI is lower than entry level -> No one is hiring new grads -> There is no new talent being developed -> use AI for everything!
I have no idea what this means.
And the second paragraph details the conspiracy is to work together to remove a certain type of employee in large numbers, so that AI tools have to be used in order to make up for that loss.
Nothing to do with thr mass exodus and offshoring of US jobs.
The BPO industry is GROWING the opposite of standard AI understanding ideas.
Also call center is a good one I was doing research myself and call center jobs overseas have GROWN pretty rapidly over time these jobs are moving not vanishing.
"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.
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 think there's also a pinch of "we've run out of ideas / high margin projects" or "we're tired of funding 'platform 2.0' projects that end up creating more problems than they solve".
But generally I agree with your assessment. Especially the Musk effect I think gets underplayed.
Nobody knows how to actually hire competent staff because it's a constantly changing bar: if you give people leetcode, they start cramming leetcode; if you review their GitHub profile, they start spending disproportionate amounts of time on projects; if you give them take-homes, they spend 5x the recommended time; if you give them real-world problems in a timed interview, that's probably harder to game, but some candidates will send a completely different person along. On top of that, some people just interview really well but aren't good 9 to 5. At a big enough company, you've always got a list of people who you incorrectly hired and want to get rid of.
DEI is a minor barrier to doing that for some cohorts. It's not that you hired incompetent people in XYZ groups to bump up your diversity numbers, it's that you hired incompetent people in every group and now you're unable to get rid of some of the ones in XYZ.
Also, let's not forget that some people are just genuinely sexist and/or racist and/or whateverist, either consciously or unconsciously. What happens when those people aren't held back by HR as strongly?
Those same people then wonder why more and more young men are turning towards "right wing" parties, how trump won, why AfD is on the rise, etc.
- collapse of Silicon Valley Bank
- section 173
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.
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.
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.
> The strategic thinking that goes into longer-horizon tasks may be something LLMs aren’t as good at, which aligns with why entry-level workers are more affected than experienced workers.
I think the article is talking in generalities, so on average entry-level software engineers have less experience with long-horizon tasks (e.g. months-long development), though there are definitely the exceptions that prove this rule.
Perverse.