But then people who work in actual tech companies come in and explicitly say they are not hiring any juniors anymore specifically because AI is good enough to do most of what juniors do, and that senior engineers can now write 3x as much code, etc.
There seems to be a desire for a narrative that AI really just can't replace productive work, and that it's all a mirage. However it seems just like common sense that if an AI can do junior-engineer-level coding work, that a company has less reason to hire a junior engineer.
https://www.employamerica.org/labor-market-analysis/dont-bla...
The New York Fed has also released some research suggesting remote work has been a major factor differentially affecting early career workers.
https://libertystreeteconomics.newyorkfed.org/2026/06/remote...
"But then people who work in actual tech companies come in and explicitly say they are not hiring any juniors anymore specifically because AI is good enough to do most of what juniors do, and that senior engineers can now write 3x as much code, etc."
If you want an anecdote: the media company I work for just started hiring interns and juniors in software career tracks again after a lengthy hiatus.
Every now and then, I actively try to make an LLM replace my tasks, and fully do greenfield projects I would accept- I don't see it. It's very good, no doubt. But I have or have been given the project parameters, and just like with a junior, failures in communication inevitably lead to breakdowns in execution.
I would have preferred someone who uses a null hypothesis and statistical tests.
Yes, and juniors aren't known for their productive work in the beginning. That's not their purpose. Their purpose is to do the mundane work, because it is important for them to become less junior and more senior.
That is robbed of them.
Which in 5-10 years means the need for senior developers is gonna shoot through the roof.
If senior engineers are even 2x more productive with AI, then it’s like there are 2x as many senior engineers.
Most likely, seniors will be 10x more productive in 5 years using AI. This outpaces the retirement rate.
All the software engineers we need for the next 20-30 years are already in the current workforce.
Only way juniors can rise to the level of seniors will be through independent projects, long unpaid internships/apprenticeships, etc.
The industry will now have heavy gatekeeping built in.
Yup, that's reflected in the data as well, no need to invoke "vibes" or whatever.
https://www.economist.com/content-assets/images/20260516_EPC...
https://www.economist.com/content-assets/images/20260516_EPC...
The likely explanation is that there's job losses happening in some sectors, but it's made up for in other sectors.
This meant that, in my example and given the same scope, one junior person with AI could easily handle multiple projects in 2026 that would require 2 or more medior people as recently as 2025. The only reasons for involving multiple people were time, learning opportunity, or increased scope - not skill.
If we paused AI progress today, and ignored the option of adding agents recursively, the current state of things still provides several alternatives to any honest & informed manager:
- Altruistic or optimistic: Keep all current jobs, and move the scope until there's a need and/or funds to hire more devs;
- Pragmatic or neutral: Keep the current jobs, but not hire anyone else as long as the scope is balanced against time (our current situation);
- Cold or pessimistic: Lay off anyone who doesn't fit the leanest model at current (or acceptably lower/higher) scope.
I don't know whether this translates into a "job's crisis", and I can only hope it won't. But hope is not a career strategy, scope is not inexhaustible, and the political, economic and social pressures are quite strong even if you removed AI.
Are you talking the big 10? Or "tech companies" in general?
> AI really just can't replace productive work
Okay. Show me the productivity gains. Those are measurable. Why is the "AI is ready" crowd never prepared to show this?
> that if an AI can do junior-engineer-level coding work
Then you have no competitive edge and most of your output probably cannot be copyrighted.
I mean, if we want to not talk about economics that’s fine, but can the AI actually do junior work at the same price? What if we don’t look only at quarterly reports, and instead include the value of having people knowing about the business having to explain it to others, who then learn it and can improve it over time?
I think it’s clear the AI is strong, there’s no doubt about that, but that’s not the whole picture.
Even if we assume it can then not hiring Juniors still doesn't make sense - where will seniors come from in the future?
If AI was truly strong, we would be seeing signs in the job market. And we would certainly be seeing a lot more subscriptions and demand for these services. For most people, AI does not improve their lives. For a lot of them, especially younger people, it makes their lives much harder and sadder.
It's always been difficult to put a number on that value, which is the problem for the MBAs running the show. There's no number on the P&L assigned to tribal knowledge, and improvements that can be made by those with that knowledge and experience within the business.
It's a mistake that businesses keep repeating, over and over again, yet never actually learning the lesson. And now the industry is going to repeat it again, until there's enough pain that they realize that lost value and start rehiring again.
How? By making a 5 person team a 4 person team + AI.
If i think about my co-workers (not excluding myself) from last 15 years, there was always someone you would accept just because it was better than not having that one person. If i can now replace them with more tokens/better models, man i wouldn't hesitate (of course i know what this means on a person to person level :/)
Nominal wages being up rejects the hypothesis that folks are being downsized into lower-paying roles.
The average person earns an average salary doing something very different to the stuff we on HN stress about.
My best advice for folks that want to get into software now is be willing to do it cheap for awhile and then jump once you've developed some skills. If you were getting into this industry for the money you're properly fucked and I hope you didn't load up on debt. If you're passionate about building stuff there's still room but the path forward is a lot murkier.
The reality is most firms are running out of projects to take that make economic sense.
Note: ECONOMIC SENSE. This has nothing to do with refactoring for the sake of refactoring. Its all to do with earnings growth with respect to the cost of capital.
Then you might as well work for yourself.
> getting into this industry for the money
I can make more money doing HVAC but I'm tired of being on hot roofs.
> the path forward is a lot murkier.
If you're just here for the money go somewhere else. If you're here because you love computer science then ignore these people and do the work. If you can't find a company get a dayjob and do it for yourself.
Definitely feel the murkiness. I've been programming as a hobby for over ten years and only recently started wanting to do it professionally. I'm actually wondering if there's a path for me.
I haven't been on the market for long luckily, but as an independent consultant I went from getting 2 to 3 contacts per week to none per months.
https://corvi.careers/blog/global_software-engineering_jobs_...
BTW, cool site!
IMO, AI coding tools have materially improved SWE productivity compared with two years ago. The open question is whether demand for new products and services will expand enough to absorb that productivity gain. If demand does not increase correspondingly, then it may only be a matter of time before we see meaningful job market pressure, starting with SWE.
The big job sites feel rife with unqualified applicants and outright fraud, leaving me frankly unsure how to proceed.
It feels like a crisis of candidates-to-job-openings matching, and AI has made it much worse.
>> The May jobs report reinforced this with nonfarm payrolls jumping by 172,000, confirming that there are no signs of workers being replaced by ChatGPT.
If you leave out healthcare, 2025 had massive job losses overall, with Boomer bedpan cleaning bringing the net number up to just above zero.
Job numbers get revised every month, in a negative direction.
New grad unemployment is high and trending higher.
New jobs exclusively are held up by addition in healthcare industry, almost every other sector is seeing some negative movement.
A lot of job openings, a good chunk of them, are just fake jobs where the company has no intention of filling them.
Pretty bold for someone to ignore all of that and come up with a claim like that.
Job opportunities differ by state and de-growth hostility to business policies and crony investments. Where I am, layoffs and offshoring continues. I hear new grads are increasingly opting for the skilled trades, which is interesting given they aren't getting use out of their degrees.
https://blog.glassdoor.com/site-us/wp-content/uploads/sites/...
https://www.glassdoor.com/blog/worklife-trends-2026/
But, CEOs figured out that if they blame layoffs on AI, stock go up a lot. Reporters know that anxiety about AI drives the clicks that write the checks.
I do think that AI anxiety is making HR around the world anxious about hiring. That's my best guess for why everyone is finding they need to apply to 500 jobs to get anywhere. So, AI is making it hard for you to find a job not because it took your job, but because HR is reading ragebait and turning it into a self-fulfilling prophecy.
This is what a buddy who works at a major consulting firm told me about the hiring trend, we are returning to all the pre-COVID hiring norms.
Lucky engineer types like me in the last group, but many of my college educated friends in that middle group. Shit is rough
Further, the graph shown is pretty noisy and I'm not sure the upward move which counters the downward trent is statistically significant.
the claim "hey there is no AI job crisis", when previous SWE of 6-figures now takes job dishwashing in McDonalds + one more gig as Uber driver + food delivery gig is "job creation! now they have 3 jobs!". does not make any sense.
You talk to any SE and it’s obvious we’re not running out of work to do since these tools became available
but of course, it is not just AI. Software is consolidating and automating even without AI, that's the whole point of software.
I’ve been unemployed and actively looking for a job for about 6 months, the longest stretch of active job searching I’ve ever done in my career. Several close friends who work in tech or tech-adjacent fields are in the same boat. Anecdata on Hacker News or LinkedIn tells the same story.
A chart showing “total number of jobs” is not meaningful. I took a temp job in a metal shop to make ends meet while I wait for the endless rounds of interviews I’ve now gone through with 4 companies. It pays less than half of what I was making before, and I am barely making ends meet. It’s not sustainable, even though the pay is more than fair for the work.
There are also a lot of job openings for home health care workers, or seasonal resort workers, that used to be filled by immigrants. Those jobs are not going to be taken by any of the engineers who just got laid off by Meta.
I have the strong impression that people who write articles like this are very disconnected from the reality of the economy right now, and that their curiosity ends at the line on the chart they cooked up to make a contrarian point.
"A chart showing “total number of jobs” is not meaningful."
It is meaningful in answering the question being asked, which is whether the hype around labor displacement, which has been growing for nearly five years now, is actually occurring in a way that would justify some of the higher valuations for AI firms.
The "AI jobs crisis" is generally understood to mean a sustained downturn in demand for all labor due to AI substituting for labor across a huge range of tasks.
I know plenty of people looking for job for months at this point. I am looking for 3+ months and have a few leads finally but nowhere close to offer. Remember, people have to live in these high cost GEOs (SF, Seattle, NYC and others). Since COVID, the US economy is effy at best with housing prices doubled, rents and household expenses doubled. Just for something simple it costs $200 subscription like mowing the lawn. Pulling stats out of some dashboard that a rando made is not how people live day to day life. Remember, people lie with statistics all the time.
So yes, there is a significant job crisis. Because Ycombinator invested in AI companies this narrative is suppressed here and moderator DANG (one of the worst) heavily moderates anything critical of AI and YC.
Personally, this makes me suspicious of calling out mods for censorship. This, in turn, makes me suspicious about the truthiness of any of the information presented.
I'm not agreeing or disagreeing with any of the opinions or experiences expressed.
I have other account but I don't want to post with that or have it banned by censorship king DANG.
If you don't want to believe that is fine but don't make surreptitious accusations.
And the link echoes that there is a job crisis. But not an AI job crisis; rather a COVID frenzy fallout job crisis.
The prediction that AI will lead to mass unemployment is largely a widespread fear more than an observation.
AI washing: On the one hand, you have periodic mass layoffs that could be related, but also could be "AI washing", with the companies citing AI to sound progressive when, in fact, they just over-hired.
Multiple independent sources find no broad signal. (Anthropic's labor research, Yale Budget Lab, Apollo's chief economist Torsten Sløk)
I'd still like to know more about the junior situation, though.
At this point the harness/applayer matters more, as different models perform better or worse on exploit classes depending on the prompt, tuning, and various other parameters.
Of course, by the time HN hyperfixates on a topic, it's already been executed on and HN is too late.
In larger models, these fine tuning techniques work more reliably/robustly. Because of this many usecases tend to prefer larger models. It is possible to work the same behaviour into the smaller model, but it requires more effort. But it's one-time. And smaller models are usually much cheaper. People make a tradeoff along this curve.
This is observed at few-B scale upto hundred-B scale. No way for us non-anthropic/openai to fine tune beyond that of course.
This literally is a brand new phenomenon; it's only a matter of time.
But yes, middle management would qualify ;-). My manager seems spooked by LLMs. Loves to use them to write his emails, but seems to internalize that since they're doing his job for him at this point, his boss may figure it out.
People are right to point out that hiring is nothing like the post-pandemic years, but it's not clear that it's any tougher out there than, say, 2018. This is from the perspective of someone with a lot of industry experience, though. I can't speak directly to the experience of junior engineers.
But to keep this out of a low-value vent, my experience has been that the _threat_ of it is there, but in my small corner of the world/industry, lots of layoffs that would have happened anywhere may just be categorized as "AI" layoffs, but the wild manpower reducing benefits aren't really there. The larger an org gets, the more of your job is dedicated to human stuff, and you can just get some of the code part done a little more quickly.
Would be interesting if we could measure how much effort is put into agentic coding harnesses, frameworks, and theory, vs labor saved using them.
At least I don't have to review PRs, we have built a team around the concept of trust and not needing to check each other's work.
There's simply too much work to start nit picking in PRs, everybody's very talented and a good engineer you can trust. They ask for feedback when they want it.
And no, we're not producing slop, if anything AI has helped improve the codebase a lot and the harnessing has definitely raised the bar at all levels.
But sure as hell, the job sucks now, and I hate it. And it's not just due to the amount of work.
On the other hand, I'm just finishing an agent-heavy piece. After getting it set up, it's been some of the most mindless and soul-destroying work I've had the displeasure of in a while. This stuff will be near minimum wage in a few years, totally unskilled babysitting.
AI really hasn't been all that bad for work, by volume at least. I know where I want to focus my efforts though.
Longer term, I don't know. I'd happily take something more secure if it came along, as long as it's not childminding for an agent. Super busy and bored out my mind the last few weeks, the worst sort of work.
If this kind of argument were generally valid, it would imply that:
- all change neither accelerates nor decelerates, which is absurd, on the face of it;
- the initial stages of a deep change are always surface-visible; for instance, cancers announce themselves when they begin to gestate, rather than when they metastasize
- A few recent points of data of questionable significance outweighs a hypothesis with considerable support from reason, intuition, and other (unpresented) data. For example, the plight of recent CS grads, which _is_ new, and _is_ on graphs, just not the one the author here chose.
So, since these implied claims are self-evidently _false_, it means that the author would, at a minimum, need to provide an explanation as to _why in this one instance, these considerations do not matter_; for example, the author could have argued that the graph positioned at the center of their argument is the one to look at (as opposed to, say, recent CS grads,) but that _itself requires further argumentation._
It also does not account for the other obvious possibilities; e.g.,that there is a delay between the (as it were) lightning and its thunder; or that even strongly nonlinear effects would have shown up by now in the metric chosen; etc. But since these contributions were not included in the original post, I have no choice but to discount it.
Much of the "AI job crisis" rhetoric was PR comms to manage conversations around corporate restructuring (even ZIRP is a lazy PR comms excuse).
Most decisionmakers by 2025 already agreed they didn't expect AI to have a significant impact on hiring [0].
I've pointed out the reasons ad nauseum on here but no one listens [1].
[0] - https://www.bloomberg.com/news/articles/2025-12-10/wall-stre...
AI has grown dramatically in capability since last year as well so I'm not sure 2025 data holds today.
That said, I agree with you that AI capabilities have grown massively, but the only way you aren't burning tokens with marginal RoI is having someone with credible domain experience using Claude Code or Codex hand-in-hand.
Basically, no one wants to spend $200K TC on training a junior who will most likely be using an AI tool already. The math works out more positively at a significantly lower TC.
*They fund these start-ups to get a good return on investment so they can get even more money. As the economy overheats the number of places to invest with a reasonable return falls so they are left with the high-risk stuff to invest in. I'm not sure what they want the money for, though, since they could already afford most of the things I would find useful...
1. AI ATS systems have made posting jobs "cheap", such that too companies post jobs that don't exist (ie "ghost jobs") to keep up appearances they're hiring or just to keep people in the pipeline in case they hire. This is a huge waste of everybody's time and should be illegal;
2. The hiring process itself gets increasingly Kafkaesque. AI screening, automated online tests, unpaid take-home work, etc. You have to get pretty far until a human gets involved. 10+ years ago this didn't happen because people needed to be involved much sooner and that's expensive;
3. In a lot of companies, getting employees to interview people is unpaid extra work effectively. They say it's important. You might even get dinged for not doing it. But anyone who has done it realizes pretty quickly a bunch of people who shouldn't get interviewed are getting interviewed and management doesn't care, even though employee time is expensive, because you essentially have to "make up the time" so it's still "free";
4. Even if you go through all that and get hired, you get laid off within a year such that income isn't dependable and you end up wasting a ton of time on the job-seeking process itself.
I've been thinking about this recently and high-information is part of the problem. In years long gone, it was hard to reach applicants so you'd have a small pool of higher-relevance candidates applying for a job. Say 10 people applying for 10 jobs. The odds were better. It was less work on everybody's side.
But now you have 200 people applying for 200 positions. This wastes everybody's time but the problem is that companies have offset this by pushing filtering onto these automated systems. People still need to enter all their bio information, etc. So it's just much more inefficient inherently even if the job opening is legitimate.
Long gone are the days when the vast majority of people just have one job. Now it's 2-3 part-time jobs because companies are exploiting a legal loophole where they only have to pay benefits for full-time employees. And then you have people doing "gig" work, which is often sub-minimum wage when you factor in expenses (eg driving for Uber and not factoring in car wear).
On top of all this we, across the Western world there's an increasing youth unemployment crisis. In 2008, entry-level jobs basically disappeared overnight and never came back. Well, that just got worse post-pandemic.
https://www.pewresearch.org/social-trends/2020/01/09/trends-...
The jobs openings / unemployed close but a bit larger than 1 is the expected value when nothing is turning the economy completely upside down right now. It means there is no new shock that is too fast for people to adapt.
What is surprising, yes, because I could easily name half a dozen shocks that I would expect to see there. But it seems US people are adapting quickly right now, or things are canceling out.
Now I have no need for anyone from a coding perspective.. I can keep up with multiple clients requests with a breeze. I don't have to manage anyone. I type of my phone while I'm on a walk and work gets done for me.
So yeah... it's not good.
If you had to hire a concierge for a hotel, would you rather take a guest-oriented, quick-witted junior or an AI? If you could either take a Waymo or an 80-year old driver, what would you take?
I don’t believe hiring managers think that one-dimensionally. Roles are unique and in some roles experience is more important than in others. Plus, juniors already balance their lack of experience with lower salary expectations.
The much easier explanation for the anecdotes: companies are more cautious at the moment and if you only have a few positions to hand out, you rather take the proven hand than risking it on someone who hasn’t shown yet that they can do it
A lot of people and companies working left and right building things which are obvious that they are coming.
We do not know yet how that impacts even more people.
And AI already removed certain jobs.
Maybe all the job cuts from ai were filled by fixers of ai output
Or maybe, no one ever heard of jevons paradox. Or maybe everyone ignored it and preached job apocalypse as risky but a high reward marketing tactic
As though the decades of work I've put into my career means I should want a job as the hamburger man or working as a ditch digger. No shade to burger flippers or ditch diggers, but these are not jobs that I'm trained for, nor are they jobs that I remotely want to do. So for me, aviation expert, programmer, ML engineer, weird IT generalist, guy with a math degree who speaks a couple languages, and isn't exactly fully capable anymore, the idea that jobs in some other field (like an RN or an Oncologist or Electrician) are just something I can pivot into is just hilarious. It's such a shallow and ignorant take. Even the premise that all jobs are equally distributed and there aren't jobs that are more or less in demand at any time is a real funny take too.
I'm so glad I started working for myself, because honestly, seeing this dogshit analyses from supposed experts means I'll be able to keep making money for a long time just be actually trying when I need to think about something.
If we need to spend just as much on salaries, while shelling out $700B/yr on AI, how does all that spending get paid for?
Even if AI is an absolutely bubble, and SpaceX, Anthropic and OpenAI all cease to exist in a year... there's simply no way that AI has not fundamentally changed work. Even if I was forever pinned to the local models I'm running and the agent harnesses they use, I would never write code for work the same way.
But I lived through the rise of the web. I remember serving dynamic websites through cgi (which meant a new instance of an interpreter was spawned per user session). I vividly remember great JavaScript books saying things like "never use JavaScript for core functionality". I recall Java engineers saying Ruby on Rails was a toy and would never take off, that Python offered nothing over Perl and that "rich web applications" where never going to replace app native interfaces. I remember when the MVC pattern from the early Smalltalk days being dusted off and repurposed for web applications, completely changing how we designed software for the web.
And all of that is just software. It wasn't until the pandemic that ebooks replaced print books in share of academic library circulation (reversing a decades long trend of reduced circulation).
In my daily use of agents for coding and other forms of problem solving, while it is a wild accelerant, it's also clear we have not even started scratching the surface of how to think about building things with these tools.
I suspect we'll adapt to AI faster, but having lived through one major tech revolution, transforming work still takes some time. I'm not surprised we don't see an immediate jobs crisis.
Not to mention the completely separate topic that huge classes of employees were not and are not all that productive, so boosts in productivity don't imply lost jobs. That would require a boost in productivity combined with pressure to create concrete value with less, looking at the SpaceX IPO we're still a ways away from working about how efficiently we create concrete value.
It just gives a general "job" going up, but
* how many of those are real jobs? I've seen the amount of scam/fake jobs go up a lot since 2022
* How many are jobs only for PhD with 20 years expereince?
* How many jobs for actual juniors / newcomers?
Just taking the aggregate is a bit scummy
