Ask HN: Is it just me, or does the job market for IT seem bad today?
I am a developer with nearly 20 years of experience trying to find a new job to try new things and naturally to get a higher salary, but even though I haven't been seriously involved in selection processes before, I used to feel that recruiters would respond right away, like you had to fend them off, and now they don't.
Do you get the same feeling?
119 comments
[ 4.8 ms ] story [ 165 ms ] threadbut a simplified TL;dr is that companies used to be able to expense that year's R&D costs (including salaries for developers etc) in that same taxable year. now, that is no longer the case, and companies must amortize those R&D costs over 5 years.
this is not ideal.
[0] https://www.axios.com/2024/01/20/taxes-irs-startups-section1...
https://blog.pragmaticengineer.com/section-174/
A recent bill has given relief for the domestic amortization but is still stuck in the senate.
https://gop-waysandmeans.house.gov/wp-content/uploads/2024/0...
https://tradingeconomics.com/country-list/interest-rate
Most businesses never operate from an operational budget but rather debt. When rates go high, the businesses have to cut the fat. Which generally means the job market tightens.
This is but 1 factor in why the governments should be obligated by constitution to balance the budget.
That model might work if they cut spending during good times.
As is, the modern government spending model is that they (almost) constantly increase whether things are going well or poorly. Then the next time there's a downtown, government spending as a portion of the (shrinking) economy is a higher % and therefore it takes more to "stabilize" it this time around.
Think of driving down the road. If it's a normal straight road in good weather, you minimally adjust the wheel back and forth and if something happens, you have many degrees of freedom and even momentum pointing you in the right direction.
Now think of a road at night in a storm with lots of curves. As you tighten your hands on the wheel and steer back and forth, things are quite a bit different. The direction you want to go now vs 10 seconds from now are a little different. And a disruption - however so small - may need a bigger reaction to avoid and a bigger after-reaction to recover. Momentum can easily work against you.
Nope. Keynes himself created this idea and then disproved it. There's 1 key factor of why this fails EVERYTIME.
Lets say the downturn was just the uranium mining industry. Government comes in, adds spending to stabilize. Then you ASSUME the government will stop once the issue is over. Nope. Never ever does. Politicians who created the subsidy move on to other issues and those temporary government measures become permanent.
This is a big gap for the difference between theory and practice.
On occasion you'll hear about the need to end corporate welfare. Nobody ever actually does it, but then when that industry bounces back, they are supercharged by free subsidies. Then the next politician is told, you can cut but it's going to cause people to lose jobs. Which is true.
Government should never be in this business to start with.
SPY trajectory is misleading as it is dominated by a small set of companies.
https://www.zerohedge.com/markets/fed-chair-powell-admits-bi...
It's a bit like RT: Before the invasion, lots of naïve Westerners used to buy that drivel too; now only the most deluded conspiracy kooks do. It's just that secondary site metastases like this ZH aren't as obviously descredited, so all the "keeping an open mind to the extent my brains have fallen out" conspiracy-flirters, semi-kooks, and other morons keep sucking the propaganda teat.
What, older people aren't allowed to work from home?
Like the benefits you can get with working from home if you e.g. have a partner and children are much much larger in practice then when you are a single student. Through also the challenges you have to overcome to make working from home viable. With the main challenge being proper working from home needing more space (in general) at home which with high apartment prices isn't always easy.
I have outsourced coworkers that I tutor. They are almost universally pathetic at software development and are a net negative overall. I get the most done during the times upper management can't overlook the lack of progress on current projects and leave the offshore people to their own devices for a spell, or have fired them as we go through cycles of hiring and firing. Management just can't stop touching themselves thinking about those sub-minimum wage salaries but somehow never grasp you get what you pay for.
Want a job in software development?
- Love it for the work not the money. If you're in it for the money you'll never be good.
- Don't chase fads. I know little of crypto, web 3.0 (what version are we on now anyway?), javascript frameworks, and use AI sparingly. But I know programming and software development well enough to ace the little pointless tests they give in job interviews.
- Don't play games with getting hired. Use people not technology to find a better job. That means being presentable and personable and networking. Indeed is not networking it's a marketplace and you're the product.
Everyone has reasons not to attempt these things, but believe me when I say I can do it so can you. I have pretty good social anxiety but we're all bozos on this bus and I just keep that in mind at all times.
It was their legit policy to fire you if you tried negotiating your salary.
[1]: https://github.com/bilbof/whoishiring-ratio
"This script outputs how many job seekers there are vs how many jobs."
Isn't it actually showing the ratio of jobs / job seekers? It does seem so by the chart and the code but the text makes it sound the opposite. (Or is it my wrong reading?)
Without that, it still says pretty much nothing.
guys from 37 signals are really pushing the post vercel era.
And I mean the desirable, well-paying jobs, not the $20/hr service/food jobs that anyone can get.
If you're a doctor, and you apply for a hospital job, do they at least reply to acknowledge your application? Do they then send a rejection letter if you're simply 'not what they're looking for'? If they then start to consider you as a candidate, do you have to 'trial cure' a couple of patients, take a cell-function exam, and endure 6 rounds of other interviews and screens before getting on the final shortlist?
Tech applications mostly just go into the void (sometimes there's an autoresponder to at least say your application has been received into their system, but often not even that), occasionally they'll give you a rejection letter if you're not suitable, but often they won't bother, and then at any time during the arduous rounds of tests and screenings you may be ghosted; just they can't even be bothered to tell you you're no longer in the funnel.
I really don't believe this hiring experience is something inherited or 'caught-up' from other industries.
I think this shows a better version of trade school is needed for software developers; computer science degrees don't seem to be enough to have confidence in someone's abilities. At least doctors have rigorous schooling and training, arguably even a bit too long and expensive.
Yes, you can refer to me as a non-techie if you want. I don't find that offensive, but it would just be inaccurate. I am tech-adjacent.
But for a long time it had been really really good to a point that even people which fairly lacking qualifications could often easily get a job in a very short time frame (through also with some common biases/discrimination e.g. against age, and some positions, like very very well paying ones are still not easy to get).
And I would say that times are about to be over soon.
Honestly what has me worried the most is that nonsensical bias against age often found, as if having experience for 5,10 years is good but dear you have 20 years of experience in server development not its bad .... wtf. Especially given that I don't buy many of the argument people in favor of it often push. Sure the tech landscape is constantly changing but a lot of the base principle aren't really changing that much only the tooling and how you apply them, even with LLM supported programming that is still true. Similar arguments like older people being less willing to work overtime as IMHO misleading as every time I have seen companies (mostly with young people) do a lot of overtime the results where quite subpar in quality to a point that it created some huge issues often just one or two month down the line (or well days later, in some cases). And sure some people really get stuck up with certain approaches with age (but not all), but as far as I can tell that as less of a risk then blindly following the newest trends (e.g. framework) without a deep understanding of the things it implies mid to long term (which is what I have often seen with younger people).
I mean I'm not old yet, but if the industries doesn't fix its mess that will be a problem for me in 10-15 years. (Generally we also need to fix teaching and sustainability of a lot of software development. But that is a different topic, one which has gotten worse through LLMs as they allow keeping unsustainable approaches for a longer time).
For every cutting edge startup chasing young people to abuse, there are a dozen existing companies with tons of legacy technical debt. They don't pay FANG salaries, but they will pay.
In my experience lately, it seems like Engineering and Developer positions have taken a hit, but backend infrastructure and DevOps type positions are still very hot, especially if you have cloud automation experience and know your way around Linux.
I'm not claiming to be very knowledgeable about AI/LLMs or the job market or "the economy" in general, but I'd be absolutely shocked to find that the AI stuff has any impact on the current IT job market.
I know that some people are using AI tools to help generate code and stuff, but it seems like those tools can't even generate correct code a lot of the time- and that's with a professional human programmer asking it to. I can't imagine that these tools are increasing productivity so much that it could actually reduce hiring.
I’ve heard people on HN suggest it’s definitely helping them, but I’ve yet to see it in person. So ehh maybe they are just working on very different things.
Not to mention that I'm so skeptical right now that I highly doubt these AI tools are helping IT workers be even 5% more productive.
Historically, every time a new web framework comes out that swears it'll make web devs so much more productive, it doesn't seem like that ever causes a noticeable change in the job market.
Maybe that “more” will be in other areas of the business or economy and thus lead to less headcount, but given how profitable tech is.. I’d assume they’d be redeployed to build more product
I think things will get better, the effects of the above will lessen with a rate cut and probably a repeal, but the golden era is past.
But things will likely improve eventually. Probably due to a mix of laid off developers finally finding new jobs, a decent percentage of developers/engineers leaving the industry, and new companies being founded in future.
The interest rate rises were intended to contract the economy and reduce wage inflation pressure by increasing unemployment, and they seem to have had that effect.
The AI bubble is also causing a lot of companies to believe in the idea of replacing their staff with AI.
Replacing generic slogan writers with AI might work, but replacing IT workers with AI at this point will be funny to watch (from a distance).
What changed is remote work. My company got rid of US/UK based employees and only hires LATAM contractors by the hundreds.
I've only heard back from three companies but I've probably applied to over 100 (5+ yoe).
I haven't had a recruiter contact me on Linkedin for over a year, in 2022 and before I was being contacted by recruiters on Linkedin all the time.
Companies have different modes - mass layoffs, small layoffs, no hires, only hiring replacements, a few new slots open, many slots open. In early-mid 2022 we were in the last category and it's moved to one of the other modes for most companies, except for LLM hires.
There are different sources of information, I found this helpful ( https://interviewing.io/blog/when-is-hiring-coming-back-pred... ) from a company with some insights into tech hiring.
My profile is marked as remote only. The offers are all on-site.
So I have my doubts that it is a match.
Now it's like submit 1000 resumes, get interviews from 4, spend 20+ hours preparing and doing interviews, get ghosted at the end.
I have a friend who works for local government. They have an open GIS programmer position. I've done GIS programming for government (as a contractor), and it's not hard stuff. She says they received many 100s of resumes and are currently interviewing 30 people. They had so many people apply that they required each applicant to submit a very involved project beforehand just to weed people out. So, there are literally 30 people out there right now, putting in 20+ hours to complete a project just to get a job that pays probably $110k / year.
I remember back in college thinking that take-home project assessments were great (as a candidate). Even after getting fleeced on those a few times (e.g., spent a week completing the assessment, everything seemed picture perfect, submitted it, waited a week, automatic rejection with zero feedback or explanation, wtf?).
These days, I would rather go through 5-6 rounds of leetcode+systems design interviews. Takes much less time, each one gets easier and faster, and zero extra overhead with each subsequent interview (while with take-home projects, each one will take around the same amount of time). And I am not even gonna dive into the whole cheating issue with take-home projects.
I simply got other things in life at this point that I would rather do, than to spend over 10 hours in a week on a single take-home project, while i can spend that exact same number of hours to complete ~2 full leetcode-centric interview loops at 2 different companies. And that’s not even mentioning that take-home project interviews also have leetcode and other rounds as follow-ups anyway. As well as the whole “the company doesn’t care about this being easily cheatable” attitude with take-home projects not really making me want to work there (acting as a weak proxy for the general quality of candidates they are hiring).
P.S. I think of leetcode as the best lowest common denominator. Yes, there are certain niche interview processes that are very one-of-a-kind that are great, but they are imo not easily transportable outside of the company conducting those. The one that came to my mind personally was Dropbox (the whole process from start to finish was fantastic, and they treated candidates like reasonable adults, no lowballing on offers or any other bs), with one specific round being “prepare to discuss any project you worked on before in depth, and a week ahead of it send a very short 1-2 paragraph description of what the project is to your interviewer.” It was one of the best interviews I ever had in my entire life (not talking about my performance on it, but rather in terms of how in-depth it went and how just great it felt to talk all that systems design and decision-making).
P.P.S. Another one I can think of is Netflix, but I didn’t interview there myself yet. A couple friends of mine did, and apparently there was almost no code involved at any point there (no leetcode, no take home projects, nothing). 3 years later, those firends are still at Netflix and seem to be loving it, so Netflix definitely did something right there.
I was fed up to see entitled wannabees be paid a lot of money to slow me down by forcing me to do some parody of scrum that they learn at one bootcamp, feeling they were in any way important as they drank they in-house latte during multiple remote meetings a day.
I much prefer a market where people I work with are actually paid according to their skill and ship code rather than power point or worse, an update to the code of conduct to include some inclusive edge case.
I would prefer a balance of all worlds, that would be fairer. But you can never quite reach it, because you pass it on the way up or down quickly as the pendulum swing.
So if I have to choose, I prefer a hard market.
I understand that people need to eat and that it's not everybody's preference, though.
Right, but in the late 90s and early 2000s, unless you were a manager climbing the ranks, you weren't making big bucks.
Even folks at Microsoft, which at the time was one of the most dominant software companies, were only millionaires if they joined pre-IPO. The company wasn't minting IC millionaires.
Now you have a whole host of FAANGMULA companies that do mint millionaires if you stay long enough and get your full vest, and don't get totally unlucky on the timing of RSU grants.
Even assuming rates will start to drop H2 this year and cost of capital will be lower, some of the new-found productivity demand will be consumed by maturing AI tools.
tl;dr this looks like the new normal.
I’m not looking for new mobile dev jobs soon but the market for native looks like shit.
I always thought fullstack meant front end and backend, not front end across multiple platforms.
I have worked on mobile app (iOS and Android) and done full-stack web and app development.
When I apply for places the recruiter literally cannot fathom this "Um. So...Are you frontend or backend?" Both. "...But which one" BOTH "Which do you prefer?" I prefer a company that can handle the concept of someone who can and HAS done it all.
The second reason is AI. The biggest use case for LLMs is writing. They are really good at writing, including writing code. Developer are more productive as a result and companies can do more with less people.
Skilled talent with a good network can find a job fast. I know multiple people who have gotten new jobs within weeks of getting laid off this year.
But I also know people who are absolutely struggling to get a job despite being quite good and putting out lots and lots of applications.
SE job listings have returned to the levels they were at before the huge ramp up in hype a few years ago. Unfortunately FRED data only goes out a few years here.
> I receive screening call in a bad moment.
> Me: "Right now it's a bit difficult for me. Can we do this at another time, like maybe in 30 minutes?"
> Them: "Sorry, I have another interview at that moment. We'll call you tomorrow."
Of course "tomorrow" is euphemism for "never", and I knew it, so I wasn't even disappointed when I didn't get the call next day.
So yeah my guess is that with so many applicants, interviewers are looking for the absolute tiniest reason to reject anyone.
Maybe it got even more worse in these last few months compared to the beginning of the year, or maybe it's specific to the company, or maybe both.
Because earlier this year I got a call early while I was still sleeping (went to sleep late the previous day) but they were chill and even laughed, saying "no worries, is it fine at [some hour]?".