Sadly no mention of Section-174, which is a contributing factor. As a reminder, TCJA (2017) amended Section-174 (with a 5-year deferral) to require amortization over 5 (domestic) or 15 (foreign) years of all R&E expenses, and included all software development under R&E. This went into effect in 2022. This causes unexpected employer tax bills over money that has been spent on employee salaries and over which these employees are also taxed.
There have been repeated promises by congress to fix it but these efforts have failed so far. The latest attempt, HR7024, received overwhelming support in the house but stalled in the senate over objections by Idaho senator Mike Crapo and was later voted down (mostly by senate Rs, but also notable by a handful of progressive Ds).
This is a scapegoat- I see this thrown around in dev forums on reddit all the time. I asked my CFO if this had any impact (small fintech startup of <200 employees globally), and he said it has no impact on our planning or any real material impact to our books.
This was a small speedbump in the financial statements of alphabet/meta/apple etc, and both revenues and earnings climbed despite the impacts. They kind of even out after 5 years regardless.
Big tech is most affected as they have the biggest r&d budgets? You still have to show a profit to owe taxes which most startups don't have and big tech does. There is a huge bag of tricks you can use to hide or show profit as well.
Find me a case study where this has curtailed hiring.
wow big companies with access to cheap financing aren't bothered by it but it destroys smaller companies that operate on a cash-flow basis, crazy who would have suspected that such a tax regime would be supported by big companies
Can you lay out your understanding of how this works?- I get the feeling you don't even understand the basics. You are only taxed on profits so why would financing even be brought into the conversation?
yeah, sure: money spent on dev salaries that are dedicated towards "research" aren't fully deductible in the year they are paid, instead you have to depreciate the cost over a number of years. that means your "profits" are higher in the first year because you can only claim a small percentage of the money you actually spent in cash terms. fine for a large company that has access to to financing and/or a large warchest to absorb the hit: in the long run it pencils out, but for smaller/newer companies that are run on a cash basis its brutal and very easy to get into a situation where taxes exceed profit.
i think everything should be done on a cash basis in general, so what do I know
> Knopp, the CEO of Pequity, says AI engineers are being offered two- to four-times the salary of a regular engineer. “That’s an extreme investment of an unknown technology,” she says. “They cannot afford to invest in other talent because of that.”
> Companies outside the tech industry are also adding AI talent. “Five years ago we did not have a board saying to a CEO where’s our AI strategy? What are we doing for AI?”
Sounds like the AI fad is sucking all the oxygen out of the room. I wish these CEOs had the courage to tell their stupid, uncreative, trend-following boards: "Nothing. Our AI strategy is nothing. We will continue to invest in our core business rather than chasing the latest trend you read about in Harvard Business Review." But they don't. They have no backbone and are about as uncreative and unfocused as their boards.
CEO where I work is driving this push - he's got enough stupid, uncreative, and trend-following ideas without the influence of a board. He came from a sales background and I don't think those guys can resist this. It's not going well either - just had a major customers board of governance kill a deal for our ridiculous llm toy. So glad I never exercised my options.
Nearly every corporate leader I've worked for charges after any trend that's hot. They don't have a tech background and can't tell if something is real or hype. They don't want to seem like a fool that's missing out, so they act a fool and chase everything.
Why not tell them something like “but let us wait till AI is strong! Then we don’t need to hire these pesky AI engineers for big money, we can rent access from OpenAI and ask it to build AI for us…”
Who are these experts saying AI isn't a good R&D investment? Like I understand an LLM is just a token predictor but it's a damn useful one if you use it right, and all the other applications are still improving too (some examples in another comment of mine)
Treating it like an R&D investment isn't a problem - you only spend what you can afford to lose on R&D. In our case, we have no R&D budget and any money/time spent playing around with llm's leads directly to degradation of the core product and risks losing customers.
Imagine if big corp waited until "the web was strong" to go online, they'd have been dead in the water. Sears was fully positioned to be Amazon, and look at them now. It's just not good business sense to not explore new things and invest in R&D - of course not every investment will pan out but you don't want to miss the important ones.
Copilot is a massive help every day. Image generation continues to improve and even animation recently made a quantum leap to the point I can see it being a useful tool to animators [1]. ChatGPT was super useful to me before they seemingly hindered it. I listen to awesome music made with AI voices all the time these days[2].
It's not just a "fad", it's a domain showing enormous potential and we need to thoroughly explore it. This is how it works with every new advancement, and it's a good thing. It would be monumentally stupid to do "nothing" when you have an R&D budget to use.
Why doesn’t our H1B system adjust the yearly number of slots based upon these metrics? From a tech worker, I just see the H1Bs as flooding an already saturated market.
I think that we should entirely abolish programs like H1B that leave people tied to a specific employer for a long period of time (and similar visas like the SWV in the UK).
These programs mean you have literal second class citizens in a the workplace who desperately need to avoid losing their jobs for fear of their life being completely upended. This disrupts the labour market and internal corporate culture significantly. Instead we should have work visas lead to permanent status much faster, but give them out more sparingly.
USA should adopt the EU model: The visa is not tied to the employer but to profession. You are only allowed to do work in the professions that is described in the original visa/residence permit reasons for up to X years, which should avoid the "in the employers mercy" situation.
EDIT: this also relieves a lot of burden on the system, as you can easily change jobs between companies by simply mailing the job change documents to the foreigners office - they will just take a look at the new position that you have taken up and simply send back the confirmation that the residence permit is still valid, instead of trying to issue a new visa with all bureaucracy that comes with.
I think that's a totally reasonable compromise which would reduce abuse both by employer and employees. However actually determining what a specific job role entails is quite complicated. Take a look at the SOC codes in the US or NOCs in Canada. Even if you figure that out, auditing people to make sure their day-to-day matches their notional job code is pretty hard.
Well, at least in Germany, the "profession" field is a free text field, there are no codes. The official is only interested in that whatever you are working as closely matches the original profession on your visa, so you can easily argue that your new title of "SRE Engineer" matches the original "Software Engineer", which they will accept as long as you are a working person in the IT Field.
The Swedish visa system is a compromise, at least when I moved here 10+ years ago.
My first work visa was tied to the company who sponsored my move, it lasted for 2 years and in that period if I changed jobs the new employer would need to sponsor me as well. After 2 years I renewed my work visa, this 2nd visa is not tied to an employer and I could freely move jobs.
As cheap as employees are they can always be cheaper. Labor costs are a huge input cost in many businesses, cost savings there adds up to a large amount of money for very few people who can pay politicians to act in their interest.
Which is why H1B wont be curtailed and why employees are blamed for inflation.
"I opened the window to listen to the news. But all I heard was the Establishment's Blues." - Sixto Rodriguez (This In Not A Song, It's An Outburst)
My hot take on the whole Amazon RTO thing is they are trying to ramp up visa workers
Amazon is the no.2 in h1bs in the US, and no. 1 in i-140.
1) Layoffs mean they have to pause both of those. Amazon even had to stop all I-140s in February until 2025. So instead of more layoffs just RTO 5 days a week. If everyone stays, great.
2) If people leave / no one applies to new jobs - now you have a case to bring to the US govt. that Americans with the required skills aren't applying
I wonder if that sort of measure would have far too much lag to be effective. Also, what signals do they go off of? Ask workers and there's too many, ask industry and there'll never be enough.
I think there are some labor board requirements to report layoffs? Maybe that could be used. Could implement some rule where you can't do layoffs at all while employing visa holders. No way that an industry going through massive layoffs needs to bring in people on visas while unemployed citizens can't get hired.
Better tooling replaces job openings. Not just AI tooling but better tooling in general. Instead of hiring that new developer, our new tools make our existing developers more efficient so it is no longer needed.
I agree fully but this thread is about why there are fewer tech jobs. Everything is a cycle. New tooling improves developer efficiency, management looks at this as a sign that they can hire less and scale back, devs get overworked, management struggles against the reality that they need to hire again. This happens all the time.
Even the baseline of improving dev efficiency is unproven. These types of decisions are being made with zero basis in facts, it’s just the whims of people in charge.
Yeah but have you ever been at a company where there weren't and infinite list of things that needed to be done? It has way less to do with "having enough developers" and more to do with, we can't afford any more developers.
If you have a machine that prints money, you will flood that machine with whatever resources it can handle, push it so it can handle more, and duplicate the machine so you can double your printing.
If you have a machine, that sort of maybe brings you value, you are going to be very cautious about how many resources you give it.
The article states due to the cost of AI engineers, they are hiring less non-AI engineers. I doubt there’s real data to work with there, more of a vibe. I wouldn’t doubt it is having a measurable effect.
I have 17 years experience and burned out in my current role. given up on job seeking while coasting until i get fired. 95% sure this will be my last ever job in tech.
as i read recently, "god's plan for me doesn't involve linkedin".
Sorry to hear you're burned out. Do you want to work after your current job? If so, curious why you feel this is your last tech job. Is it because of ageism or something else?
thanks. i no longer find the work interesting, and the job market is very punishing. and i've seen 2-3 responsibilities being repackaged into a single job became normalised in the last 10 years.
When I was starting college twenty years ago, the prevailing sentiment was all programming jobs were going to be outsourced and that getting a CS degree was a fool's errand. I literally had a guidance counselor try to talk me out of it. I did it anyway.
When I finished in 2006, there was a real vacuum in the developer space, presumably because everyone got the same advice I did, and jobs were relatively easy to find.
By about 2008 there was just no one. We had open developer positions for months with zero applicants.
I have seen vacuums and gluts come and go.
My advice is just if you are interested, go for it. The market will come around.
True. However the future may be different if the government doesn’t make changes. It’s easier and cheaper to now hire off-shore devs than Americans. There goes your pay raises and jobs or competitive salaries.
Don't worry. There is plenty to be done. These stories come up every few years. Previously it was outsourcing to India, or low-code tooling, etc. AI can make you more productive but there will still be plenty of jobs.
I think that if AI "eats the world", which it will, all other jobs will go away. AI will, for instance, design and run factories. That means most jobs in factories will go away. Those factories will need developers, sysadmins, networking, data analysis, ... to operate, and they don't really have those functions today. It won't be 100% the same as a developer job now, but much closer to developer than to engineer.
Tech jobs are definitely coming back. I think more than ever before.
This, I think, has an incredible potential to advance the economy, and will require A LOT of programmers, AI/ML engineers and sysadmins, most in places where there is almost zero need for them today.
Have you ever been in a factory? Perhaps you’ve seen automated material handling systems, robotic assembly systems, and so on. Have you seen the plumbing? Have you seen a pump fail and be replaced? We are a long, long, long way from the lights-out factory. A steady reduction in the number of machine operators does not a self-sufficient plant make.
It's interesting to think how much better and cheaper robotics would have to be to cover the super long and hairy tail of issues that a single human operator can handle for <$100k / year.
Developing economies purchase a lot from the west. Germany and Japan have export-oriented stances because their local markets are developed to the point that local growth is slow...
Seems like the right call. I had a few friends who graduated in 2002. Some of them tried and failed to find a job straight out of Uni, and that really screwed them. They started working retail to keep food on their table and then they were somebody with no relevant experience but not really suitable for the new grad hiring pipeline.
OTOH those who went into the master's program in 2002 were fine, they came out a couple of years later when the market was hot.
> Find some specialization/niche you're interested in software wise, and start learning + building in public about it.
This is a very different and mostly non-overlapping skillset compared to focused product engineering at a company. I believe it's a good way to get you there if you have the time and drive to develop those skills. But then if you do are you going to be satisfied with the prize being a job where you don't use them?
I'm not discounting this as effective, I just don't think it can work for very many people for several different reasons.
I read that article yesterday too, I read Blind a lot as well
The software engineering jobs are there. The candidate selection and interview process is worse than ever. The article was relatable about the ratio of applications moving forward.
This article had some other gaffs though, like the engineer lamenting that they had to go outside of their network to apply to job for the first time. What? They responded to a recruiter DM on linkedin for the first time? Sound the alarms ya’ll
A 5% pay cut? Big tech employees join the mid size and startup world demanding no less than $400k total compensation…. at the beginning of their journey. Which market was the engineer in this article in? Not enough info
Definitely more indication that a boom is over, and plateaus are in.
Yes, the title is very much clickbait, granted there's some discussion SWE related matters but really they're referring to the "tech industry" as whole, and which industry isn't squeezing at the minute...
Some nontechnical workers in the industry, including marketing, human resources and recruiters, have been laid off multiple times.
Yeah, I think a a better headline is the section header about "No More Red Carpet". Tech jobs are down from 2020? The tech job market in 2020 was unsustainable smoking hot like nothing I've ever seen. I remember the 2000 bubble and bust. That bubble was nothing compared to the one we just came out of and this "bust" is nothing compared to 2002. This is a correction. And you can attribute maybe 5% of it to AI, but the other 95% is just interest rates. The ZIRBs are learning about real life. With tightening officially over, we are likely going through the worst of it right now and the worst of it is still pretty good.
I'm a full stack dev with over a decade of regular web app development roles and I can't even get a phone screen right now. I've been applying for months.
I have gone through 10 final rounds… I feel there are so many candidates that small things are getting called into question. I’ve gotten dinged for being too verbose, for being too terse, for being nervous, for having to debug, for not debugging enough… I just had one where I got knocked for having the right time complexity, but using two more methods than needed.
> Moore, who learned how to code by taking online classes, says not having a college degree didn’t stop him from finding work six years ago.
> Myron Lucan, a 31-year-old in Dallas, recently went to coding school
> nontechnical workers in the industry, including marketing, human resources and recruiters, have been laid off multiple times
THIS is the issue. If you have a CS degree (EDIT: And Experience, internship or FT), you will find some job.
If you went to a coding bootcamp, are a self taught programmer, and/or are working in a non-technical role you don't have an edge compared to the multiple recent and mid-career graduates who have the same experience AND a college degree.
I've been frank about this multiple times - if you don't have fundamental skills and knowledge in Algorithms and Systems (an actual Data Structures, Algorithms, and OS class), no offense but you are just a script monkey.
As a hiring manager, I pay the big bucks for demonstrated domain experience AND critical thinking skills. If I need boilerplate code EPAM and TCS (and in 10 years AI-driven code generation platforms) can do that for me.
> If you went to a coding bootcamp, are a self taught programmer, or are working in a non-technical role you don't have an edge compared to the multiple recent and mid-career graduates who have the same experience AND a college degree.
unless you have an edge that isn't a CS degree. CS degree is not the only edge in this industry.
I started seeing the trickle of bootcamp grads coming in maybe 7-8 years ago. Many of them were just as good as the CS grads. They have no issue finding new jobs today, some are even at FAANG
Someone who went from bootcamp to coding 6 years ago shouldn't have any issue finding a coding job today, if they're actually good and progressed over those 6 years and have worked a year or 2 as a senior engineer at a good/known company. My guess is they aren't
Companies today want people who can hit the ground running. Less jobs for juniors, less jobs for ramping up within 3-4 months.
> They have no issue finding new jobs today, some are even at FAANG
I went this way and then later taught at a bootcamp, so my network is heavily skewed towards this kind of engineer. It is not true that they have no issue finding jobs. The very few that ended up at big tech companies, or who have a prestigious non-tech degree, are mostly doing fine.
The ones who just got normal developer jobs at normal non-famous companies are having a hard time from what I know of my personal connections. They were more likely to get laid off and less likely to have found something else after, or are stuck in their current role because they can't even get interviews. And I mean the 2015-2018 bootcamp cohorts, firmly mid-level with 5+ years experience. I don't have many connections to the newer waves of code school grads and am not talking about them.
I think the other commenter is basically correct. The time when you could have a secure development career without a degree quietly ended over the last couple years.
I'd take a risk on hiring a non-traditional student who taught themselves to code and is a maintainer of a decent sized OSS project. Not on a bootcamp grad with no coding experience outside of the bootcamp.
> Companies today want people who can hit the ground running. Less jobs for juniors, less jobs for ramping up within 3-4 months.
Looks like market is correcting against translators[1].
Usually, when a recession(or hypothetical one) occurs, expensive work forces can’t be paid adequately or adjustments become expensive, so employers try to suppress wage by doing risky layoff[2]. That then becomes a trend for other insignificant players to also follow. After a layoff, expectation for inflation adjustment disappears because one has bills to pay, family to feed, hence desperation leads to wage stagnation.
Then suddenly after 1-2 years hiring boom begins and the wage skips a few cycle of correction but most with decent experience can command good pay despite lower adjustments.
One problem with such play is that, big players can hold out for few years before the dire consequences of the layoff begin to emerge, but before things get serious they start hiring again. The sad part is that, small players are screwed because they are now left with consequences and expense of hiring new workforce.
Also this time it is not only about wage suppression, a massive covert selling of “AI as replacement for those pesky expensive code monkeys” is being peddled(if you look between the lines carefully on layoff/poor productivity/ ai increases productivity play). This will have consequences much dire for the smaller players scrambling to drink the AI coolaid without understanding how it helps, because by the time money had been wasted, the cost will be manyfold to recover the productive workforce and some will fail and become insolvent, but AI will become common place and just another tool like class generator, advanced IDE, no-code tools, uml-to-class generator.
AI(we call these LLMs now) of current generation is expensive and it needs massive investments to become viable(think of main frames of older decades), but we are past the age of when saying something will be useful and getting money because most business saw similar(blockchain/no-code/outsourcing/offshoring) trends before. So the same old play, “productivity gains with this cool tech minus the cost of a human” gets big management excited. Once investment keeps flowing, more development and innovation will take place in AI(LLMs) space and it will get cheaper eventually(mainframe to your smartphone) but until then peddling is needed to fund these gambles.
I see for what the trend it is, I hope others see it too.
[1] I call people learning to code(without formal education on the topic) and then write rote code on what they learned as translator. Code-Monkey is very insulting and should not be a term we use against other human beings.
[2] Like I said, it is short term risky because if things go south too fast they can hire immediately and of course new hires usually have 150% more productivity on their first 6month to a year so any lost productivity of the layoff is adequately compensated quickly.
I have huge hopes that the expense of training and more importantly running the current generation of AI models will mean that for the first time since the 90s the cost of customer will be higher than the maximum possible advertising revenue per customer.
Nearly free is the price where markets fail completely.
Imagine being the customer of the next google rather than the product.
Also looking myself for the last couple of months. No luck. 8 yrs exp, but generic full stack web experience in financial services (this might be my problem, having generic and common skills).
Could you elaborate on your exp? Wondering if you have specialist skills that are making you more in demand.
Honestly it's nothing too special. Boring enterprise .NET development, mostly focused on moving companies from old monolithic WinForms etc applications to modern stacks. Have been doing this for about 8 years.
Might be a localized thing though, I work and live in The Netherlands.
I’ve actually had a few recruiters reach out in the last few weeks. It’s nowhere near the daily calls I used to get, but it seems like there’s been a little bit of an uptick.
There have been repeated promises by congress to fix it but these efforts have failed so far. The latest attempt, HR7024, received overwhelming support in the house but stalled in the senate over objections by Idaho senator Mike Crapo and was later voted down (mostly by senate Rs, but also notable by a handful of progressive Ds).