"But for the rest of us, the era of cushy jobs with healthy salaries and light responsibilities might be ending."
Wait. That was a thing? This article also mentions, "Last year saw the demise of middle managers." Is that all this article is talking about?
Because a more appropriate way of addressing this is probably talking about what market shift is currently happening, and how economies experience cycles of expansion and contraction because of industry growth and changes.
My guess is this has more to do with companies optimizing their financial situation through layoffs (reducing number of available positions) and with remote work acceptance enabling offshoring. I doubt it is about middle managers specially - despite being made fun of they are important and necessary in most companies.
Middle managers basically do all the lifting at my company. If anything I feel like the directors and the executives have relatively light responsibilities.
1) Globalization. Many countries have better talent for lower-cost than the US now. Our school system sucks. This is not a bad thing, since it improves lives for hundreds of millions of people, just not here.
2) Automation. GenAI isn't taking over roles, but it is allowing them to be fulfilled with fewer people.
3) Inelastic supply curve. If demand falls, wages plummet. If demand rises, wages skyrocket. This means that something like a 10% layoff can have a disproportionate effect on both job availability and salary.
4) As relevant to HN, K-12 schools are making coding increasingly mandatory, so that skillset is less and less exclusive.
There's an open question about how much of this is a cycle. The Communist Manifesto was published in 1850, and has a nice discussion about how highly-skilled, respected artisan jobs were wiped out by industrialization, where a master blacksmith and unskilled labor had the same salary as factory labourers, which is worth a reread, and seeing which parts turned out wrong, and which parts turned out right.
A lot of what we might be seeing might be macroeconomic trends, rather than a cyclical.
> but it is allowing them to be fulfilled with fewer people.
Are you seeing this in the wild? I see a few people using it here an there to be a little more efficient, but I haven't seen anything better than a marginal improvement yet.
I'd be curious to hear how people are using this stuff for meaningful work.
I work at a law firm and CoCounsel[1] is pretty good. It's not impacting jobs yet, but I suspect next year will be the year planning with automation in mind will happen.
I spoke to a CTO of a firm doing relatively cookie-cutter work (custom software for things like waste management). They were able to lay off a significant portion of their workforce due to AI. I think it's eating in more at the bottom-end of coding right now.
It definitely makes me more productive in high-end, by more than what I'd call marginal, but not by enough that it's risking jobs.
However, the article, and my comment, were more general than coding. The impact is different industry-by-industry, from zero to very high.
More than likely it was the rose tinted glasses he wore.
It just means the remaining engineers had to work harder. AI helps you generate boilerplate, that you still have to check because it will fuck up. It doesn't do entire projects for you.
He described specific functions which AI mostly took over. Those tended to focus on things like developing test cases, where before they had a test team.
I didn't pick up that the people on that team were particularly impressive. I would guess someone still had to check their work because they would fuck it up. I'm not sure how AI is worse.
I think it's a little bit different. When I was in tech, I had a set of bars of quality for employees:
1. Google (keeping in mind early Google was much more selective than today
2. Elite schools / HN
3. Typical tech
When I worked for a non-tech, I learned that there are about a dozen rungs below that. The ladder runs very deep. I was surprised at just how incompetent people were if you stepped into almost any company outside of tech.
I am guessing there is no way I could hire you (a random HN'er) to make custom software for some waste management firm, but they need someone to do it. It's not particular complex or interesting, but it's stuff which needs to get done. That's actually a huge part of the industry. On a prior thread, someone asked how IBM makes their money. The pieces I've seen are based on contracting out this sort of boring but very necessary work from a place that's completely incompetent in tech (random big non-tech firm) to one which is only moderately incompetent (IBM).
My suspicion is that there are uses for technical tasks where a proper software developer is unavailable or the accounting math doesn't work where Gen-Ai would be useful. I'd really like to see how someone puts something like that together and how it functions in a business.
> As relevant to HN, K-12 schools are making coding increasingly mandatory, so that skillset is less and less exclusive.
The majority of those people have neither the ability nor the inclination to build software. The idea that everyone will write code in the future is a fantasy pushed by the business bobbleheads who want any excuse to pay less for engineering talent.
No, but when Microsoft was founded, the pathway was open to Bill Gates since he went to a $$$$ elite private school which could afford a computer. That's right. One computer.
Today, that path is open to everyone. The idea isn't that everyone will write code. The idea is that everyone will have the opportunity to learn, and if 1% run with it, that's 3 million programmers in the US, and 80 million worldwide.
I have no idea what that percentage is (5%? 0.1%), but whatever it is, it's a skill that's growing much less exclusive.
> 4) As relevant to HN, K-12 schools are making coding increasingly mandatory, so that skillset is less and less exclusive
I'm not buying this. Literacy is declining in the US - and it's not like the older people have forgotten how to read, so we can safely conclude that fewer and fewer of the younger people can read and you think somehow these illiterate people are learning to code?
One could argue, and many have, that a bachelor's degree in computer science doesn't prepare one for being a professional developer. A secondary education level developer? On average (discarding the very few exceptional students)? Nope. They're not in your competition pool.
Most people cannot even describe their projects in enough detail that an exceptional developer could build it. Now they are supposed to just do it all themselves because they wrote fizzbuzz in high school once? That seems very optimistic
Literacy's decline can be partially blamed on literally liberal elitists sitting in an ivory tower pushing an absolutely scientifically unproven and bullshit method of teaching kids to read and the dividends are paying real hard now.
You know schools are locally run and controlled, right? Those so-called "liberal elitists" can only do what a locale allows them to do, meaning, those "liberal elitists" aren't the problem.
OP's political undertones aside, schools like Stanford and MIT fake a lot of data.
I have zero ability to tell a school that some study they're relying on is faked if the author is at an elite school. This specific rot is in universities and not local to school districts.
The most common scientific method used for most high-impact education research:
1. Pick a story people (teachers, parents, journalists, fellow academics) want to hear.
2. Bake data to support it.
3. Publish it.
4. See people pick up on it.
5. Tenure.
Other domains of research have related, but different, scientific methods. There is rot local to school districts, but it's not the same rot.
This stuff hurts kids. I'm increasingly moving towards the strict liability side of the debate. If a university publishes results like this, and my district wastes a million dollars on pseudoscience in reliance on that research, that ought to come back as damages. Ditto for people harmed by bad medical research and other domains.
All of this would happen just as surely if the universities were staffed by conservatives as by liberals. It's how we've set up incentives.
Not buying what? That the coding instruction is happening, or that the skillset is less exclusive?
My 14 year-old came to me to help him debug a Python script that he had to write for his Technology class. I didn't even know that they were teaching programming until then. And this is for a rural school district in farming country. So yeah, it's happening.
As far as whether or not the skillset is less exclusive, I don't think that we have much to worry about. It doesn't seem that people are having less trouble learning to program today than they did 30 years ago when I was tutoring in CS. And this is with far more resources available now.
It's not just the US anymore. A child in rural Jamaica can learn to program, if sufficiently motivated and ambitious.
Arguendo, if the US has a gap there, that would only strengthen the macroeconomic trend I described.
Why should I pay an illiterate American 150k, when a Jamaican student can do the same better for a fraction of the cost? And yes, many of the Caribbean education systems are increasingly impressive.
Me too but - remember the jobless recovery from the early 1990's? This time around feels a lot more like it did then. Corporate profits are soaring, Wall Street is happy - there's lots of money being made. Except employees have been left out of the party and are on the outside looking in. This feels very much like that. Then again, that era presided one of the largest periods of economic growth in US history! If history repeats itself then we're all going to be in for a great time in about five years! Of course, that may be of little help to you if in the meanwhile you've depleted your life savings and have been foreclosed on.
Media can't gamify the facts: corporate profits are soaring and the stock indices have been doing great. GDP is up, US-based manufacturing is up, and job growth is at the highest levels in the modern era. The fact that's not helping the guy out on Main Street much is exactly what we saw in the early 90's.
The one thing I would say is different between now and the early 90's is we have a lot more retirees, and retirees as a group are much more exposed to inflation. Though its largely the Boomers who are that group of retirees and they've been through periods of much higher inflation and so if they thought Quantitative Easing was going to last forever then they were just being foolish.
They're also the generation rapidly losing power and so they like to complain so they can still feel like they have some power. If anything, because of the Boomers I'd say the media is being overly pessimistic about our current situation. It's still better than where we were in the early 90's.
These cuts are stock boosts and a result of lending rates rises coming from fed inflation response.
The fed even.reported theyd want unemployment to rise to help.
The inflation itself is driven by corporate price gouging, so its mostly rich people fighting eachother and seeing which part.of the proletariat is easiest.to downsize.
Honestly, the best electives I took in college was microeconomics, macroeconomics, political science, and world history. After those four courses the world made a thousand times more sense to me. It's a shame these courses were electives and I didn't even have to take them, they simply fulfilled an elective requirement. It just seemed to me that if I were to regard myself as having a "well-rounded" education then I should know these things. BTW - I was a lot older when I went to college, I was something like 25 or 26 when I started and had already worked for two startups and had even lived and worked abroad. I had a lot of experience coming into college and so I was very mindful of my electives. My 18 year old self would have never made such smart choices.
People are complaining about "high inflation" and I'm thinking today inflation is finally back to normal. I was around in the 70's. Back then people would have considered the inflation we have today as an answer to their prayers!
The rate of change is much more important than the absolute inflation number in terms of how it effects workers and the economy. It's the rate of change in the inflation number that causes pain, more than the absolute percentage.
There's nothing scientific about the Fed or other central banks targeting for 2% versus, say 5% or 10%. A finance minister in New Zealand suggested the 2% target in the 80's:
https://www.cfr.org/blog/history-and-future-federal-reserves.... Other central banks copied it and now it's set in stone.
What matters is that everyone has an agreed upon expectation for what inflation will be. This is because the economy is based on expectations. Things only get bad when inflation moves suddenly higher or lower and it catches everyone off guard. All their annual projections are wrong and they have to make adjustments.
So the absolute inflation number doesn't really matter, so long as it is steady and predictable. It is true though, that psychologically the higher number may have a negative effect on people.
The 70's and 80's did have big inflation numbers, but what really causes the pain is the fluctuations. And in hindsight it's clear that the actual inflation number wasn't the problem, it was that no one knew what next years inflation rate would be.
The point is: on an absolute basis, today's inflation rates are lower. But that's not what we should be paying so much attention to. It's the rate of change of inflation that matters.
Moving from 2% to 8%, is effectively a 4x adjustment in expectations. Whereas going from 10% to 20% is just a 2x adjustment.
Yes, there's a part of me that thinks "you think these times are bad? They're not so much, you just weren't around when things were really bad, so you have nothing to compare to."
But I temper that with the understanding that people only really know what they've experienced. If today is the worst you've seen, then to you it's legitimately horrific. That's real.
That the situation today is downright luxurious compared to actual hard times isn't really emotionally relevant if you weren't around for the hard times.
> We also have AI based efficiency boosting tools and greater offshoring to contend with
I think that AI is parroted around but there is little to no substance to the claims.
Some activities can and will be impacted by generative AI, but if you are not a copywriter or a graphics artist then your work is not significantly impacted by the current blend of commercial AI services.
Software engineering is a particularly popular target for these baseless claims, as if the critical path of each coding task is writing code. It isn't. In fact, I'm yet to see a single concrete claim on how generative AI impacts productivity beyond filling in boilerplate code and acting as a glorified template engine, and this assumes AI produces code that work which more often than not fails to do.
More importantly, implementing business services to automate business processes has a more direct impact on eliminating jobs. Things like opening an online store instead a brick and mortar shop eliminates the need for a bunch of positions. Nothing AI does comes even close to have that impact. But we still have brick and mortar shops.
It's also important to note that some software engineering positions didn't even existed a couple of decades ago, and they are high paying jobs. A couple of decades ago mobile development wasn't even a field, and right now it's absolutely clear that iOS developer and Android developer are entirely separate specializations.
I think it's possible that the increased efficiency of the employee due to generative AI leads towards cuts short term but growth long term as companies become increasingly competitive with each other due to a more leveled playing field.
At my company I'm in it's 6th reorganization in 13 years. The term used here is golden handcuffs. Where you're paid very well but you're in a middle of nowhere facility so any position you apply for has to be just right or it's not worth it.
50 comments
[ 3.2 ms ] story [ 127 ms ] threadWait. That was a thing? This article also mentions, "Last year saw the demise of middle managers." Is that all this article is talking about?
Because a more appropriate way of addressing this is probably talking about what market shift is currently happening, and how economies experience cycles of expansion and contraction because of industry growth and changes.
And I feel like there are several headwinds:
1) Globalization. Many countries have better talent for lower-cost than the US now. Our school system sucks. This is not a bad thing, since it improves lives for hundreds of millions of people, just not here.
2) Automation. GenAI isn't taking over roles, but it is allowing them to be fulfilled with fewer people.
3) Inelastic supply curve. If demand falls, wages plummet. If demand rises, wages skyrocket. This means that something like a 10% layoff can have a disproportionate effect on both job availability and salary.
4) As relevant to HN, K-12 schools are making coding increasingly mandatory, so that skillset is less and less exclusive.
There's an open question about how much of this is a cycle. The Communist Manifesto was published in 1850, and has a nice discussion about how highly-skilled, respected artisan jobs were wiped out by industrialization, where a master blacksmith and unskilled labor had the same salary as factory labourers, which is worth a reread, and seeing which parts turned out wrong, and which parts turned out right.
A lot of what we might be seeing might be macroeconomic trends, rather than a cyclical.
Are you seeing this in the wild? I see a few people using it here an there to be a little more efficient, but I haven't seen anything better than a marginal improvement yet.
I'd be curious to hear how people are using this stuff for meaningful work.
[1]https://www.thomsonreuters.com/en/artificial-intelligence.ht...
I use it for coding all the time. All day everyday. I don't know about 'fewer people' claim though
It definitely makes me more productive in high-end, by more than what I'd call marginal, but not by enough that it's risking jobs.
However, the article, and my comment, were more general than coding. The impact is different industry-by-industry, from zero to very high.
More than likely it was the rose tinted glasses he wore.
It just means the remaining engineers had to work harder. AI helps you generate boilerplate, that you still have to check because it will fuck up. It doesn't do entire projects for you.
I didn't pick up that the people on that team were particularly impressive. I would guess someone still had to check their work because they would fuck it up. I'm not sure how AI is worse.
1. Google (keeping in mind early Google was much more selective than today
2. Elite schools / HN
3. Typical tech
When I worked for a non-tech, I learned that there are about a dozen rungs below that. The ladder runs very deep. I was surprised at just how incompetent people were if you stepped into almost any company outside of tech.
I am guessing there is no way I could hire you (a random HN'er) to make custom software for some waste management firm, but they need someone to do it. It's not particular complex or interesting, but it's stuff which needs to get done. That's actually a huge part of the industry. On a prior thread, someone asked how IBM makes their money. The pieces I've seen are based on contracting out this sort of boring but very necessary work from a place that's completely incompetent in tech (random big non-tech firm) to one which is only moderately incompetent (IBM).
The majority of those people have neither the ability nor the inclination to build software. The idea that everyone will write code in the future is a fantasy pushed by the business bobbleheads who want any excuse to pay less for engineering talent.
Today, that path is open to everyone. The idea isn't that everyone will write code. The idea is that everyone will have the opportunity to learn, and if 1% run with it, that's 3 million programmers in the US, and 80 million worldwide.
I have no idea what that percentage is (5%? 0.1%), but whatever it is, it's a skill that's growing much less exclusive.
I'm not buying this. Literacy is declining in the US - and it's not like the older people have forgotten how to read, so we can safely conclude that fewer and fewer of the younger people can read and you think somehow these illiterate people are learning to code?
One could argue, and many have, that a bachelor's degree in computer science doesn't prepare one for being a professional developer. A secondary education level developer? On average (discarding the very few exceptional students)? Nope. They're not in your competition pool.
https://apnews.com/article/phonics-science-reading-c715dea43...
https://slate.com/human-interest/2023/10/reading-phonics-lit...
https://www.apmreports.org/episode/2019/08/22/whats-wrong-ho...
I have zero ability to tell a school that some study they're relying on is faked if the author is at an elite school. This specific rot is in universities and not local to school districts.
The most common scientific method used for most high-impact education research:
1. Pick a story people (teachers, parents, journalists, fellow academics) want to hear.
2. Bake data to support it.
3. Publish it.
4. See people pick up on it.
5. Tenure.
Other domains of research have related, but different, scientific methods. There is rot local to school districts, but it's not the same rot.
This stuff hurts kids. I'm increasingly moving towards the strict liability side of the debate. If a university publishes results like this, and my district wastes a million dollars on pseudoscience in reliance on that research, that ought to come back as damages. Ditto for people harmed by bad medical research and other domains.
All of this would happen just as surely if the universities were staffed by conservatives as by liberals. It's how we've set up incentives.
My 14 year-old came to me to help him debug a Python script that he had to write for his Technology class. I didn't even know that they were teaching programming until then. And this is for a rural school district in farming country. So yeah, it's happening.
As far as whether or not the skillset is less exclusive, I don't think that we have much to worry about. It doesn't seem that people are having less trouble learning to program today than they did 30 years ago when I was tutoring in CS. And this is with far more resources available now.
Arguendo, if the US has a gap there, that would only strengthen the macroeconomic trend I described.
Why should I pay an illiterate American 150k, when a Jamaican student can do the same better for a fraction of the cost? And yes, many of the Caribbean education systems are increasingly impressive.
this is a BI article. reset your expectations
yes it was def a thing. Lots of people i know had ( and still have) 2 or 3 jobs at the same time.
The same articles were written each time. Once growth starts up again things will be the same.
With media so gamified and managed, it's hard to know what's true or what big spenders want us to think is true about the economy.
The one thing I would say is different between now and the early 90's is we have a lot more retirees, and retirees as a group are much more exposed to inflation. Though its largely the Boomers who are that group of retirees and they've been through periods of much higher inflation and so if they thought Quantitative Easing was going to last forever then they were just being foolish.
They're also the generation rapidly losing power and so they like to complain so they can still feel like they have some power. If anything, because of the Boomers I'd say the media is being overly pessimistic about our current situation. It's still better than where we were in the early 90's.
The fed even.reported theyd want unemployment to rise to help.
The inflation itself is driven by corporate price gouging, so its mostly rich people fighting eachother and seeing which part.of the proletariat is easiest.to downsize.
However, I also don't understand macroeconomics enough to evaluate whether your statement is true.
It's as if I doubt myself more than I doubt you.
We also have AI based efficiency boosting tools and greater offshoring to contend with
That just goes to show perspective is everything!
There's nothing scientific about the Fed or other central banks targeting for 2% versus, say 5% or 10%. A finance minister in New Zealand suggested the 2% target in the 80's: https://www.cfr.org/blog/history-and-future-federal-reserves.... Other central banks copied it and now it's set in stone.
What matters is that everyone has an agreed upon expectation for what inflation will be. This is because the economy is based on expectations. Things only get bad when inflation moves suddenly higher or lower and it catches everyone off guard. All their annual projections are wrong and they have to make adjustments.
So the absolute inflation number doesn't really matter, so long as it is steady and predictable. It is true though, that psychologically the higher number may have a negative effect on people.
The 70's and 80's did have big inflation numbers, but what really causes the pain is the fluctuations. And in hindsight it's clear that the actual inflation number wasn't the problem, it was that no one knew what next years inflation rate would be.
The point is: on an absolute basis, today's inflation rates are lower. But that's not what we should be paying so much attention to. It's the rate of change of inflation that matters.
Moving from 2% to 8%, is effectively a 4x adjustment in expectations. Whereas going from 10% to 20% is just a 2x adjustment.
But I temper that with the understanding that people only really know what they've experienced. If today is the worst you've seen, then to you it's legitimately horrific. That's real.
That the situation today is downright luxurious compared to actual hard times isn't really emotionally relevant if you weren't around for the hard times.
I think that AI is parroted around but there is little to no substance to the claims.
Some activities can and will be impacted by generative AI, but if you are not a copywriter or a graphics artist then your work is not significantly impacted by the current blend of commercial AI services.
Software engineering is a particularly popular target for these baseless claims, as if the critical path of each coding task is writing code. It isn't. In fact, I'm yet to see a single concrete claim on how generative AI impacts productivity beyond filling in boilerplate code and acting as a glorified template engine, and this assumes AI produces code that work which more often than not fails to do.
More importantly, implementing business services to automate business processes has a more direct impact on eliminating jobs. Things like opening an online store instead a brick and mortar shop eliminates the need for a bunch of positions. Nothing AI does comes even close to have that impact. But we still have brick and mortar shops.
It's also important to note that some software engineering positions didn't even existed a couple of decades ago, and they are high paying jobs. A couple of decades ago mobile development wasn't even a field, and right now it's absolutely clear that iOS developer and Android developer are entirely separate specializations.
Its like an advanced mythical man month fallacy.
They’ve been saying that for at least 20 years.
HIRE - not really.