43
votes
LinkedIn executive says that the bottom rung of the career ladder is breaking
Link information
This data is scraped automatically and may be incorrect.
- Title
- Opinion | LinkedIn Executive: A.I. Is Coming for Entry-Level Jobs
- Published
- May 19 2025
- Word count
- 1013 words
I know it's just an OP-ED, but I feel that the author is really trying to cram a point to which reality does not yet quite fit. In particular, the evidence just isn't there that "AI" is in any way related to hiring slowdowns.
It's absolutely true that newgrad unemployment is high - it's at a 40 year high, apparently. But even if you take it as granted that "AI" will cause large changes in white collar employment patterns, there has not nearly been enough time. Meanwhile, there's an elephant in the room that the article briefly acknowledges, then just kinda... ignores again?
...
That's backwards, in order. It's mainly the weird economic malaise we're in from the erraticism of the administrations economic policy. It's a very weird state - the US economy is still fine, but it's also at the precipice of multiple cliffs. This means that newgrad employment takes a disproportional axe - because they are an investment, and companies are loath to invest in anything next to a cliff.
It's all too convenient to take that as evidence of "AI", when even if you believe the "AI" revolution will occur, it's still too early to show up in the numbers.
Yea and people cheering the firing of probationary employees in the name of "efficiency" is not helping.
I've been seeing people writing about this trend since long before GPT3 dropped, let alone before OpenAI threw a chat UI on it. Companies have been shrinking their non-senior positions for close to a decade, with no foresight that you can't have new senior-level engineers if nobody hires juniors.
This intensified after ZIRP ended, and that also started the layoff trend. Because even outside of startups that run on venture money, any large business involves large amounts of credit.
And we also had that tax change from Trump's first term kick in around 2022, which really screwed software related fields: companies can no longer deduct R&D costs as expenses (i.e. software engineer payroll) in the calendar year they accrued. They are now required to be amortized over five years, which effectively causes a liquidity crunch and incentivizes reducing R&D costs. A bill was recently introduced to repeal this change, but I haven't been following it closely.
Now we have a drunk orangutan making unilateral changes to economic policy and running a reverse pump-and-dump on the entire stock market.
LLMs are just another blockchain: a shiny and mostly useless thing for MBAs to fixate on, torch piles of money, and then forget about when the bubble bursts.
I wonder if ever-increasing shareholder expectations aren’t also a factor in all of this. With a few exceptions, increasingly for publicly traded companies to do well on the market they can’t just be in the green quarter to quarter but also need to command huge profit margins. The easiest way to pump those numbers of course is to do a round of layoffs, stop hiring high-investment employees (newbies), and squeeze the seniors harder.
I don’t think that’s really true. Is there a numerical metric that indicates that that is a trend?
Anecdata: For years, I've been seeing steadily increasing outsourcing of infrastructure management and coding tasks to other countries with relatively high numbers of English speakers and low labor costs. India and Eastern Europe, for the most part. Technical services companies like Accenture and Optum Health may be U.S.-based, but they have huge contract teams elsewhere, and don't do entry-level hiring in the U.S.
You can't talk about "AI" displacement without any mention of the impact of non-domestic, globalized tech labor. As the saying went after the COVID-19 lockdowns, if you can do your job remotely, there's no reason people anywhere in the world couldn't do it cheaper. [Reality: language barriers, wildly varying skill and quality standards, time zone issues, fraud...]
Additional anecdata: the brainrot generation has hit the workforce and they're less capable than any previous generation.
At the same time, the global workforce's English and technical skills are improving. Given these two trends, on the macro level, it's getting harder to justify paying domestic graduates top-tier salaries for second-tier work, when you can get second-tier work for second-tier salaries through contracting.
There is also this article claiming that entry level tech hiring is down 50 percent
@vord,
@redwall_hp,
@tonestones
A lot of the economic malaise we're in has been due to the US Fed Reserve's response to inflation/high employment/low unemployment/high wages.
Money (loans) was too cheap. The interest rate was at or near zero.
Many big businesses were leveraging those free loans to invest in riskier ventures. They could afford to take those bigger risks because the low interest rates meant that they could afford to have some of those ventures tank as long as enough had outsized returns.
Also, these loans helped prop up the other side of the tech economy: venture capital. Cheap money meant more high risk/absurd startups meant more failures but also more home runs—not necessarily as a matter of percentage but in total.
But it can feel as if capitalism needs its sacrifice in blood so that the rest may prosper. When too many prosper, demand goes up, driving up prices. And so the economy requires adjusting so that more people suffer. If you read quotes from the Fed Chair, in 2022 and 2023, he literally said this—just cloaked in slightly more clinical terms.
And, so, the Fed made money more expensive.
This is where I often fall back on the concept that "The purpose of a system is what it does." Capitalism works exactly as intended: it protects capital. The further you are from it, the more you bleed.
I don't think this is a good read on the what and why of raising interest rates. It's not a net negative - it's good to slow the economy down when it's growing too quickly. Inflation is not a good thing on its own.
The main issue is that when demand grows at a rate that's too fast, it outpaces the capacity for supply to grow. Unlike demand, supply is much more chained by physical realities - it takes time to make more things. When demand vastly outpaces supply, it eventually crashes, and that's worse - it causes lots of thrash.
And either way, the correction is going to happen - uncontrolled demand growth is a mirage that will eventually be chained by back supply.
It's like parking a car. When you're parking the car, you could apply the brakes and slowly reduce your speed (demand growth) to the parking spot (the maximum theoretical demand a current economy can sustain). Or you can just crash the car into the car in front of you. Both ends up at the same spot, but the latter is... damaging.
But is protecting capital necessarily bad?
There are historical precedents where the destruction of capital directly resulted in widespread economic and human misery, like land confiscation in Zimbabwe, nationalization in Venezuela, and the collectivization efforts in Communist China and Cambodia.
I'm not arguing that capitalism is perfect. But it does seem to me that it's important to not only protect capital but deploy it more effectively. The critical flaw that you've pointed here is not that capital is protected, but that ineffective deployment of capital isn't penalized enough, therefore the stewards of capital are incentivized to be reckless with it, like lenders were leading up to the 2008 crisis.
But to that point, we must remember that our prosperity is built on systems that de-risk entrepreneurial activity. Silicon Valley is so prosperous because it's relatively easy for entrepreneurs to find investors to fund an idea, and entrepreneurial failures are actually valued as experience. Elsewhere in the world, entrepreneurs who fail aren't given a second shot by investors; with such investors, why would those entrepreneurs ever dream big?
I'd be willing to believe that both could be potentially true... In a way. I think you're absolutely right that it's the weird and unstable economic conditions that are causing companies to slow down/stop hiring for some entry level positions.
But I wonder if the reason all these companies feel empowered to do so is because they've been sold on the lie that AI will be able to replace all their worker drones in the next few years and they think they just need to hold out a little longer and they'll be saving tons of money. Whenever I see anyone who has fallen for the AI hype machine, it's usually a high level executive at a company, and not people who actually would have to work with it.
Just speculation, but it fits pretty nicely in my perspective on how these executives make their decisions.
A few points I’d like to make.
(A) People are definitely using AI in their jobs. ChatGPT’s monthly average users do not come from nowhere. It’s possible the 20/80 heuristic still holds and the people using AI are the people who weren’t that productive pre-AI. I don’t use AI very much. It’s hard to know.
(B) There’s a lot of money to be made automating white-collar work. If it’s economically feasible (read: if equivalent compute is cheaper than a human salary), it will happen.
(C) Some tech companies are outsourcing their intro-level jobs to contracting firms, some of which are located overseas, and all of which pay substantially less. Some of those jobs exist in the U.S. and they are hiring some very bright engineers at pretty low wages. Capital will do what capital does.
I see a lot of skepticism here and maybe for tech jobs that's warranted. But for law in particular (my job), I expect this to be completely catastrophic.
You can't outsource legal work - differences in training and jurisdiction specific knowledge (not to mention tight regulation) prevents this. But AI tools are now at the point that the assistant can do everything a senior lawyer would ask a law student or junior lawyer to do for them, but much quicker and easier. I do not have an answer other than training juniors will become an explicit expense for firm survival. The higher order legal skills I employ are much harder for AI to mimic, but good lord do I feel for law students today.
It's not that far off from the state of tech; I see some similarities here. The differences would be that senior engineers aren't using AI to pawn off work that would have gone to junior engineers, rather AI means their own productivity goes up a little, reducing overall hiring needs. I know what I need to write but rather than write it, I can have AI spit out 90% of what I need and tweak it. It sounds like lawyers are explicitly replacing junior work though?
Do junior lawyers typically stay with the same firm for long enough with a firm to recuperate the costs of training them? In tech the stereotype is that engineers jump every 2-5 years and depending on the salaries, that can mean they're not profitable to take on.
In law, the typical breakdown is that students are a net loss but they'll stay at your firm to become a big money maker. The 3-8 years of experience is the sweet spot for junior associates who rack up billable time doing awful boring tasks like research and document review that make a ton of money. The juniors either graduate to partner and pawn off the shit work or leave to in house/government jobs or other firms.
What stops/discourages law firms from poaching lawyers that have just gotten past the training phase and to a phase where they're profitable?
Nothing except wages and working hours. Poaching associates is a very common game between firms. The typical advice given to students and juniors is to tough it out until you've had two years of experience as an associate, and you suddenly have a new world of opportunity open up for you. Who knows how long that will last now.
Archive link
Everything is noise except the end result. Which is. They need less legal restrictions from exploiting foreign workers from poor areas.
Its obvious how no other solution is ever discussed. Just the sky is falling..blah...blah "Exploit foreign workers...less laws"
Your "representatives" are all cowards or owned by corporations. No one is getting saved. This is how things fall apart.
@mycketforvirrad
can you please change the group to life, not life.men
Sure! I was a little confused, but thought you might have had a mysterious angle. ☺️
Nope. My mistake. Thanks.