41% of Employers Worldwide Say They'll Reduce Staff by 2030 Due to AI
130 comments
·January 10, 2025keyle
anigbrowl
The World Economic Forum released its bi-annual survey on what employers worldwide expect their businesses to look like in the future and much of the attention is on generative AI. And while a majority (77%) expect to help train their existing staff to work with AI, 41% say they expect to reduce the number of staff they employ as AI automates more tasks on the job.
The survey includes 1,000 employers worldwide, which covers more than 14 million workers in 22 different industry clusters, according to the new report. One of the big problems that emerges from the survey is that employers believe many of their workers don’t have the skills needed to do their jobs as technology evolves.
While I share your skepticism about the viability or specificity of predictions, large think tanks do this sort of research all the time and provide at least some indication of where corporate weathervanes are pointing and trends in business management, be they fundamental or faddish.
This information was all laid out in the first 2 paragraphs. You could have critiqued their track record or methods rather than asserting such questions are meaningless. Why do you think it's so outlandish to survey ~1000 corporations, especially if that's your raison d'etre and you do it twice a year, every year?
41% of the business today won't exist by 2028.
You think 40% of businesses are going to disappear in the next 3 years? That seems unlikely. I assume you're being sarcastic.
makeitdouble
> You think 40% of businesses are going to disappear in the next 3 years? That seems unlikely. I assume you're being sarcastic.
The average life span of a business is short. Even for S&P500 companies it's only 15 years[0], if we take in all the startup and small/individual businesses, parent's 3 year bet is aggressive, but not completely unreasonable.
[0] https://www.ey.com/en_gl/insights/consulting/how-businesses-...
hn_throwaway_99
I went to the referenced article about the 15 years (https://eprints.kingston.ac.uk/id/eprint/42134/), but I'm very skeptical as to what that actually means without more info (I just read the abstract, don't have access to the full article). I'm sure a huge part of that is there is much more M&A than there was 80 years ago. I also don't even understand how the quote in the abstract can make sense, because it states "The average lifespan of a US S&P 500 company has fallen by 80% in the last 80 years", but that article was written in 2018 and the S&P 500 was only created in 1957. There were no "S&P 500 companies" 80 years ago.
I'm also guessing the WEF's survey list skews heavily, if not entirely, to large, established employers.
subarctic
Hmm I guess it's a good point that a lot of those 1000 companies are not startups and will be around in 3 years
benterix
> You think 40% of businesses are going to disappear in the next 3 years? That seems unlikely. I assume you're being sarcastic.
Not necessarily. When you look at the lifecycle of a company, the numbers are quite harsh[0]:
> 23.2% of private sector businesses in the U.S. fail within the first year. After five years, 48.0% have faltered. After 10 years, 65.3% of businesses have closed.
So the parent's figure is exaggerated but not by a large margin.
[0] https://www.lendingtree.com/business/small/failure-rate/
joshdavham
> 41% of incendiary articles are purely written to gather clicks.
Agreed. This is just fear-baiting.
benreesman
The year is 2029. The LinkedIn/Meta merger has resulted in a 100% rate of LLM recruiter agents and applicant agents sending emails with keywords set not to display via the “color” CSS consuming all Internet bandwidth.
The survivors envy the dead: what few resources remain are devoted to an arms race between Jeff Bezos and Elon Musk on building the tallest and most efficient phallus to nowhere.
The only way for anyone, wherever they were born, to be part of the rocket class is to apply for an H1B visa. The lowest rate of accepted applications is in East Palo Alto.
Surfer Zuck hasn’t been seen in decades.
NBJack
> Surfer Zuck hasn’t been seen in decades.
But a suspicious and perpetually young version of him, in the exact same office background and a weirdly off voice, regularly appears in video messages before every earnings call. He speaks of great opportunity and growth for the Meta Quest X, and assures the investors that the scant few humans remaining are perfectly targeted for the next wave of MetaM-8s to find them, corner them, then display enough ads on their chest screen until the victims give in and willingly max out their Bezos Card as each bot meticulously records their frantic taps through the checkout process.
In the distance, the next set of competing space rockets roar off into an orange sky choked with delivery drones.
benterix
I just wonder - don't these current billionaries see how their actions literally shape the world to be a worse place to live in? I believe Gates, in spite all the hate towards him, actually has this kind of conscience and good intention of wondering, "How could I actually help the world in an optimal way?" and choosing his way (saving lives through vaccines; whether these are happy lives or not it's another question).
But seeing that your businesses make the world worse and not doing anything seems completely short sighted. These folks have children, too - what kind of world they want to leave them?
benreesman
Your satire is better than mine! This is some dark shit and I’ll thank you for giving me a little humor to go with the Black Mirror screenplay happening in real time.
null
throwaway290
You don't need to interview all employers. It's called sampling.
forgetfreeman
I'm not sure what's sillier, banking on 3rd party bullshit generators to handle work product or ignoring the likelihood that executive tier depravity extends to these kinds of depths when the prospect of cutting labor costs is presented.
imgabe
I don’t know, we manage all right with the 1st party bullshit generators.
forgetfreeman
You want to review the hiring slump in software development or perhaps the massive layoffs in middle management nationwide? We could also go take a look at what generative AI is doing to creatives income streams. Careful with that "we", it may not cover as much as you think. Edit: it occurred to me that I may not fully understand what your comment was meant to convey Are you saying labor is doing fine or work product created by generative AI is of acceptable quality?
ijidak
Yeah. Even if generative AI assisted workers are not as productive as their managers imagine, craven managers will still cut or refuse to hire staff and tell their employees to just use more AI to get work done.
I think many of us on Hacker News underestimate the depths to which middle tier managers will go to appease higher ups.
Even in the case where AI productivity is total fiction, that doesn't mean it won't have a huge impact on the workload given to a single worker.
All the way back to Pharaoh in the Bible when he is quoted as saying, 'They are relaxing! Let them make bricks without straw!'
Hiring reduction doesn't require an entirely real productivity boost.
It only requires the MBA and Management Consulting led perception that hiring fewer humans is now the smart thing to do!
Over2Chars
Totes bro.
safety1st
No you've got your head in the sand. 41% of firms having plans to reduce staff with AI as part of the reason - that figure 100% tracks for me from conversations with upper management and C-suite at client sites.
Whether they will all get around to it or not I don't know but they enthusiasm they have for it is tremendous. I have no plans to cut staff within my own company but I also see very little headcount growth as being necessary in the next few years.
The mechanisms here are simple and plausible,
1) The task now is to train people on how to use generative AI to get more done - if you think this trend doesn't exist you really do have your head in the sand, this is happening, now, today, some roles and people do better with it than others but it is happening.
2) The upcoming task will be to let go of the least efficient performers, who have not increased their productivity by incorporating generative AI into their workflow. It will work, a part of the staff will complete X% more tasks, this will translate into X% of jobs being made redundant. When this large cost cutting opportunity exists and execs are basically chomping at the bit to make it happen, it will be super hard to achieve a positive hiring climate.
Now in eras past this might have been blunted by a number of factors, but all these factors are on their last legs:
A) Quality is a real issue when you double someone's workload and tell them to get it done using GenAI. However the global economy is increasingly consolidated under large firms that possess a lot of market power - they are too big and entrenched to be unseated, so they don't really have to deliver quality, they know this, they can just be shitty monopolies, cut costs, and feast anyway because the barriers to entry for competition are too high, consumer switching costs are too high etc. Quality is just going to go down and that is just going to be the new world. Big tech understands this opportunity better than anyone else, every single one of them is a criminal enterprise, and due to weak government enforcement every single one of them is still posting amazing profits
B) Demographics are now working against us instead of for us -- turns out everyone has decided not to have kids, which means an end to population growth, consumption growth, ergo hiring growth. The nature of compounding numbers is that it's a slow burn until suddenly it isn't and if you look at a 10+ year horizon absolutely firms are thinking about this and they know there is no reversal to this trend, there is no gearing up for a future where everything suddenly doubles, long term the goal is to prepare for the shuttering of a lot of operations, adapting to markets being smaller, and this of course means managing a secular downward trend for your headcount, not upward.
It is a very precarious time we are entering, we either need to start busting trusts and making babies again, or the way of life that we knew is going to end. AI if anything is what is going to help us limp along into the twilight without a complete collapse of the value chain. Either way there is going to be a pretty limited appetite for new hiring in the foreseeable future due to a confluence of trends.
bwfan123
Speak bro. gen ai is the wet dream of mgmt types tired of ceding control to the pesky workers, not to mention highly paid entitled devs. How dare they poke fun at us suits. Now is the time for the managers of the world to wrest control back from labor - mgmt unites in strength with robots and ai. Let’s rule
rajnathani
There will be more companies to absorb the labor force changes, think small boutique-size (read: nimbler too) companies; the greater number of companies will create further competition in the market and further raise standard-of-living for all besides initially raising standard-of-living due to making at least non-blue collar work easier aided by AI.
nemoniac
I wonder what David Graeber would have made of this. Would he have said that the staff to be reduced were doing "bullshit jobs" anyway? Or is there even a significant overlap between the reduced staff and the bullshit jobs?
walterbell
WEF 2018, https://www.weforum.org/publications/the-future-of-jobs-repo...
> by 2022, 62% of organization’s information and data processing and information search and transmission tasks will be performed by machines compared to 46% today. Even those work tasks that have thus far remained overwhelmingly human—communicating and interacting (23%); coordinating, developing, managing and advising (20%); as well as reasoning and decision-making (18%)—will begin to be automated (30%, 29%, and 27% respectively).
WEF 2020, https://www.weforum.org/publications/the-future-of-jobs-repo...
> we estimate that by 2025, 85 million jobs may be displaced by a shift in the division of labour between humans and machines, while 97 million new roles may emerge that are more adapted to the new division of labour between humans, machines and algorithms.
Asymptotic Intelligence is almost here!
neom
I run an accelerator out of a huge law firm in Canada. I'm not much involved with the firm outside of using the startup lawyers to get the work done for my founders, but lunch room chatting happens. It's been interesting to hear the discussion: inbound is actually increasing due to AI, AI is both giving bad advice (or more often, good advice badly) and drafting nonsense, or missing so much case law it build weak cases etc. I don't think we will hire less, I don't think we'll downsize, I think we'll just take on work we said no to perviously because it wasn't interesting or whatever, and I think client satisfaction will increase. (also AI tools for huge firm lawyers doing big client files are comically bad, one large legacy player recently turned on the AI feature and it kept telling the lawyers to consult a lawyer)
fl4tul4
The tools will get better in time, with human assistance.
They will learn how to get better at providing results.
Nowadays, there's confusion, however, there is room for improvements.
neom
I agree. One funny thing I observed happen was it seemed maybe last year we'd slow associate hiring, but they took a pretty measured wait and see approach, and they set up an AI team to basically very quickly end to end test every tool they could. Coming out of the year two things were clear: AI was still not great for the type of work they do (it's "enterprise law", think $1k/hr lawyers), but the EAs and Paralegals etc will be able to get way more done in a day (but ChatGPT or whatever is fine for their work) and 2025 was going to be a pretty busy year. I'm starting to think the firm will only grow as a result of AI.
a year ago I thought they would all be replaced, but as I spend time with them and see their work, I realized I knew nothing about big law, half the job is therapist and talking people out of being emotional and brash before you can even get to the lawyering. Lots of tacit knowledge too, this regulator has this guy who always rejects the application if the format isn't this way, or this judge never accepts that argument, or yeah I've negotiated against him 30 times, he hates yellow so wear yellow.
The legal system is a very fundamental human thing after all.
blharr
"They'll just get better" is a huge non sequiter.
AznHisoka
Have they used GC.AI? Heard some buzz about it but kinda skeptical myself of these tools
neom
No, but it doesn't look very good to me at all, tbh.
The best thing I've seen is https://www.spellbook.legal/ but our firm is too big for them right now.
throwaway243123
Is that Relativity?
BuyMyBitcoins
How many other people here had some “dazzling” in-house AI tool quietly fade away because it simply wasn’t useful? Where I work the AI hype faded and executive leadership went back to the tried and true “just outsource this to overseas contractors” method of trying to reduce staffing costs.
forgotoldacc
The company I work at went all in on ChatGPT subscriptions to help programmers. They tried to get us to use it for debugging and everything.
People went from using it nonstop to maybe using it once a week. All subscriptions were canceled and it's basically relegated to toy status now.
bhhaskin
It's just not that good beyond basic tasks.
TuringNYC
>> It's just not that good beyond basic tasks.
A ton of dev work is basic tasks, esp writing unit tests.
quacker
Agreed. I finally tried Github Copilot for a bit at work. It felt like babysitting a bad developer. I have no confidence in the correctness of the code it produces, so everything requires careful review. Sometimes it spits out very broken code on non trivial tasks.
I’m still figuring out if/how it can make me more productive. First impression is it’s more a drain on productivity than not.
RSHEPP
I would guess most people in leadership have no actual clue how they would replace the human in the seat with AI. They are riding the hype that the companies who provide the tools are putting out, but when it comes time to execute they don't have a plan. The steps between generated code to running in production is completely being ignored.
BuyMyBitcoins
Ironically, LLM’s seem most suited towards replacing many of the functions middle managers perform. But those folks would never allow their own role to be outsourced or automated away.
SpicyLemonZest
That's not really the dynamic. If director A used to oversee 3 teams, and she discovers that with AI she can now oversee 6 or 9 teams just as efficiently, she's not going to avoid doing it out solidarity with manager B who's looking for space to move up the ladder.
tbrownaw
> have no actual clue how they would replace the human in the seat with AI
That's not how it works. It's not "your job is done by an AI now", it's "you have an AI to help with these specific tasks, so the team is now half the size and you each get twice as much work".
anothernewdude
I'm hoping the work in fixing AI hype projects is just as lucrative as fixing overseas contracting projects has been.
iambateman
No one in the world knows within an order of magnitude what the effects of AI will be on hiring in 2030.
To pretend otherwise is foolish.
o11c
I can guarantee that AI will be used as an excuse to lower wages.
somenameforme
Nominal wages are basically never lowered, but real wages can be easily lowered by giving out annual raises that are below inflation. Since 2020 the CPI is up more than 18%. [1] So if you're not earning more than 18% than you were then, you're receiving a lower real wage.
This is one of the many reasons inflation is such a horrible system. It makes it really easy to invisibly abuse labor. People are trained not to spit on "gifts", but if you received a 2% raise each of the past 4 years, your real wages would now be down by nearly 10%.
gruez
>but real wages can be easily lowered by giving out annual raises that are below inflation. Since 2020 the CPI is up more than 18%. [1] So if you're not earning more than 18% than you were then, you're receiving a lower real wage.
If you looked around on the same st louis fed site, you'll see that real (ie. inflation adjusted) wages has not gone down.
SoftTalker
Inflation is a way to raise taxes without actually raising taxes. Gov't just creates more money. Gov't pays its bills, people's spending power declines, same as if they were taxed.
vkou
That is a reason inflation sucks, but there are many reasons for why alternatives suck even more. Deflation strangles the economy, while basing your monetary supply off the amount of metal you dig out of the ground results in a tag-team of uncontrollable deflation and inflation taking turns.
Unless you're planning on doing away with money entirely, it's the best of all possible worlds.
jasdi
I can guarantee the opposite will happen.
Just have to read some history. A great example is the Gilded Age and what followed after the peak.
Why didn't the grand masters like Rockefeller, Carnegie, JP Morgan control where the story went? How did the system prop up an Ida Tarbell, Upton Sinclair or Teddy Roosevelt? Similarly when interesting characters like that don't emerge eg China or Russia the story goes in very different directions for the people who run Companies. Companies and what they are upto can look very mesmerizing but they really aren't. The British East India Company had more resources and land than the British Govt yet where are they today?
Firms (or any large grouping) and what ever they believe, are a part of much larger systems. They are not The System but part of the Ecosystem.
People inside and outside the firm forget that all the time. History is full of examples of what happens after that. Another great example is why did Central Banks emerge all around the world? What do you think was happening before when the power to issue Currency was in the hands of individual Banks or Kings?
If AI does replace jobs massively we will get UBI type systems just like we got Central Banks whose goal is Ecosystem Stability.
chii
> If AI does replace jobs massively we will get UBI type systems
wishful thinking. A more viable and realistic alternative is to let those who are useless post-ai to die.
UBI only works if the production of resources is "free", and i do not foresee such a future until the entire globe is united under one gov't, and we become a type I civilization capable of harvesting all energy on the planet.
RIMR
To be fair, businesses will use any excuse they can find to lower wages.
fuzztester
To be unfair, you mean.
lotsofpulp
I also shop around for lower prices. I think many people do, at both work and at home.
WhyOhWhyQ
Without looking at the actual report, it is a pretty meaningless statistic. Maybe it's more significant by job category, but if hiring can either go up or down, and you say about half think it's going down, you haven't suggested a non-trivial conclusion.
llm_trw
The more I work with Ai the more I feel like a programmer in 1950 dreaming about the www.
It will get here eventually, but barely within my lifetime and only if hardware keeps advancing the way its been for decades.
otabdeveloper4
I'm not sure hardware is the bottleneck here.
I've had "AI" regurgitate reddit and 4chan threads verbatim to me. What happens when this mine of free data runs out? (Basically it already has, in no small part due to the effects of "AI" itself.) Will we need to actually pay creators for content? Seems economically unfeasible, the cost of content creator + AI is more than just hiring the content creator directly.
llm_trw
Hardware is very much the limit.
The more compute you have to train your models the better the models become.
I've been working on a family of (new) deep learning models which outperform everything else that's been published, at between x1e3 to 1e9 the compute on the same data. If I had the GPT4 training cluster I'd be able to run it on cifar10 without having to make sacrifices, yes 32x32 color images solved with a trillion parameters.
Juliate
Also no one knows what the main events/news drivers will be in 5 years: climate, water, energy, war, peace? AI and economics, employment are way too little of concern at this point.
YetAnotherNick
Everyone in the world knows that with current AI, you need probably a quarter of content writers to deliver the same output.
To pretend that there is a possibility that there would be no effects on hiring is foolish. Not saying it would increase unemployment, just shift the work.
tonyedgecombe
Or you write four times as much content to keep up with your competitors who are doing the same.
roncesvalles
Mostly blogspam.
AI is death for people whose primary skillset was being proficient in modern western hemisphere English grammar, because now you can input a few messy sentences and have it output grammatically perfect, if not stylistically perfect, English.
YetAnotherNick
Blogspam and clickbaity content is not niche. What's your point?
Pxtl
If you could replace 3 of your 4 writers with gen AI then they weren't writing content they were writing filler.
AIs write noise, not signal.
brigandish
It’s irrelevant whether they were writing filler or not, they had jobs, and someone was paying them, which means their produce had some value, both to the publisher and the consumer.
Otherwise, they’d simply be cut at some point, not replaced.
toss1
YUP
They also predicted that the personal computer on everyone's desk, and every major software category, etc. would reduce employment. It just made more work
darth_avocado
Those employers are going to have to find solutions for keeping themselves around. There’s always a tipping point where enough population is unemployed/impoverished that society starts to collapse. Of enough people are unemployed, who is going to buy all the stuff? Who is going to pay the taxes? Who is going to make sure the employed people can ensure their own safety?
dare944
Maybe they'll hire some people back as bodyguards for the CEO?
darth_avocado
CEOs won’t be the only people who would need bodyguards. All you have to do to understand what happens in these situations is to look at failed states across the world.
itake
There is literally an infinite amount of work for tech to do. I think there is a deeper issues going on.
bb88
I think it's called greed.
from-nibly
Greed is a lazy excuse. Why not just say badness or villiany?
null
Juliate
That is putting intent where it is neither needed nor in capacity. Greed is enough.
BobbyTables2
Wouldn’t mind seeing 41% of executives unemployed…
matthest
Imo the best case scenario for AI is that it allows almost anyone to become an executive. AKA a small business boom.
RIMR
There is no group of people more fitting to be replaced by machines. Unfortunately, they have the means to ensure that they will be the last to be replaced.
Brace yourselves for technofeudalism...
cellis
I disagree. In fact they’ll be the FIRST to be replaced, as that is the richest target for replacement. Think about it. You wouldn’t build a robot (R+D costs) to do the job of someone cheap unless you can build many many robots and amortize the cost over millions.
Soldiers for instance are EXTREMELY expensive; so in modern theaters they’re being replaced with drones and missiles.
And executives are being replaced with automation in that there are less and less of them and the ones that remain have much higher leverage.
lyu07282
The managerial class has class consciousness, the working class doesn't. That's why they will be safe, they will all protect each other.
est31
Executives are paid handsomely because of the value they bring to the company. Part of an executive's job is to meet with executives of other companies and make deals with executives of other companies. Those deals work, but they require the human element. Executives don't do their jet set lifestyle just for fun, they do it because human connection creates the best deals.
There is humans all around executives, they are in meetings with humans all the time, both inside and outside of the company. They even have humans to interact with computers for them because they are so busy interacting with other humans.
Where do you see executives being replaced?
Maro
I think some jobs will definitely disappear / change due to LLMs. It's already happening, I can tell how I personally effectively elimitated 1-2 jobs around me in 2024 thanks to my $10/mo ChatGPT subscription.
I work at a large non-tech company. We have a lot of "Strategy people" and "Strategy teams". What these teams actually do is: create and submit to each other internal plans/KPIs/OKRs, quarterly updates, budgets, and so on. Strategy people are all ex-consultants (many are ex-McKinsey), they have have little domain knowledge.
Before ChatGPT, the way it would go is that they'd ask me something like, what is our annual plan for "AI", what is my quarterly update on "Data Platforms", etc. I'd give them some raw input, sometimes I have to do extra work to make it presentable, sometimes they do most of it. The final product is almost always some sort of PPT. Effectively there's 3 phases:
1. I write actual raw (semi-technical) content: usually about 5-10, maximum a page worth of condensed bullet points. I can do this very quickly, in 10-20-30 minutes, since I just have to throw down what I'm doing all day anyway (well, the teams I manage).
2. Convert those raw points into some sort of narrative (if it's a plan), or summary (if it's an update). What part is relevant for a non-technical exec? What part makes the org look good? etc.
3. Convert the narrative to a PTT.
Before ChatGPT, I would do 1 and some of 2, and the strategy people would do most of 2 and 3. After ChatGPT, I do 1, I use ChatGPT to do 2 (takes 20-60 minutes), and strategy people now only do 3. This means that in our org, ChatGPT has taken out about 1-2 jobs out of the strategy team (headcount was closed).
Once AI is good enough to make the PPTs (upload 2-3 older PPTs to pick up the template and style, plus upload the current narrative, to build new PPT), I will be able to make the whole deck in 1-2 hours, without any strategy people. (Not that I want this, because then I'd be on the hook for the whole process, emailing, etc.) Note that since this is a recurring internal exercise, actual quality doesn't have to be super high.
Perhaps once we reach this level of automation people will notice that the whole exercise is inefficient and look for better ways to manage the company.
Over2Chars
41% of employers won't be here by 2030 due to AI.
41% of middle managers will be replaced in 2030 due to AI.
41% of incendiary articles are purely written to gather clicks.
Think of the hilarity and irony of the title itself.
How could you, really, interview All of the employers Worldwide, and ask them an even relatively correct estimate of a 2030 figure, when they can't even predict what Q3 will look like accurately. "Oh but it's just trend gathering".
41% of the business today won't exist by 2028.
And reduce, reduce staff by what? 2%, 10%, 50%?
This is just so silly.