'Forget Coders. The Real AI Threat Is In the Back Office' (thestar.com.my) 79
Which jobs are most threatened by AI? "Programmers, software engineers and other tech industry employees," goes one common answer.
"But many economists are more concerned about a different, larger group of white-collar workers," reports the New York Times: customer service reps, bookkeepers, payroll clerks and HR specialists, "who fly under the radar but collectively account for tens of millions of jobs..." They are spread across the country and throughout the economy, working in every industry, in big cities and small towns, at major corporations and mom-and-pop businesses... These jobs typically offer a middle-class salary or a pathway to achieving one — much as manufacturing jobs did for men before decades of globalisation and automation wiped many of them away... For now, such an outcome is a fear, not a forecast. Despite high-profile layoffs in tech and finance, there is little firm evidence that AI has hurt the labour market as a whole.
Economists have become increasingly convinced that disruptions are likely, but they say it is too early to know where or how widespread they will be. They remain broadly sceptical of claims that the technology will lead to mass unemployment in the near future. Some AI industry leaders have walked back such predictions in recent weeks. But given the extraordinary pace at which companies are adopting AI — and at which the technology is improving — economists say policymakers need to consider the potential effects on the labour market. And they say they are concerned that the public debate has focused too much on software engineers and a relative handful of other high-status careers — lawyers, consultants, economists — rather than the workers who could be most vulnerable...
Economists at Northwestern University recently recalculated measures of AI exposure based on the makeup of the total workforce, not just the people using the technology. Administrative and front-line roles, such as customer service representatives, rose to the top of the list. "The most affected jobs are secretaries, are routine clerks," said Michelle Yin, one of the working paper's authors. "They're not computer scientists or data scientists at all."
The article also includes this counterpoint from an economist at the University of Illinois who has studied earlier waves of white-collar automation: that like other disruptive technologies, AI likely will also create new jobs. So the possibility exists AI will make workers more productive and allow them to earn more. "I would be cautious about just focusing on what are we losing as opposed to what are we going to gain on the other side."
"But many economists are more concerned about a different, larger group of white-collar workers," reports the New York Times: customer service reps, bookkeepers, payroll clerks and HR specialists, "who fly under the radar but collectively account for tens of millions of jobs..." They are spread across the country and throughout the economy, working in every industry, in big cities and small towns, at major corporations and mom-and-pop businesses... These jobs typically offer a middle-class salary or a pathway to achieving one — much as manufacturing jobs did for men before decades of globalisation and automation wiped many of them away... For now, such an outcome is a fear, not a forecast. Despite high-profile layoffs in tech and finance, there is little firm evidence that AI has hurt the labour market as a whole.
Economists have become increasingly convinced that disruptions are likely, but they say it is too early to know where or how widespread they will be. They remain broadly sceptical of claims that the technology will lead to mass unemployment in the near future. Some AI industry leaders have walked back such predictions in recent weeks. But given the extraordinary pace at which companies are adopting AI — and at which the technology is improving — economists say policymakers need to consider the potential effects on the labour market. And they say they are concerned that the public debate has focused too much on software engineers and a relative handful of other high-status careers — lawyers, consultants, economists — rather than the workers who could be most vulnerable...
Economists at Northwestern University recently recalculated measures of AI exposure based on the makeup of the total workforce, not just the people using the technology. Administrative and front-line roles, such as customer service representatives, rose to the top of the list. "The most affected jobs are secretaries, are routine clerks," said Michelle Yin, one of the working paper's authors. "They're not computer scientists or data scientists at all."
The article also includes this counterpoint from an economist at the University of Illinois who has studied earlier waves of white-collar automation: that like other disruptive technologies, AI likely will also create new jobs. So the possibility exists AI will make workers more productive and allow them to earn more. "I would be cautious about just focusing on what are we losing as opposed to what are we going to gain on the other side."
Lol (Score:5, Insightful)
Some statistics on who works those jobs (Score:5, Informative)
https://www.bls.gov/cps/cpsaat... [bls.gov]
Human resources managers - 314,000 workers - 79% women - 80% are white
Human resources workers - 897,000 workers - 78% women - 78% are white
Human resources assistants, except payroll and timekeeping 124,000 workers - 79% women - 73% are white
Bookkeeping, accounting, and auditing clerks - 1,254,000 workers - 82% women - 81% are white
Payroll and timekeeping clerks 166,000 workers -82% women - 81% are white
Customer service representatives 2,600, workers - 64% women - 71% are white
Re: (Score:2, Informative)
I have no what your racist angle is supposed to show here. Let me just add one data point:
US population: 332 million people, 75% are white.
Re: Some statistics on who works those jobs (Score:2)
Re: (Score:3, Informative)
You really do not? Fascinating. He mentioned race in a context where it had no purpose. That does it.
Re: (Score:3)
Well, after you jumped in, it has purpose.
As it clearly shows that more white women are in those areas than you have as percentage in the population.
Also: it is absolutely not racist to mention a race, for no purpose.
It is racist to be derogative about a race.
Also again: he simply quoted it from the original text ... so that was racist, too?
Re: (Score:1)
You are mistaken.
Re: (Score:2)
It is not racist to point out the majority group and the majority ethnicity.
This is the not the large industry of agitators who constantly comparing the ethnicity / gender percentages in only high-prestige / high-paid jobs to the same ethnicity /gender percentages in the general population - yet for only for a favored ethnicity / gender.
That comparison is to divide groups, ensure the agitators have jobs, and keep the victim hood narrative in place to help win elections.
Dividing us into angry groups fighting
He was right, it was irrelevant (Score:2)
No idea why you think the OP's post was "racist."
At best, it's completely irrelevant, but sounds like race-baiting to me. Sorry, we're not that dumb. That's like me saying "I'm NOT saying your wife is a fat drunk who smells like rancid lunchmeat"...wait...why are you offended? I was saying she's NOT those things!!!
Minor Refinement (Score:2)
Re: Some statistics on who works those jobs (Score:2)
If you point out a significant demographic that is perhaps useful so others can make assumptions.
As I assume you point out that Americans are 75% white so we can afford 71% unemployment of white peoples?
Re: (Score:2)
But this is both racist and sexist? How can /. allow this?
Re: (Score:2)
and it's your fault when the AI fucks up.
No, I can't ever believe that! ;p
And when they don't pay enough for AI it is your fault for not using it 'efficiently' enough.
Re: (Score:3)
If AI makes human workers more productive, it becomes more profitable to hire human workers and the demand for them increases.
If AI makes human workers unnecessary, it becomes less profitable to hire human workers and the demand for them decreases.
The balance is subtle and hard to predict. If you still need humans but they can do more, that leads to more jobs. If you no longer need the humans at all, that leads to fewer jobs. In the near term, we're probably going to see a lot of both cases. Some jobs w
Re: (Score:2)
Well, problematic it is in "niche industries", where only a few players dominate the market.
For example Zeiss or ASLM. Zeiss started to use AI pretty extensive since a few month.
Their mantra is: they can now find solutions in their data, which they could not find before. Which helps them to create products which they could not do before. And for the ideas about those products they need the workers they already have. Especially in Germany lay offs are unlikely. Will they hire more "specialists", for creating
Re: (Score:2)
Probably. And do that long enough and you find nobody wants to work for you. Nice way to kill an enterprise slowly.
Re: (Score:2)
Nobody wants to work for you anyway. They do it because otherwise their family doesn't eat.
Re: (Score:2)
Some people actually like to work.
Software Developers, everything in healthcare etc. are prime examples.
Artists, Athlets, Astronauts :P
Body workers like massage specialists, Yoga teachers, martial arts teachers.
Plenty of more I guess, rock climbing guides, safari guides etc.
Pilots of course.
Re: (Score:2)
Indeed. In fact, there are quite a few jobs you _cannot_ be good at unless you enjoy them. Basically all of STEM qualifies. Yes, I get that economics grads hate their jobs and think that is how it must be. May explain why they are near universally bad at it.
Re: (Score:2)
[...]economics grads[...] May explain why they are near universally bad at it.
Because economics usually has nothing to do with running the business they are employed in.
If they would love their job, they would pick up the trade and figure how their economics skills fit into it.
However they lean back and tell everyone if the oil price is increasing, everything gets more expensive.
Well ... not sure if I need an Master of BA or Master of economics to grasp that.
The next thing is: we can fire half the people, t
Re: (Score:2)
AI is today already scanning the receipts at my workspace and it sucks because it is a hit and miss if it scans the sums correctly or not and I always have to double-check the sums that were correctly scanned and it's an 80% risk that it's not properly done.
Re: (Score:2)
It is certainly not bookkeeping, that is a job that requires high accuracy, and is mostly automatable by traditional automation techniques.
Who are they? Penguins! (Score:2)
These unseen minions are penguins! And they fly under the radar because they don't need to fly at all.
Here's the article's actual answer to the question (not in the summary): "âoeThe most affected jobs are secretaries, are routine clerks,â said Michelle Yin, one of the working paperâ(TM)s authors. âoeTheyâ(TM)re not computer scientists or data scientists at all.â
Who are they? Secretaries (Score:5, Insightful)
Here's the article's actual answer to the question (not in the summary): "The most affected jobs are secretaries, are routine clerks," said Michelle Yin, one of the working paper's authors...
Secretarial jobs were already cataclysmically wiped out by the word processing/computer revolution. It's hard to remember anymore how ubiquitous secretaries were to businesses.
However, turns out that this previous revolution didn't reduce the workforce, because the number of IT support personnel required increased directly as the number of secretaries decreased.
AI customer service gets it wrong at times (Score:2)
AI customer service gets it wrong at times and then someone needs to fix it.
Re: (Score:2)
Hello, people get it w7 sometimes and other people have to figure it out.
Re: (Score:2)
These jobs aren't ALL going away (Score:3)
Really? (Score:2)
Which jobs are most threatened by AI? "Programmers, software engineers and other tech industry employees,"
I never assumed that at all. I always figured it would be the "customer service reps, bookkeepers, payroll clerks and HR specialists" mentioned in TFS who would lose their jobs first.
Realistically, only construction workers, skilled tradespeople, doctors, scientists, and some tech workers are unlikely to be replaced by AI in the near term. And as robotics and AI improve, even those "safe" fields may start to shed human workers in favour of automation. But of course, by that time there will be little to buil
Re: Really? (Score:2)
Re: (Score:2)
On the one hand, humans are still much more adaptive, creative, flexible than programming or machines.
On the other hand, we may yet become a Philip K. Dick story.
Re: (Score:2)
Re: (Score:2)
Yes ... and no.
Some tasks become easier to automate. Like coding.
But, it is still left to interpret what the user wants. AI can't do that ... yet. And perhaps never. This isn't because AI is deficient, it's because the user does not know what they want.
What will be replaced is a worker who inputs data, checks the input matches expected output and then clicks SEND. Very little to zero value added. And in every workplace, there are a lot of those types of people. Never, ever had an original thought in the wor
AI will remove all the clerks (Score:5, Insightful)
If your job is filling out forms or collating information to produce reports, if it's taking notes, if it's taking inventory, if it's managing schedules, if it's producing documentation...
All those jobs are going to fall to IT. Not entirely, but it'll be human oversight and an AI replacing a team of white collar workers.
At the same time, it'll be embodied in robots and unskilled manual labor jobs will evaporate (this is already happening).
Good luck adjusting when the disruption is broad, deep, and rapid throughout the economy and workers can't retrain as quickly as jobs are eliminated. This isn't the automobile, this is "cheap obedient slaves with almost no support cost for those who can afford the upfront price tag".
Re: (Score:2)
I'm not sure where you get the "almost no support cost". The article I read about the Chinese factory said that the robots were just about all being replaces (by newer models) after only a year.
Re: (Score:2)
Well at least my job down at the robot factory should be safe...
Re: (Score:2)
Clerking has been historically vulnerable, even before AI.
The BLS has separate categories for "accountant" and accounting clerk". After spreadsheets, the former continued the historic grown and the latter cratered.
The BLS split "software developer" and "programmer" over twenty years ago. The former continued the historic grown and the latter cratered.
Accounting oddly is resilient (Score:5, Interesting)
My old company tried workday to simplify time tracking and accounting. It was a miserable failure despite a huge committment and time investment. The bulk of accounting is significantly more shades of gray than black and white which makes it terrible for AI/automation. You can make it easier to automate by designing your systems and processes around automation, but that can force you to lose some of the benefits of the original system.
My guess is that HR will be much more gutted, as the whole point of HR today is to make everything a repeatable process and eliminate subjective bias. Customer service automation is miserable so far, but the decision tree/script does lend itself to automation, at least for first tier support (if you don't care about actually providing service to your customers).
Re: (Score:2)
Accounting is a more rational and regulated industry than many. It's ripe for AI maximization.
Re: (Score:2)
Pure accounting and auditing are, as is tax preparation. But most business/cost accounting is much more complex which makes it more difficult.
Re: (Score:2)
Do you mean 'creative'? As in using the rules in interesting ways...? Pure AI application. The prompt starts with, for instance, "how do...".
Re: (Score:3)
No it isn't. Accounting isn't going to be improved using a glorified autocomplete. I think you're confusing "AI" with "computerization in general", and you're also failing to learn the lesson of the last 30-40 years that shows that, actually, no you can't automate away all of accounting. You can reduce the number of people involved, sure, but someone has to understand the logic.
AI isn't about logic, it isn't about reason, therefore it cannot replace people whose jobs are 100% logic and reasoning. And it's n
Re: (Score:2)
I had some outside look into accounting as an IT/IT security auditor (side, job, now in the past). My impression was that these people need to make quite a lot of judgment calls and if they get it wrong, they may be in criminal territory very fast. They are also critical for fraud detection and deliver a lot of the data needed for decision making and long-term planning. Hence not glamorous, not flashy and definitely only a job for people that enjoy this type of work, but it is quite critical this work is do
Re:Accounting oddly is resilient (Score:5, Insightful)
No it's not? HR's only role is to protect the company from their employees.
HR is not your friend. Raise any concerns with HR and you will be flagged as problematic. They are not there to resolve conflicts. They're there to AVOID conflicts. And the best way to avoid conflict is not having "conflictive" people in the first place.
HR does not work in your interest. HR works in the interest of the company.
HR is there to document your mistakes so the company can fire you cleanly. That's HR's main function.
Re: (Score:2)
You must be American. There is a rest of the world, you know.
Re: (Score:2)
I'm Argentinian. And I'm speaking from experience.
Re: (Score:2)
Ya, this is just awfulizing repeating an online mantra. To be fair, it CAN be like that in a non-union, private company *especially*, if you're not a particularly valued worker. But if you work in the public sphere (for a city, county, state, federal government, or public school), it's much, much better because those orgs are MUCH more risk averse.
Re: (Score:2)
So flag misbehaviour that may be detrimental to the company with high likeliness to be found out, you're likely good. Flag a conflict with an internal VIP, you're not likely going to win. Flag something the
Re: (Score:2)
Ya, I came to say something like this. I remember when the whole concern about AI taking jobs was resulting in daily articles on Slashdot and people were talking about how even the AI of that time could replace accountants... which is so very evident that people don't actually understand the work of an accountant.
It's not just a person managing a simple spreadsheet and putting out a report once per quarter. Accountants are the translators between, law (federal, state, county, city), local policy, past pract
Economics is pipe dreams (Score:2)
Their own data reveals that hasn't happened for 40 years: Why would it change when there are fewer paying jobs available?
We will see (Score:2, Interesting)
The threat to coders is slowly collapsing and it is more a threat to pretty bad coders (whose work is of negative worth anyways) and we do not have enough good coders to do all the review needed. It also looks like LLM-coding done right is more expensive than competent human coding, and they are not yet charging for the "tokens" what they need to charge to become profitable.
The problem I see with the back-office is that while the work is conceptually simpler, errors are even more costly. Hence I think the b
Re: (Score:3)
We recently got access to Claude Enterprise and found how expensive it is. We were given $45 a month of budget. Everyone in the team blew through that in 2 days. And considering this is still being "subsidized" I honestly don't see what's the future for "AI Coding".
Re: (Score:2)
Exactly.
Re: (Score:3)
Probably because they do not set the "context" correctly. And the AI tool is analyzing the whole project for every simple question again and again.
However for intensive usage $10 or so per day is kind of normal, so you had a budget for roughly a week.
Re: (Score:2)
It would be a valid point if one or two devs got it wrong. But it happend for all of us (and we work in different projects).
Re: (Score:2)
If you have no formal training how to use AI, I actually would not wonder if you all did it wrong.
Anyway, I guess you figure what the problem is sooner or later.
Re: (Score:2)
The problem is that we were on the Team plan which is a flat rate plan.
You raise a valid point but it has no substance. Management confronted us and said "you are doing it wrong!" And I stopped them "no, we aren't. even if we are, we can try to economize on tokens. We can get a 100% improvement, 200%? 500%?
It's still order of magnitude LESS than what we were doing and it's costing us, in dollars, twice what we were paying before.
You see, the problem with the Enterprise plan is that, unlike Team or Pro, it d
Dirty Secret: new models are more expensive (Score:2)
We recently got access to Claude Enterprise and found how expensive it is. We were given $45 a month of budget. Everyone in the team blew through that in 2 days. And considering this is still being "subsidized" I honestly don't see what's the future for "AI Coding".
The whole AI revolution is predicated that today's models are clumsy, error-prone, and EXPENSIVE...just like the early computers, but advancements in technology will make them more reliable and cheaper and that in 10 years, doing the same work will be a fraction of today's cost with greater accuracy. The problem is accuracy is improving VERY slowly. Yeah, on benchmarks they cheat on, you can show whatever you like, but Claude still errors daily for me and has pretty consistently for the last 2 years. If
Re: (Score:2)
We recently got access to Claude Enterprise and found how expensive it is. We were given $45 a month of budget. Everyone in the team blew through that in 2 days. And considering this is still being "subsidized" I honestly don't see what's the future for "AI Coding".
So, $22.50 per day, or $113 per week. How much does one of your people cost, all-in, including benefits? It's unlikely that it's less than $100k per year, and very likely at least double or triple that, if not five times or more. At $200k/year for 50 weeks, that's $4k/week. At the current token price, the AI makes a $200k engineer 3% more productive, the company is breaking even. If it makes them 1.5X or 2X as productive it's a clear and unquestionable win, even with higher token prices.
As for me, th
Re: (Score:2)
So, let me see if I understand correctly:
Your leadership gave you access to a new tool, without any training on how to use it.
A budget to only spend $45 worth of "work" from it. Each. Per month.
Assuming you make about 90K in yearly salary that works out to roughly 1 hour of your time. So if you can use the tool, only spend $45 in tokens, and get MORE than one hours worth of saving from it (magically, without any training in it mind you), you are WAY ahead.
"And we blew through that in 2 days" - what did yo
Re: (Score:2)
They were expecting it would last for the entire month, like the Team plan was.
But it turns out the Pro and Team plans are subsidized and include far more tokens, that reset every week, than the "actual price" that they charge large companies.
Re: (Score:2)
That they could spend $45 a month and double your output? Wouldn't that be a treat!
They probably actually indeed believed that.
I worked a lot with AI, but I had free access, and have no clue how many tokens I used to waste.
However I learned very early how to instruct the LLM to focus on the context i want it to work on.
At the moment I only use Google Gemini sparsely when I want to research something, I am logged into it with my google mail account, I have no idea if a "normal user" has quota, and if/when I
Re: (Score:2)
AI Helpdesk Agents (Score:2)
Re: (Score:2)
I've found that it is great at tech support. But i'm using it for tech support of widespread consumer products, not niche enterprise products.
Re: (Score:2)
...By which I mean that using AI without bothering to contact the company's tech support at all. Of course the company's tech support is utter garbage whether you reach a human or not. In fact, they are stretching the meaning of "human."
CSRs (Score:2)
Let's talk about CSRs. ChatGPT is much better than those morons. Companies should not be concerned, they should e happy they can finally get rid of the morons.
You mean THE University of Illinois at Urbana? (Score:2)
It's a bit ironic that the study is published by the SAME university that built that crazy AI that killed the crew of that ship (IIRC Discovery) we sent to Jupiter back in 2001.
The Quintessential Irony (Score:1)