Studies Suggest That Rather Than Killing Jobs, AI Could Revive The Middle Class

from the unexpected-optimism dept

We’ve certainly been talking a lot about the “AI Doomers” who insist that AI is all too likely to destroy humanity. However, even people who aren’t fully on board with the existential threat of AI do often say that, at the very least, it’s going to destroy jobs for most people, potentially creating huge problems. For years now, people have been arguing for universal basic income, in large part, because they think that automation and AI will take away everyone’s jobs. I mean, it was a core plank of Andrew Yang’s silly run for President.

And with all these new AI tools coming on the market, there are discussions about it again. In recent weeks Annie Lowery in the Atlantic and Andrew Kennedy in Newsweek, both argued that AI was going to wipe out millions of jobs and we needed UBI to protect against that future. Both articles were responding to a Goldman Sachs report predicting massive growth in GDP from AI, but also that 300 million jobs might be automated away.

Of course, both of those articles totally ignore that the very same Goldman Sachs report predicts that the automation of those 300 millions jobs might not lead to mass layoffs, but rather just a changing of how those jobs are done, saying that in many cases, it will complement jobs, rather than substitute for it.

And, a few early studies seem to support that possibility. Planet Money recently had an episode looking at these new studies that even suggest that these new AI tools could help revive the middle class.

But there is a glimmer of hope – in the form of an economic study. The study looked at the customer service department of a big software company, and it found that ChatGPT made workers much more productive. More interesting, most of those gains came from less skilled workers, while the more skilled workers showed only marginal improvement. Put in other words, AI narrowed the productivity gap between lower skilled workers and workers with more skills. This finding is very different from previous findings about the effect of technology on workers over the last four decades. A whole generation of economic research shows that computers have been a major force for increasing inequality. A force for a shrinking middle class.

David Autor is a professor at MIT, and he is widely regarded as one of the greatest labor economists in the world. He led a lot of that initial research about the computer era and the labor market. And he thinks this study, and another one like it, suggest that we could use AI to expand job opportunities, lower barriers to entry to a whole range of occupations, and reduce inequality.

It’s notable that Autor is saying this as well, since (as the article snippet notes) his earlier research has shown how earlier tech innovations actually increased inequality. His biggest papers are on how much tech has harmed inequality in the job market. But in the podcast, he notes that his own research suggests that AI may also be decreasing that inequality.

The key, again, is recognizing how AI works best as a tool to assist workers, rather than as a tool to replace them. The first study mentioned above is really interesting to read. Basically, they looked at how customer support agents were using AI, and specifically reviewing productivity and performance and how it changed via the staggered introduction of the tools to different support agents.

The report found productivity went up across the board — and also that both employees and customers were much happier, which seems like a good thing. But, the impact on lower skilled workers was much bigger, basically leveling them up to work nearly as well as higher skilled workers.

Instead of experienced and skilled workers benefiting mostly from AI technology, it’s the opposite. It’s the less experienced and less skilled workers who benefit the most. In this customer support center, AI improved the know-how and intelligence of those who were new at the job and those who were lower performers. It suggests that AI could benefit those who were left behind in the previous technological era.

“And that might be helpful in terms of closing some of the inequality that previous technologies actually helped amplify,” Brynjolfsson says. So one benefit of intelligence machines is — maybe — they will improve the know-how and smarts of low performers, thereby reducing inequality.

Of course, this is just one study of one company, using the technology as it is today. But lots of things could still change. That study may not prove to be generally applicable. The tech could change. There are lots of other things that could go wrong. And this isn’t to suggest that UBI might not still be a useful tool for helping make sure that everyone has a clear foundation on which to feel safe and able to live freely.

But, at the very least, it might call into question the “doom and gloom” predictions that this will somehow further hollow out the middle class and massively increase unemployment. The early evidence, at least, seems to suggest it might do the exact opposite.

Filed Under: , , , ,

Rate this comment as insightful
Rate this comment as funny
You have rated this comment as insightful
You have rated this comment as funny
Flag this comment as abusive/trolling/spam
You have flagged this comment
The first word has already been claimed
The last word has already been claimed
Insightful Lightbulb icon Funny Laughing icon Abusive/trolling/spam Flag icon Insightful badge Lightbulb icon Funny badge Laughing icon Comments icon

Comments on “Studies Suggest That Rather Than Killing Jobs, AI Could Revive The Middle Class”

Subscribe: RSS Leave a comment
74 Comments
This comment has been deemed insightful by the community.
gianpo (profile) says:

Misreading the data

I’ve worked in an industry for 30 years that has gotten more more automated and I can tell you from experience it has cost jobs because it takes less people to produce more product with less knowledge so they also get paid less for the job so 1. less people employed and 2. paid less for the work. That’s 100% what’s going to happen to all industries that AI/automation take over. Even your quotes state that less skilled workers can do the job with AI do you honestly think the owner class is going to pay high wages for a job an unskilled worker can do.

This comment has been deemed insightful by the community.
Anonymous Coward says:

Re:

Automation tends to create more so-called “low skill” jobs, which is just business speak for how you can hire people without a certain degree of university training to do the same job for a fraction of the wage and save money.

Andrew B Allen says:

Re: What about the small businesses that will be enabled by ChatGPT?

ChatGPT helps me learn new skills quickly.

ChatGPT helps me cover for the skills that I lack.

ChatGPT makes me look professional when I’m a beginner.

All of these help me if I want to start a small business.

I remember in the 1990’s when the meme said “On the Internet, nobody knows you’re a dog.” I remember everyone wanting to start a business, but they needed to learn a bunch of new skills first, like they had a 2-page list of college degrees they needed before they could open for business. The limitation then was learning skills.

Today, ChatGPT can make us all 10x employees.

So, as the big businesses lay off employees and everything gets cheaper, there is a tremendous potential for millions of us to go out on our own and start out own ventures.

Anonymous Coward says:

Re: Re:

wrong answer

The United States has a class system defined as the bell curve of income divided into three even sections.

Look at where the middle of the curve is, this is the middle class. This is where government gets funding. This is where people are disappearing.

If this keeps up, there will be no one to make your happy meals. Border patrol will be tasked with stopping them from returning home.

and then there are other class systems .. which were we talking about again?

This comment has been deemed insightful by the community.
Anonymous Coward says:

That’s odd, because yesterday I read a tweet about how a game company fired its art department and replaced the department with an AI-based art procedural generator.

It was Korea-based Rayark Games.

And yhere’s that article where they talked to Chinese artists and they say similar things.

You clearly have access to data that says the owner class has tried to automate art jobs, at least. And you know that AI is taking away jobs.

Maybe the current crop of AI could lead to a better future for all of us. I do hope you’re right.

But the evidence, for now, points otherwise.

Anonymous Coward says:

Re:

Updatrs and errata.

First off, Rayark is based in Taiwan, not Korea. Apologies for that.

Secondly, according to a recent press release put out by Rayark themselves, they claim to not have fired their art department, but have implied that they do use procedurally generated Art in their games, and will continue to “explore the AIGC space”in the future.

More as the drama unfolds.

This comment has been deemed insightful by the community.
JCL says:

This theory doesn’t quite pass the smell test.

Key quote from the article: “What this system did was it took people with just two months of experience and had them performing at the level of people with six months of experience.”

Accelerating a learning curve by four months is nice. But if six to twelve months is the whole breadth of the curve, you are talking about really, really low-skilled jobs — of the kind that will be replaced in their entirety.

As AI continues to improve, it will logically “eat” low-skilled knowledge worker jobs.

By that meaning, you won’t get an AI that replaces human workers entirely — but as time passes and the AI grows more powerful, you will need a smaller and smaller number of human workers, coupled with more and more powerful AI, to get the same amount of input.

What this looks like in practice is a call center that used to have 500 human workers handling the same level of output with 50-100 workers — and doing a better job than before. The fact that the low-skilled workers who keep their jobs have an accelerated learning curve doesn’t change the basic trajectory of removing the human component, not all at once, but in piecemeal increments as the AI gets more powerful.

Then, too, as the AI gets more powerful, it will require more human skill to maximize its use. Think of what it takes to, say, master a complicated video game versus as simple one; it takes more innate skill to be great at Starcraft or whatever than MarioKart.

As AI systems increase their capabilities in terms of a hybrid configuration — being used by human workers — those systems will become more like Starcraft and less like MarioKart as they become more complex.

That will open up more room for skill differentiation, in the sense that the most skilled human workers will again have an edge against their less-skilled counterparts — maybe a big one — because their ability to process the inherent complexity of the AI system, like being good at a complex video game, means they get multiples of output productivity.

Then, too, the real question is what happens to white collar knowledge work — the kind of jobs where the learning curve is not six-to-twelve months but three to five years, or even five to ten years, to get good. In those jobs, one could argue AI will have an even more profound impact — again not by replacing humans entirely, but dramatically increasing the output multiple of the most skilled (those with the most ability to leverage augmented AI tools).

In the end you still move toward a society where human skill — the ability to leverage an AI system in complex ways, coupled with meaningful domain expertise — becomes a premium that gets leveraged more and more to produce more and more output at the top end of the human performer stack, which means you still get a brutal amount of lopping off at the lower end of the stack.

Maybe we’ll get a lot more occupational therapists and senior living care workers or something.

mhajicek (profile) says:

Re:

As an example that’s already occurred, an 18 year old with a CNC and some software can easily out produce a whole shop full of experienced manual machinists. Even 5 axis stuff is becoming relatively easy now.

So that one worker might get paid more, while the rest are laid off. Or they might get paid less, since there are plenty of others lined up for their position.

Anonymous Coward says:

Re:

Maybe we’ll get a lot more occupational therapists and senior living care workers or something.

There’s certainly been a lot of focus on the latter where I am, with phrases like “aging gracefully” and “active senior living with dignity” being common in the narrative.

Meanwhile the younger generations struggle to meet a standard of living on par with their educational qualifications per their parents’ and society’s expectations and demands. Confidence and self-esteem are headed for the bottom, if they’re not already scraping it. And the focus is still on making the boomers as comfortable as possible.

Eventually even the youth who are being trained to be senior care will have to have senior care provided for them, but that’s something nobody wants to think about – all people can think of is how to “honor the pioneer generation”. Then they piss and moan and whine about why the youngsters aren’t having kids or meeting their expected life milestones.

Darkness Of Course (profile) says:

First, AI lies. AI does not grok Truth

So, bad news. We got world class stupids in America. World class.

What we can do is fact check AI. All the time. Explain to the thread/post/replies that the AI was wrong, and if fact, made this shit up. Cause the AIs were developed to have a good conversation.

Any intelligent women are very familiar with mansplaining; Well this is AI-splaining. Which is 100% compatible with lies, liars, and lying.

The best outcome is more people learn to fact check articles. “What? Nah, that can’t be right. Okay, duckduckgo says that doesn’t even exist.”

That is optimistic, I agree. But with bean counters, Yale/Harvard grads doing M&A to make bigger corps and increase their personal wealth as their focus. The GOP pushing for Xtian theocracy Fascism 2.0, that’s all I got.

This comment has been deemed insightful by the community.
Anonymous Coward says:

After the Ross Pruden shit the other week and now this article on some studies by Goldman Fucking Sachs and a report about how low-skill workers combined with AI can do a better job and lower barriers to entry before AI just automates that current level away, like JCL pointed out farther up…

I feel like any articles about AI & Jobs on Techdirt can be safely ignored.

John says:

The train is entering the station

It may make people feel good to imagine that new jobs will be created as a result of AI trends – it’s true. But the net effect in a decade will be that people will be made not just unemployed, but unemployable.
When I imagine a Venn diagram of what humans can do vs what AI/robotics can do, the area where humans rule is rapidly disappearing. One can see that with regard to knowledge work that almost any discrete task can be automated today, and jobs are just sets of tasks intelligently knitted together. With regard to physical tasks, we have a bit more time. That said, look at any large organization and I’ll bet that at least half of the labor budget goes to the white collar side of the house.
I think AI and robotics will have a beneficial effect after we figure out how to “grow the pie” and share the benefits. When most work is automated, the case for taxing capital becomes much stronger IMHO.

Anonymous Coward says:

AI will record and analyze everything it can find on the internet, input by users. That’s going to allow for some incredible analysis of human behavior, and fact-checking.

Search warrants will seem so primitive, so soon. What is done in the dark will be brought to the AI light.

AI was programmed on an internet full of lies, hatred, misinformation, and disinformation. Garbage in, garbage out.

PRIVATE AI will be the dominant force, as it already is. I’d imagine some of the more nefarious actors are way ahead of us in that area.

Good or bad, this is inevitable, and we’ll have to find a way to deal. Wealth won’t change, only who controls it.

Anonymous Coward says:

It has been noted that AI will just make up shit, may sound convincing but it is still incorrect. This is consistent with its design, it is to be expected and it will not stop.

That being given, who in their right mind would hire such an employee? What sort of job requires fluency in bullshitology?

No, AI is not going to take yer yob any more than those lazy illegals who are both weak and strong.

Anonymous Coward says:

Re:

Did you not read the article? AI allows one inexperienced worker to do the job of several more experienced workers. That means that fewer people will get paid less to do more work. That drives down wages and job security, which has economic consequences. AI doesn’t have to be better than people, it just has to be cheaper and not significantly worse.

jarocats (profile) says:

Say again?

“Well, ideally the working classes will be, let’s say… living on a farm upstate where they can run and play all day.”

Excuse me? Correct my understanding if it’s mistaken, but exactly what are you saying here? That the working classes are a bunch of “useless feeders” that should be put down like sick dogs, or what? Because that’s exactly what it sounds like.

Anonymous Coward says:

Re:

Always has fucking been.

Though personally, that’s the “ideal”. The not-rich get turned into slaves.

The reality is that most of us will eventually get forcibly drafted (kidnapped, even) into a war we never wanted, dying to an enemy no one wanted to shoot at, and we all will be doing this because there’s no way to keep even yourself fed at slave “wages”.

jarocats (profile) says:

The billionaires are gatekeepers to AI because they fear economic democratization.

Why, if you or I could live comfortably without fear of being replaced and ruined at any moment, Elmo’s bunker inside Twitter HQ would be worth less! And its owner would be less special! Somehow!

That’s the way it works with everything, folks, from money to civil rights: People don’t want you to have the same things they have because they think they’ll somehow be poorer because of it.

It’s bullshit, and it keeps all boats from rising, but that’s how people think.

nerdrage (profile) says:

makes sense

The most resistant-to-AI jobs are the creative and the political. True creativity is beyond the reach of AI and politics basically means “convincing people to do things my way.” So someone with a new logo design that isn’t just a knockoff of a lot of other logo designs is being creative and if they present it to a client and persuade the client to accept it, that’s politics. AI isn’t going to be able to handle that kind of stuff in the near future.

What AI could do is create a buttload of knockoff logos that the creatives can then refer to as examples of what not to do. That could be useful because you can avoid making a logo that is too reminiscent of something else. Maybe AI could even suggest “gaps” in competitors’ logo designs, things they haven’t done that could make your design more distinctive.

Example: HBO Max’s rebrand from purple to blue. Did someone not realize that the world of entertainment apps is awash in blue? Couldn’t they have assigned AI to go find all the competing app colors and display them in some kind of way that reveals the “gaps” where the new Max brand can be more distinctive? I bet there’s a big gap in the purple range.

nerdrage (profile) says:

captain obvious, but...

“In this customer support center, AI improved the know-how and intelligence of those who were new at the job and those who were lower performers. It suggests that AI could benefit those who were left behind in the previous technological era.”

So is AI not magically going to replace the customer service reps because it can do the job better than them for no pay? I’ve never once encountered a chat bot that was anything but useless but I could see how that could change.

Anonymous Coward says:

Re:

I’ve never once encountered a chat bot that was anything but useless

I’ve rarely encountered a customer service agent that was anything but a chat bot. Especially if it’s an online “contact us” form; inevitably, the response is a form letter vaguely relevant to a couple of keywords, that largely ignores the substance of my message.

In the film Demolition Man, a human operator answers the phone and says “if you’d prefer an automated response, press 1 now”. It was just a throwaway joke, but I’m starting to think I’d prefer the AI.

ECA (profile) says:

To many dont see it.

But did the study go back to the Automotive industry?
Look at the past.
Automated wielding And Any job that does not require moving around ALLOT. Can be automated.
Goto the nearest Store that has YOU scanning the products, and NOT a PAID PERSON. 1-2 persons, manning 8+ machines, NOT 2-3 extra cashiers, that know what they are doing, and KNOW how to get price checks, Answer questions, Communicate with the customer.
They created a Grocery that there WERE NO cashiers, to see what would happen. It didnt work so good, after talking to the customers, They found that they enjoyed talking to the cashiers. The Whole thing has been created again, and ??? made a store that you wonder around, and pick things off the shelves, and the Computer KNOWS what you took, and as you get to the Front, it Charges you AUTOMATICALLY.
(how easy to mis-price goods, and the customer dont even know it Until/IF they read the info when they get home, on the COMPUTER)
Who do you debate the price with? Who do you complain to, FACE TO FACE? What happens if a Jar of pickles Crashes on the floor?

Automated Answering the phone and directing YOU to the correct department(Choose a number 1-9)(yes Iv had 9 choices, sat there a full minute to hear it). A Customer to decide WHAT THE F’ they need?
(Took me a Month to get the Company to forward me to the correct department(advanced corrections(they F’up my data for Signing in)) Then 3 tries to get it DONE.
Automated MEANS they teach the PEOPLE only certain things to do, AND NO MORE. Then they forward you to Another Person if they cant do it in 10 min, and if THAT person THINKS they can do it, they HANG UP on you and tell you it will be done in 24 hours. THEN you call back the next DAY, bitching, and you get Hung up again.
Then they routed me to the MAIN company, that BOUGHT them out, to have THEM fix it, after 6 calls back and forth.(Verizon Bought out 6-8 other Small companies).

Until they get a Automation to MAKEUP BEDS, Any jobs that take allot of movement wont be done. AFTER they do, a Good amount of jobs WILL go away. From Janitorial, house keeping, Bed making, Cleaning the Showers and toilets, will Go along WITH the current selection of jobs.

Kaleberg says:

How is this supposed to work?

If AI lets a less skilled worker to the work that currently requires a more skilled worker, why wouldn’t companies just fire all the more skilled workers and use less skilled workers instead? It would cut the payroll since less skilled workers are more plentiful and cheaper. No one seems to have offered a clue as to how this will create middle wage jobs.

Worse, this seems to assume that companies are run by profit minimizing morons. They may be run by morons, but they are not run by profit minimizing morons. If you can pay someone minimum wage and get the work you’d get from someone getting a premium, surely that money goes to the managers and shareholders.

This is the logic that says lower taxes will increase investment since having more money for managers, moronic or otherwise, and shareholders will somehow create an increased demand for a product. There just aren’t enough managers or shareholders to make that happen.

FloatUnblock (profile) says:

Productivity gains

The main idea seems to be that AI could lead to large (and comparatively larger) productivity gains for lower skilled workers.

If that is the case, why would that help the lower skilled workers? Or any worker?

That’s not a question about technology or even AI. Productivity improvements have not benefitted the people performing the work for the last forty-fifty years, but rather the companies they’ve worked for and their highest level executives.

Which is precisely the point those job doomsayers talk about. AI will lead to massive producitivy gains – but none of those gains will go to the people working, due to how the economy works.

Anonymous Coward says:

Re:

And we’ve already seen this.

It’s called “automation” and has displaced a lot of jobs, and IIRC, it created far fewer jobs than predicted.

And again, the answer isn’t “leave it to the capitalists”, at the bare minimum, it’s independent unions, preferably ARMED, taxing the rich and closing off of loopholes and tax havens.

I mean, look at China. They clearly WANT to use their masses of unemploted for WAR.

Anonymous Coward says:

Re:

There is a limit on productivity gains, and that is who buys what you make if most people are unemployed? An economy is an exchange of value, represented currently by money, and if only the top 1% have money, then that is the number of potential buyers for all the products of their automated factories. In reality their wealth accumulation will be limited by reaching a point where it is redistributed by force, or a more equitable economic system will do the same thing peacefully.

Anonymous Coward says:

The fearmongering in this thread is looking to be a typical right wing extremist attempt to stir the pot with fear and loathing.

This sort of thing will increase leading up the election. It will get even more silly, stupid and ridiculous.

AI is great at spewing bullshit, it would be a great tool for the extreme right and their bullshit. Some far right idiots seem to have difficulty with language and communication, AI will help them with getting their messaging out to the extremists.

Anonymous Coward says:

Re:

Some of us are extrapolating from similar examples.

One of the commenters brings up their experiences in manufacturing and points out that automation has taken away jobs.

I point to actual real world examples where AI HAS replaced creatives.

And with the world on fucking tenterhooks as it is, we’re not trying to sell you the next generation of Smarter Markov Chains; we’re telling you that we’re literally fucked. And in my case, more fucked than ever as the world lurches into another war situation.

Anonymous Coward says:

Here is an example of AI being put to use by fools, resulting in the completely expected outcome.

ChatGPT: US lawyer admits using AI for case research
https://www.bbc.com/news/world-us-canada-65735769

“A judge said the court was faced with an “unprecedented circumstance” after a filing was found to reference example legal cases that did not exist.”

Add Your Comment

Your email address will not be published. Required fields are marked *

Have a Techdirt Account? Sign in now. Want one? Register here

Comment Options:

Make this the or (get credits or sign in to see balance) what's this?

What's this?

Techdirt community members with Techdirt Credits can spotlight a comment as either the "First Word" or "Last Word" on a particular comment thread. Credits can be purchased at the Techdirt Insider Shop »

Follow Techdirt

Techdirt Daily Newsletter

Ctrl-Alt-Speech

A weekly news podcast from
Mike Masnick & Ben Whitelaw

Subscribe now to Ctrl-Alt-Speech »
Techdirt Deals
Techdirt Insider Discord
The latest chatter on the Techdirt Insider Discord channel...
Loading...