Randy Scott

What if AI transformed your job so you had four-day weekends or never even had to work again?

You're listening to Where What If Becomes What's Next from Carnegie Mellon University, where we're exploring what AI means to industries, government, society, and you.

On this episode, we're delving into artificial intelligence's impact on the future of work. How will AI reshape job markets, the workplace, incomes, and the economy? Let's find out.

Joining us today are three experts on AI in business and work. R. Ravi is the Andres Zoltners Professor of Business, and Professor of Operations Research and Computer Science at Carnegie Mellon University. Laurence Ales is a Professor of Economics at the Tepper School of Business at Carnegie Mellon University. And Calum Chace is a global speaker, advisor, and the bestselling author of “Surviving AI” and “The Economic Singularity.”

The rapid advancement of artificial intelligence is poised to dramatically transform the landscape of work. This transformation is not limited to automation replacing manual jobs, but extends to deeper changes in job roles, the structure of industries, and broader economic dynamics. As AI technologies evolve, they are reshaping the very nature of work, creating new opportunities, and presenting unique challenges for workers and businesses alike.

According to a recent report by Goldman Sachs Research, around 4 % of all U .S. firms have adopted AI into the workplace. And that is expected to double in the next six months and rise to more than 70 % of U .S. companies embracing AI by the year 2030. So how impactful will AI be in the workplace? Laurence, what's your take?

Laurence Ales

It's a transformative technology and the impact is going to be widespread.

What is perhaps of great interest is the speed at which the technology will have an impact. This is not the first transformative technology that we've lived through as a society. Previous technologies might have included anything from the rise of the assembly line, from electrification, from the rise of IT, rise of internet and other forms of communication. What is unique in terms of AI,is that now we're thinking about a technology which is intangible, meaning this is a technology that you cannot go and touch. I could tell you go and touch your computer, and you cannot do this with an algorithm. Now, this distinction is important because with technologies that are tangible, the physical entity is a limit on how quickly we can build them and deploy them.

With an intangible technology, something you cannot touch, you can scale it up really quickly, meaning that within a week, we could all adopt to the latest form of AI or latest form of generative AI, and that could scale and serve all of us. So this is something that is unprecedented in this space, and that's why the adoption is being so widespread and will only be more widespread as we go forward.

Randy Scott

Ravi, you have a different perspective.

 R. Ravi

My own personal take is that it probably will not be as impactful as, say, electricity or even the internet, but it's going to have a profound influence on sectors of business where information is the workhorse. And by this, I mean media, news, and various forms of entertainment and technologies where information is the product.

Randy Scott

AI has the potential to bring enormous benefits to the workplace. Among them, automation of repetitive or manual tasks that creates greater productivity and efficiency, more time for better decision-making, opportunities for innovation, enhanced capabilities for remote working, better recruiting efficiency and hiring results, even real -time translation for breaking through language barriers. What's your view on where AI is going to impact the workplace?  Ravi?

R. Ravi

As far as these routine tasks that I think many of us are accustomed to, like answering routine emails, reconciling, doing reimbursements, planning travel, simple data processing tasks, a lot of which involves low-level decisions, but I think the general direction is clear. These are activities that a large number of information workers are actually currently undertaking, and this is a significant part of their time. And I think such loosely structured information tasks will get automated in the future. So what that means is the time that these workers have spent on these loosely structured tasks of converting what we might call unstructured knowledge into more codified or structured knowledge.

So the time that's freed up by doing that, the key question is what will we do with that time? Are we going to try to insist that these workers do more with the time that's released? Are we going to have some new kinds of abilities or new products that these workers will be working on? I think that's the big question.

Randy Scott

Laurence, what's your take?

Laurence Ales

There are, at least I would think, two ways in which this could play out. One is this scenario which you're describing, which is now I don't have to send all these emails and I can reallocate my time into doing something that I'm perhaps better at. I'm a professor in a university, so I would do more research or more teaching, which are the things that I'm being hired for doing here. There is an alternative, however, which is we've seen in the past, which is if my time gets reallocated, an alternative could be that that time goes away.

Meaning, let's take an example where a worker whose job description is spending half of his or her time sending emails and communication. Suppose that now that we can cut that in half. So what you're describing first was like, well, this person can move into doing something else, or we could have fewer of these people doing that set of tasks. Now, that's important because the moment I need fewer worker per every amount of output, what could happen is the following scenario, which is suddenly the cost of creating output goes down. When things get cheaper, people start demanding more of them. So we can actually have a loop where we end up having more and more workers to begin with.

Randy Scott Let's talk about jobs. There is a lot of hype and fear that AI is going to replace some jobs, a lot of jobs. But on the flip side, AI is creating new jobs in tech, analytics and more. So perhaps it's about displacement of jobs more than disappearance. Net loss, net gain. Either way, AI will be a job disruptor. Ravi, can you shed some light on AI and jobs?

R. Ravi

We have seen with the earlier innovations and revolutions like electricity, the automobile, very similar fear existed in the marketplace about, you know, coachmen losing their jobs.

I think this time is slightly different though, simply because we have grown a healthy amount of white collar workers just to mainly digest, process information. So I think what might be more important here is people who are simply processing and calibrating and classifying information, those jobs might go away. But people using human judgment to act on that summarized information, their jobs are going to get even more important.

So the big question in my mind is, is there a healthy proportion of these white collar jobs and knowledge jobs that involve judgment? And if I look around and informally think about it, my worry is that there is a significant portion of knowledge workers and white collar jobs, which are simply doing the calibration and the classification, the tasks that could be taken over by AI.

So I believe this is going to be different than the earlier revolutions like electricity or even the internet and the telephone. I think there is going to be a loss of a significant fraction of these knowledge jobs.

Randy Scott

Calum?

Calum Chace

I think reports about how many jobs it's going to destroy and how many jobs it's going to create in the next 12 months, 24 months, whatever, they're heroic and completely pointless. They are going to be wrong by orders of magnitude.

We just don't know. But what I think we can say is that AI will lead to jobs changing. Jobs are made up of tasks, and a lot of those tasks will be automated so that machines can do them. And the jobs that those tasks are part of will be transformed as a result. And that means that people are going to need to change jobs. They're going to need to upskill themselves so that they can do the more interesting things that the machines can't yet do. Over and over again, I call this the great churn. I think we're at the beginning of a period of great churn as people need to change jobs and quite often change companies, change industries as the jobs that they're doing are transformed.

Randy Scott

And what about wages? High skill workers in AI driven industries might see wage increases due to demand for advanced skills. Conversely, middle skill jobs might experience wage stagnation or decline as AI automates certain tasks. How will AI affect people's paychecks? Laurence?

Laurence Ales

In general, we should not think, however, that technology is always polarizing. Our own research with colleagues here at Carnegie Mellon has emphasized that some technology can indeed be beneficial to lower paid workers and increasing their wages. The idea here would be that an AI might help lower worker achieve occupations or perform tasks that before this technology, they were just out of reach. Indeed, if we go back even further, the story of the Industrial Revolution is a story of enabling workers, manual workers, that before had no access to middle income jobs, to now suddenly be moved to the middle of the income distribution. So in that sense, that technology, that set of technology that you know, we associate with industrial revolution were not polarizing, were quite the opposite. They were hitting at the top by lowering wages and employment of artisan and pushing everybody into the middle. So what will ultimately happen to AI? I guess it depends on how we'll be adopted and used.

Randy Scott Winners, losers, are there certain industries, sectors or jobs that are in greater danger or poised to thrive in an AI powered workplace? Ravi?

R. Ravi

Information based industries are going to be at a disadvantage. So broadly, I think people working in media production, entertainment, they're going to see an uphill battle to compete against the tools that will automate away and maybe even reduce the, you know, increase the amount of output that is possible now in things like, you know, music production and so on.

That means that in that area, there's going to be more inequality. The heavy hitters, the best performers, the ones that produce very unique content are going to command much more of a premium than is possible in these information-oriented fields.

Turning to populations, we already said white collar workers, there is a great risk, right? But I think what that's going to do is that the more tactile skilled workers, people who are doing plumbing, carpentry, I think the demand for them as well as the ability for them to accomplish more in terms of what their job descriptions are is going to increase. So we are going to see an improvement in the job opportunities both below and above the white collar workers. The management class is going to have a much larger set of knowledge tools.

and the tactile class, the worker class below the knowledge workers are also going to get a lot of gains from information tools. Suppose you're a plumber and you're trying to figure out how to solve a complicated problem, the AI tools are going to help you, able to debug what you took several hours earlier. So you're going to see these productivity gains supercharge employment in these two areas.

Randy Scott

Calum, your thoughts?

Calum Chace

Undoubtedly, yes, but we can't predict what order they will you know, what order different jobs and different industries will be susceptible to automation in what we can't say who's going to be first. Two or three years ago, everybody assumed it would be drivers because self -driving vehicles were progressing fast. Now it turns out it's translators and writers of not terribly good marketing copy. You know, those people having to find new jobs. It's a very fluid situation because everything's changing so fast and they're always winners and losers. So.

In the last few years, if you happen to be good at wrangling algorithms for deep learning systems, you could name your own salary. You still can. And there are other people whose whole livelihoods are going to disappear because what they did can now be done by a machine. But they can retrain and there are other jobs that they can do.

Randy Scott

We're talking about AI and the future of work. When we come back, we'll talk about how you should adapt to this change the dangers and risks of AI in the workplace, and hear from our guests on what's next for the workplace.

Sponsored Content

What if we could reimagine how scientists conduct science? What if researchers on opposite sides of the globe could collaborate on experiments in real time? And what if it all could accelerate the pace of discovery and ignite a new scientific revolution? Carnegie Mellon University's AI for Science has answers. And with the first AI -enhanced laboratory of its kind, Carnegie Mellon is driving what's next in the quest for safe, responsible, automated science. Learn more about AI for Science now at ai.cmu.edu.

Randy Scott

Welcome back. You're listening to Where What If Becomes What's Next from Carnegie Mellon University. And today, we're talking about AI and the future of work. While experts can differ on who will win or lose in an AI -powered workplace, they can all agree on one thing. Work will change, and workers will have to adapt themselves to prosper. How should a worker prepare for an AI workplace? What skills should they learn? Calum.

Calum Chace

Well, the most important thing is to get familiar with AI, to get as much hands-on experience you can with generative AI, which is the current leading form of AI. Use it in your personal life, use it in your professional life to the degree that you're allowed to, and become as expert at using it as you possibly can.

Randy Scott

And Ravi, what would you advise those entering the job market? How should they prepare themselves?

R. Ravi

I would advise to focus on one of two different directions.

One, is get technically strong enough so that you are involved in the creation of these next generation of AI tools that allow us to do the summarization, that allow us to develop these recommendations. So that's definitely a growing job market. And for the very technically gifted amongst us, that's what I would recommend. For the others, I would urge them to study areas that would encourage better decision-making. And now I'm thinking of things like sociology, psychology, human judgment, ethics, and to some extent economics, because economics is going to help us translate the gains that we get and distribute them.

Randy Scott

In addition to learning AI skills, a worker may also require upskilling, which means learning new skills for a worker's existing job or role, and re -skilling, which means training for an entirely different job or role all while embracing lifelong learning. That could be something of a challenge, right? Laurence, your thoughts?

Laurence Ales

Indeed, it is a challenge because we usually traditionally think about an investment in human capital in education that lasts a lifetime. And even in the face of previous technological changes, that was enough because if there were technology change, they would happen slowly enough that, I guess, unless you were really unlucky they would not be impacting so much your career. Now the landscape that we have five, 10 years from now could be very different from the ones today. My recommendation to students would be, well, instead of getting too hung up in learning one thing or learning one topic, one superpower that you may have is learn how to learn.

Randy Scott

Are there risks, dangers, and ethical issues we should consider? For example, as AI tools learn more about employees, are there invasions of privacy we should consider? Are there built -in biases when AI algorithms read and then recommend resumes? Do we need safety social nets to protect disadvantaged workers? Any regulations? Laurence?

Laurence Ales

Yes, I could maybe make two points with that on that point, which is, first of all, one thing that is of potential concern here is speed. You know, we are thinking about...

change in technology, which is unprecedented in terms of speed. And we should be cognizant that our institution, our workers, our society works at a completely different time horizon. And whenever there are changes that are faster than the society can adapt to, the outcomes are rarely positive. So for a policymaker, one thing I would urge them to think about is are there grounds for here to you know, is this the right speed at which we should progressing and thinking about what else is being left behind.

Randy Scott

Ravi?

R. Ravi

Very much so. When we see domains of work shift and also the fraction of the population engaged in different types of work move, we have a responsibility to think about, you know, the society's obligation to the advantaged and the disadvantaged populations. So there's going to be a societal question. The way we have done this typically is via politics and we have not done this very well. We have done this in a very shattered way across the world. So one thing that we might have to face is an international consortium of some sort to try and get at the political ways of redistributing productivity gains coming from AI. So the big questions are going to pertain to things like, should there be a universal income? What should we do with all the productivity and the increase in GDP that comes with these tools? I think that's going to be another big conversation.

Randy Scott

Calum?

Calum Chace

The issue of regulation is another vexed one. So on the one hand, people quite reasonably say most regulators do not have the expertise to be able to frame sensible guardrails around AI. And there's a danger that if you try to regulate AI, then all you really do is you hobble the creation of very good systems in one country and people in a different country, and let's call that different country China or Russia, don't have those regulations and they advance more quickly. And there's something in that argument, clearly. On the other hand, people who say, therefore, just let it rip, that will not fly. Artificial intelligence is our most powerful technology. It's going to change everything in the world. It is inevitable and right and proper that the politicians that we elect have a role in saying which bits should slow down and which bits should not be allowed and which bits are fine. We have to regulate it. We just don't have an option not to.

Randy Scott

Looking into your crystal ball five or 10 years out, what's next for AI in the workplace? Ravi, what's next?

R. Ravi

I think that the next phase in work has to be collaboration. I think there is going to be a human need and also a societal need to work across disciplines to make something new. So I think AI is going to somehow force us to come together because we're going to get tired of all the rule -based tools that we're going to see. And work better mean working with other humans for us to feel more meaning and dignity.

Randy Scott

Laurence, what's next?

Laurence Ales

We will work less. It doesn't matter how complementary or substitutable you are with this technology. Nowadays, we are thinking about a two -day weekend. Let's go back 100 years ago. That would have been unheard of. Will we start having three, four-day weekend in the future? Most likely.

Randy Scott

Any final words?

Calum Chace

Whenever anybody says to you, AI will do this or it won't do that, ask them what timeframe they're talking about, because what it can do today is wildly different from what it will do in five years time.

Randy Scott

Thank you to all our guests for offering their insights and thanks to you for listening. Please check the show notes where you'll find links to additional resources for this episode. And to learn more or contact us with a story, and to learn more or contact us with a story idea or comments, please visit ai.cmu.edu/podcast.

If you are enjoying this show, please rate and review us on your favorite podcast site and share with friends. And be sure to subscribe so you'll never miss an episode. You've been listening to Where What If Becomes What's Next from Carnegie Mellon University, where we are exploring what AI means to industries, government, society, and you.

Stay tuned and stay curious.