As Automation Advances, Are Jobs in Danger?

IRA FLATOW: This is Science Friday. I’m Ira Flatow.

It’s been a common idea since the dawn of the Industrial Age– machines taking jobs away from people that used to do them. Remember? We call that automation.

And the beat goes on. Speakers at the World Economic Forum this year said, quote, “about seven million jobs will be lost, two million gained, as a result of technological change in 15 major developed and emerging economies.” Have we reached that point where more jobs are being lost than created by machines? Certainly machines do excel at some jobs, and there are jobs where human labor is still invaluable.

And as machines and robots and the internet of things, technology, continue to improve, what does the future of work look like? That’s what we’re going to be talking about. Our number 800-844-724-8255, 844-724-TALK. You can also tweet us @SciFri.

James Manyika is a Senior Partner at McKinsey & Company. He is a Director of the McKinsey Global Institute in San Francisco. He was one of the authors of a recent article analyzing which industries are and are not becoming automated.

He joins me from KQED in San Francisco. Welcome to Science Friday.

JAMES MANYIKA: Thank you, Ira. Glad to be here.

IRA FLATOW: Howie Choset is a Professor of Robotics at Carnegie Mellon University in Pittsburgh, Pennsylvania. And you may have seen some of his snake-like bio-inspired devices. He also works on medical robots. Welcome to Science Friday.

HOWIE CHOSET: Thank you for having me.

IRA FLATOW: James, you’ve looked across industries at where automation has really taken hold and where it might be in the future. And what did you find there?

JAMES MANYIKA: Well, this is very interesting, Ira, because what we did was we took an activity-level analysis and looked at something like 800 occupations across the US economy. And what we found is that if you take an activity view, something like 45% of activities that are being done in the economy could be automated with already demonstrated technology. But if you translate that into, so how many whole jobs could be automated, we found that the estimate, at least for the foreseeable future, for the near future, was about 5%.

But this is the thing that was quite interesting, Ira, which is even though we may not see, at least in the next decade, whole-scale automation, we did find that something like 30% of activities in about 60% of jobs could be automated. So in other words, the jobs will actually change quite dramatically.

IRA FLATOW: So what happens to the people who normally would do those jobs? I mean, the skill set level of those people. Can they find other jobs doing other things?

JAMES MANYIKA: Well, first of all, the people are not likely to be totally replaced. They’ll find that they’ll be working a lot more with machines. In other words, machines will complement some of their activities, the 30% that I mentioned. And so those people are going to need to be competent, or the machines are going to have to be designed in such a way that the people can work with them.

But I think to the question of people who may find their whole occupations automated, the roughly 5%, it’s actually important to note that most of those are actually going to be in the middle skill category. And in that category, the middle skill category, the rate of automation may actually be as high as between 15% and 20%. So that’ll be the hardest hit group.

And I think retraining, education–

IRA FLATOW: Well, what kind of jobs are we talking about there?

JAMES MANYIKA: So these are jobs where they’ll typically have three characteristics. There’s a lot of data processing involved. There’s a lot of data collection involved. And/or they work in highly-structured environments.

So think about somebody working on an assembly line in manufacturing. Or think about a clerk who is capturing accounting records, for example. These are all jobs where there’s a lot of data processing involved and there’s a lot of data collection involved and you work in a highly structured environment, which is very easy to automate.

IRA FLATOW: Mm-hm. Howie, you’re a roboticist. Where do you see the future here?

HOWIE CHOSET: So I see the future with robots creating jobs, empowering the American workers on today’s jobs. I think we have only begun to really see the true potential of what technology can do for employment and for productivity.

IRA FLATOW: Mm-hm. Give me an example, for example.

HOWIE CHOSET: So for example, using one of the automation companies with which we work, they want to use these robots called collaborative robots. Robots that work in close proximity to people and with people at the same time in order for that person to carry out his task. Not only will he carry it out more quickly, they’re actually more concerned about the quality of the task that this person will be able to carry out. Fewer errors, it’s safer, and it’s more ergonomic for the person as well.

IRA FLATOW: Yeah, no. You brought up safety. I always think of one of the great advantages robots have is that they can go and work in unsafe places.

HOWIE CHOSET: Yes. One of the benefits of robots is that they can get into, you know, hard-to-reach difficult places to do tasks that people shouldn’t be doing. Like, you know, in a nuclear power plant disaster, for example.

In this case here, what I mean by “safety” is that a worker, you know, an associate working on a line, often they have to reach into configurations that are just hurtful to their bodies. They’re not ergonomic. Now with this robot that’s both a tool and a partner, they can carry out the task and it still be ergonomic. They’re not going to get hurt. They’re going to be more safe.

IRA FLATOW: Is that something like your snake-like robots get into hard-to-reach places?

HOWIE CHOSET: Well, I do have a bias that I like thinking about snake robots getting into highly constrained spaces, you know, to do inspections and repairs and an assembly tasks. But, yes, that would be an example.

IRA FLATOW: Mm-hm. Our number 844-724-8255. Let’s go to the phones. People with some interesting ideas. Let’s go to San Francisco and [INAUDIBLE]. Hi, welcome to Science Friday.

SPEAKER 1: Hi. Thank you for letting me on. I’m a really big fan of the program.

IRA FLATOW: Well, thank you. Go ahead with the question.

SPEAKER 1: So certainly. So actually right here in San Francisco, I’m creating jobs through automation. We run 3D printing shops. So I’m actually curious about the guests, like, you know, with us, we’re using 3D printers to make things locally instead of sourcing the jobs overseas. I was wondering what the guests thought about how, you know, this might– you know, granted we’re taking jobs away from, you know, over in China, but we’re actually bringing them back to America. So I was wondering about your thoughts on that.

IRA FLATOW: James Manyika?

JAMES MANYIKA: Well, I’m actually very excited by the potential for 3D printing in terms of what it my due to fueling a Renaissance, if you like, in manufacturing. So I think, you know, what you’re describing and you’re working on is tremendous, because it’s going to enable more individuals to be able to participate in manufacturing. So I think that’s a good thing.

And I also think it’s important to go back to something that Howie said, which is one of the things that we should think about when we come to these questions of automation is that quite often the output is actually better. In other words, it’s not just about matching what humans can do, but it’s also about going beyond that in many cases. Less errors, more perfection, better outputs, better products. So I’m actually quite supportive. And I think I’m excited about the possibilities of 3D printing.

IRA FLATOW: Do you ever see robots taking the place of where the human touch is needed? Like in caregiving, hospitals, nursing, things like that. Howie, do you think that those are places that they really need a real person?

HOWIE CHOSET: So in caregiving, obviously you need a real person for that kind of relationship. But I do view us developing intelligent technology, robotic tools, that can do some of the more mundane tasks that I don’t think the caregiver wants to deliver, nor I think the patient wants to receive from a person. You know, certain kinds of behaving and cleaning, for example. I think a person would rather be able to do it himself with a mechanism than having another person come in and clean him.

IRA FLATOW: You know, we hear of, and we know of, robotic surgeon– robotic surgery going on. I mean, it’s not going to put surgeons out of business, is it?

HOWIE CHOSET: Oh, no. What we’re really doing is we’re– what we’re doing is– this is a common theme– we’re democratizing the ability to deliver medical care. Now with robot surgery, as an example, we’re offloading a lot of cases that surgeons may not even want to do. They want to do the special, the more complicated work. You know, that’s why they’re surgeons.

We can offload them to non-surgeons who can now deliver the same level of care. And I might add not only minimally invasively, so it hurts less, but also at a lower cost and not requiring all the infrastructure that you would need in a hospital. So now you have a situation where robotics, surgical robotics, are creating jobs for medical care to be delivered. And more medical care can be delivered at a lower price so everyone benefits.

JAMES MANYIKA: Well, in fact, if I could just add one point to that, Ira. if I may.

IRA FLATOW: Please.

JAMES MANYIKA: I mean, I think there’s a useful historical analogy to what Howie’s just describing. Think about the bank teller, right? So the job of the bank teller three, four decades ago involved helping you get money out of your savings account. Well, we automated that part of it. And it doesn’t mean that bank tellers went away. In fact, people were predicting that there’d be no more bank tellers.

But what actually happened is the opposite. So the role of what a bank teller is changed quite dramatically to doing these higher order, if you feel like, human interactions and other functions that didn’t involve simply helping you get money out. In fact, the number of bank tellers went up, mostly driven by the fact that there’s more demand. We opened more branch banks, and so the number of people working as banks tellers actually went up.

IRA FLATOW: Mm-hm. Well, you know, as we create robots, we also will need to create the software, the artificial intelligence, just the code that has to be programmed into these robots. So is that a job gain there?

HOWIE CHOSET: So in that regard–

JAMES MANYIKA: Oh, absolutely.

HOWIE CHOSET: Oh, I’m sorry.

JAMES MANYIKA: Sorry. Go ahead.

HOWIE CHOSET: Yeah, so in that regard, not only are we creating jobs because we’ve empowered the American worker, but we’re also creating jobs for people to deploy, maintain, and perhaps, you know, operate these robots. And these are, you know, fine jobs to have. On top of that, we’re also creating lots of jobs for software developers. You know, higher-level thinking and creativity to make our mechanisms that much more capable.

IRA FLATOW: We now have lots of websites, videos, whatever, that teach people things. But they’re not teachers. Are we going to be replacing teachers? Or do teachers have a special place because in schools they’re more hands-on with the students? What do you what do you think about that, James?

JAMES MANYIKA: Well, I think when we did our analysis, I think teachers were actually in that category of the least likely to be automated, for most of the reasons you just described. Because keep in mind that one of the things that’s quite fascinating on this question of jobs and automation is that you sort of have to think through, you know, four or five different factors. One is a question of technical feasibility. Can you actually technically automate the task.

But there’s a whole other set of considerations. The cost of doing that. The alternative labor supply dynamics. The benefits of doing that. But also the social acceptance. I mean, there are things that we, I think in a social sense, think ought to be done by humans. So I think when you take all that into account, education and teachers and a lot of caregiving are likely to be things that we’re going to, at least for the near-term anyway, want to have humans do that.

IRA FLATOW: Mm-hm. I’m Ira Flatow. This is Science Friday from PRI, Public Radio International, talking about robotics and about automation. Old topics that we’ve been talking about for over a hundred years about what’s the workplace going to be like.

Let’s go to the phones. Let’s go to Pittsburgh, PA. Matt, welcome to Science Friday.

MATT: Hi. Yeah, thanks for having me.

IRA FLATOW: You’re welcome. Go ahead.

MATT: So I work for a delivery company, I won’t name them, but they don’t have any infrastructure in the area. I just work through the app. So it’s kind of like my job is working under a robot already. So hearing this segment kind of just made me think about– well, first of all, delivery jobs probably won’t be there in the future, because it’s an obvious place for automation.

And second of all, it’s kind of different when you don’t have a boss that you talk to every day. I mean, you just signed in through an app, and, like, get punished through the app or have to call off through the app, it’s a little different.

IRA FLATOW: Tell me. Wait a minute. What do you mean you get punished through the app?

MATT: Well, the app will be like, oh, you didn’t pick up that order fast enough. Like, what happened? And then they’ll be like, if I don’t make the order on time, then I don’t get paid the full amount that I’m owed.

IRA FLATOW: And you have no recourse to go to your boss, because there’s no boss sitting in the office next door?

MATT: Yeah, exactly. The boss is in Philadelphia. They don’t have any infrastructure in the Pittsburgh area.

IRA FLATOW: Wow. Uh, let me ask my guests. Howie, what do you think of that? James? Are you guys there?

HOWIE CHOSET: Yeah, for that–

JAMES MANYIKA: Yeah. So one of the– sorry. Go ahead, Howie.

HOWIE CHOSET: No, please. Please. You please go ahead.

JAMES MANYIKA: No, I was going to say, one of the things– to the caller’s point– one of the tasks that’s actually very hard to automate, at least in the work that we looked at, was the idea of managing and goal setting and the kinds of things that managers do. So while the task may be easy to complete in the way that the caller’s describing, the whole process of managing and setting goals is actually very difficult. So there is something lost when, in fact, as the caller described, there’s no way to recourse, no way to discuss and apply judgment to the things that the caller described. Maybe punishing or not. So that is something that could be lost.

IRA FLATOW: Howie, do you have something to add?

HOWIE CHOSET: I have nothing to add to that.

IRA FLATOW: OK. Let’s go to see if we can get our last call in in this hour. Let’s go to Rochester, New York. Hi, Mike. Welcome.

MIKE: Thank you. Great show.

IRA FLATOW: Thank you.

MIKE: So my concern is as automation continues to progress, how are we going to restructure our world so it works for everybody?

IRA FLATOW: Hm. Small topic.

[LAUGHTER]

Well, it is a worldwide– you know, the earth is flat. The world is flat now. Do you think about this, James, at McKinsey?

JAMES MANYIKA: I actually do. And I thank the caller for asking the question, because I think it depends what we’re trying to solve for. Because on the one hand, if we’re trying to solve for work, meaning other things for people to do, I think there’ll be plenty of things for people to do.

But if the question is how do we solve for things like incomes and wages, that may be a slightly different question. Because one of the things that happens, by the way, economically, is that as people work in a way that’s augmented with technology, in many instances– not all– but in many instances that sometimes has a depressive effect on the wages for the person working with the technology. Not always.

So the question is if we’re trying to solve for other things for people to do, I think that’s an easier thing to solve for. But if it’s trying to solve for incomes and standards of living and so forth, that becomes much more problematic. We may have to think about, you know, when much of the work driving output in the economy’s being done by machines, what wage models look like in that world. And that’s a big question. And I think it gets to some fundamental, sometimes ideological, questions about work, wages, and incomes.

IRA FLATOW: Yeah. It’s being debated, certainly, in this political season. I thank both of you for taking time with us today. James Manyika is a Senior Partner at McKinsey & Company and Director of the McKinsey Global Institute in San Francisco. Howie Choset is a Professor of Robotics at Carnegie Mellon in Pittsburgh, Pennsylvania.

Thank you, both, for taking time to be with us today.

HOWIE CHOSET: Thank you.

JAMES MANYIKA: Thank you, Ira.

IRA FLATOW: You’re welcome. Have a good weekend.

Copyright © 2016 Science Friday Initiative. All rights reserved. Science Friday transcripts are produced on a tight deadline by 3Play Media. Fidelity to the original aired/published audio or video file might vary, and text might be updated or amended in the future. For the authoritative record of ScienceFriday’s programming, please visit the original aired/published recording. For terms of use and more information, visit our policies pages at http://www.sciencefriday.com/about/policies