To Thrive Alongside AI, Find the Bottlenecks (with Ben Jones)

View the full episode transcript.

Will AI actually take over every job ever? Economist Ben Jones doesn’t think so. Drawing from historical evidence around innovation, Jones argues that automation never conquers everything — it inevitably creates operational bottlenecks that only humans can solve. By leaning into these bottlenecks, we can protect meaningful, well-paid work. But his optimism comes with a stark caveat: there is one nightmare scenario where this historical pattern breaks down entirely. And it doesn’t involve superintelligence. It involves mediocre AI.

Further Reading: 

How AI Could Benefit Workers, Even If It Displaces Most Jobs — Benjamin Jones, AI Frontiers

This is a computer-generated transcript. While our team has reviewed it, there may be errors.

Jess Love: I’m Jess Love. This is Life Automated. For generations of Americans, the question was, where were you when JFK was shot? 

[Breaking news bulletin announcing the death of John F. Kennedy]

 Good afternoon, ladies and gentlemen. You’ll excuse the fact that I’m out of breath, but about 10 or 15 minutes ago, a tragic thing from all indications at this point has happened to this city.

Jess Love: When Neil Armstrong walked on the moon?

[Audio Clip of Neil Armstrong on the moon]

That’s one small step for man, one giant- 

Jess Love: Or the towers fell on 9/11?

[News bulletin from September 11, 2001]

There’s smoke billowing out of the World Trade Center. 

Jess Love: These are moments seared into our collective psyche, and now we have a new moment, only this one arrived a bit differently, a browser at a time.

Jess Love: Where were you when you first used ChatGPT?

This is Life, Automated. I’m Jess Love, and I can tell you exactly where I was. The year was 2022, and I was visiting my in-laws in Florida. It was that lazy week between Christmas and New Year’s that I usually spend in a food coma. I sat on my sister-in-law’s childhood bed and signed up for a free subscription to ChatGPT.

I started asking the machine questions, easy ones at first. What did it know about butterflies native to Chicago or the most common tourist attractions in Italy? But it wasn’t too long before I asked what I really wanted to know. Could it do my job? 

See, much of the work I’ve done over the past decade is in science communication, essentially translating original research written by and for experts into something comprehensible to most people.

So, I asked ChatGPT to summarize some papers I was familiar with, and it did okay. But what it lacked in finesse, and especially in those earlier iterations, accurate citations, it more than made up for in sheer efficiency. In just seconds, it could do work that would take a human the better part of a day.

If I didn’t like the first draft that was conjured, I could just tell it to try again from a different perspective now. Write it shorter. Write it longer. Write it for a fifth grader. Write it with sports metaphors. I was stunned. It seemed clear we were entering a new era of science communication where all the world’s knowledge was not only available, but personally customizable on demand.

If you wanted a given study to be relayed in the style of Mr. Darcy flirting with Elizabeth Bennet, you were now suddenly in luck. 

[AI-generated summary of Pride and Prejudice in style of research article]

Darcy’s mouth tightened in a way that suggested the accusation had landed, but he proceeded anyway. My dear Miss Bennet, said Mr. Darcy with an air of careful restraint, I have been reading “AIG in Hindsight.” It is an instructive account of how a firm may appear a paragon of prudence until one… 

Jess Love: The thing I did to make my living that made me special, that made me employable, that I loved to do, was now available to anybody. How could I possibly compete? What did all of my expertise even mean anymore?

And the more I thought about ChatGPT, and the more its implications for a lot of other professions became clear, the more I started wondering about all of us. I have a seven-year-old daughter. What opportunities are gonna be available to her? I’m not alone here. The majority of Americans feel pretty anxious about AI.

And it’s not just jobs. Public polling data suggests Americans are really concerned about how AI will affect education, democracy, our environment, our information ecosystem, and just how we relate to each other. But even as I share all these concerns, I gotta say, over time, a big part of me did start rooting for this technology.

Partly, this is because I now work at a research institute, the Ryan Institute on Complexity. I use AI tools every day, and so do the people around me. I see firsthand how useful they can be. But there’s a bigger reason. I have another daughter, an 11-year-old. She was born with a rare neurological disorder, so rare that it doesn’t actually have a name.

You can currently count the number of people in the world with her specific genetic mutation on two fingers. And this genetic difference, it affects how she talks, how she moves, how she uses objects like forks and scissors, how she thinks and learns. My daughter lives in a world that wasn’t made for her, with a brain that science doesn’t really understand.

So, technology that can harness the world’s cumulative knowledge in order to make her life a little easier sounds pretty incredible. AI that can predict which medications might be useful or offer step-by-step directions on how to operate household appliances, and then wait for her to finish each step.

[Audio of AI Microwave]

First, press the button above the handle and open the door. Your food will go inside that box when it’s ready. Now close the door. Did it click? 

Jess Love: Or AI that can adapt educational materials to her specifically, on demand, maybe in the voice of her favorite Moana character. 

[Audio of Maui from Moana] 

You’re welcome. 

Jess Love: Sign me up. So yeah, that’s me. One person, one employee, one parent, reckoning with this technology. But of course, it isn’t just me. There are millions, billions of us wondering what AI and other kinds of automation really mean for our lives and our businesses and our communities, and what we should be doing to prepare. If you are one of these people, I’d like to invite you on a journey.

Because nobody can predict the future, nobody has all the answers about where we’re headed, but there are a lot of people with some answers, and I’ll be speaking with them on Life, Automated, a new podcast from Kellogg’s Ryan Institute on Complexity, distributed by KQED. I’ll talk with scientists and philosophers, business leaders and engineers and artists, all with unique insights into the world we’re collectively building.

I’m not gonna try to freak you out. I’m not gonna try to sell you anything. I am gonna get nerdy. I’m gonna try to understand, as best I can, where we’re headed. And you’re invited.

Shortly after I had my initial ChatGPT moment several years ago, I tracked down Ben Jones. He’s an economist and a co-director here at the Ryan Institute where he studies economic development, innovation, entrepreneurship, and the scientific process. And it turns out, Ben has been paying attention to new developments in generative AI for much longer than me.

When we talked three and a half years ago, I was kind of at peak anxiety; but Ben, he’d had a head start. Speaking with him helped me feel better about where we were headed, in part because he gave me a helpful way of thinking about AI. So, I wanted to bring him back for another conversation, this time in a recording studio. I started off by getting him talking about the first moment he realized AI had the potential to bring about big changes.

Did you have an AI moment, kind of a single moment that sort of shook you to the core and made you think, wow, something about these newest models is really fundamentally different? 

Ben Jones: I did. I’ve had several AI moments along the way. I think many people, you know, when you see what AI can do, it’s uncanny.

Sometimes the hair stands up on the back of my neck. But my first AI moment where the hair stood up on the back of my neck was in 2017, and I was in Toronto for a conference that mixed economists and technologists, people from Geoff Hinton’s lab, who would go on to lead many of the top AI groups at the top companies.

And we were shown over dinner, it was a very nice dinner, a video. The video was of an AI learning how to play a first-person shooter game. And as you may know, with first-person shooter games, the experience is if you’re looking through the eyes of the player. So, we’re sort of looking through the eyes of the AI.

And without any particular directives, except to sort of survive as long as possible, the AI was training itself to play this game better and better. And at first, you’re sitting there looking through the AI’s eyes, and you’re kind of like a drunken sailor. You’re, like, stumbling around, you’re banging into walls, you’re getting trapped in corners.

You never pick up health or a weapon. You just don’t last very long. The video showed us the AI’s progression kind of in a fast-forward, high speed over maybe 24 hours. And so around the middle of the training, over, maybe 12 hours, it started running around the maze extremely effectively, but it wasn’t always picking up things on the floor, and it was no longer falling off bridges.

By the end of the day, it was not only picking everything up and firing the weapons accurately, it was killing every single person in the game, all the real players. It didn’t miss. It didn’t miss in any way. And just watching this progression from this AI algorithm was a really uncanny moment, and to me, opened my mind to how powerful but also potentially worrisome AI is.

Jess Love: Yeah, and I think, you know, on some level it’s not that surprising that a computer can learn to master a computer game. But what makes this story so incredible is just that the computer had no instructions that told us what it was supposed to be doing or what it was supposed to be optimizing other than just, like, live.

And it was able to just learn all of those techniques just via trial and error. 

Ben Jones: Right. Just the certain algorithms for training AI with fairly simple objectives lead to this incredible set of capabilities, which are truly uncanny. Many people have probably had that first moment now with AI, I think a lot with ChatGPT, where you ask it a question and you can’t believe the answer.

You’re like, “How could it possibly be coming up with this?” And AI absolutely has that quality. So, for me, it kind of happened a bit earlier, and so I started writing about it a bit earlier than most because I’d had that experience a bit earlier than most. 

Jess Love: Yeah, which was very convenient for me personally because it meant that, when I was still in my AI freak-out phase, you were somebody who had been doing research on this for years, and were able to kind of be this, calming presence, in order to give me some concrete information about this technology.

So, you have kind of an interesting perspective, which is that AI, the current generative AI that we’re talking about with ChatGPT, it’s really just part of a long process of automation. 

Ben Jones: Correct. And it’s true that certain technologies for a long time, and AI is one of them, are just hands down amazing.

They do things that are just amazing. But that doesn’t mean that they disemploy everybody.

Jess Love: Phew.

Ben Jones: And, how can those two things be true? Well, let me just give you another example. You know, obviously computers have been around for a long time now, and their capacity to do sort of certain kinds of calculations are just, more or less infinitely, virtually infinitely better than humans.

You know, in my field and in many research fields, you use data and statistical regressions. If I were to run a regression by hand, with a large number of observations, it would take me literally my entire life. And I would make mistakes. You know, my computer can do a regression with a million data points in a minute.

And so, we have these technologies that basically are infinitely better than us to do these tasks that are really important to what we do. And then you can ask, okay, well, has our understanding of economics or astronomy or physics multiplied by a similar factor? And you’d say no, because although we can do certain things really, really, really incredibly well, other things are holding us back.

But that doesn’t mean that you don’t have these other tasks, which I would call bottlenecks.

Jess Love: Okay, and this is the thing that I have thought about so often since our conversation. It kind of blew my mind. So, can you talk to us about what a bottleneck is and why it’s so important?

Ben Jones: Well, of course, it’s a metaphor from a bottle, and what do we know about bottles? We know that the speed at which water or wine will flow from the bottle depends on the width of the neck. Doesn’t really depend on how big the bottle is otherwise, how much wine there is in it, how big the base is. It really just depends on the neck.

And in many processes in the economy and in the economy overall, it looks like there are certain things that have to happen that act as bottlenecks. Let’s say you’re going out to see your cousins for Thanksgiving dinner, and you’ve gotta get everyone in your house into the car, and you’ve gotta pack up the pies that you’re bringing and contributing to the festivities.

Well, everyone in the house could have gotten up early, gotten dressed, put on their Thanksgiving sweaters, gotten the pies into the car. But you’re not leaving the house until everyone has their shoes on and their socks. And maybe there’s someone in the family who just is getting real slow about getting their socks and their shoes on.

How fast is it that you leave the house? It doesn’t really depend on how fast everyone is. It depends on how slow the slowest person is. That person becomes a bottleneck. Maybe an example we can all relate to. Much of the economy has that flavor, where it’s not how good you are at a lot of things. What’s really sort of determining the rate of the economy is the things we’re quite slow at, the things we’re not very good at.

Jess Love: So basically, it’s this task that is slowing down the speed with which the entire overall goal or objective can be met. 

Ben Jones: Correct. And the reason that’s important is because often in innovation, we get really focused on the shiny thing that’s going really fast, and we’re really excited about how fast it can compute things and these kinds of new technologies.

But often what really matters is the thing we’re not looking at. It’s the things that we’re quite bad at. The things that we’re really bad at are actually going to slow down the whole system, and if we can’t improve those, the gains from these kind of shiny new objects and new technologies won’t be as great as we hope.

Jess Love: I thought about these bottlenecks a lot after our initial conversation. Like, whenever I heard a tech leader say AI was going to be ready to colonize the galaxy in five years, or whenever I started spiraling about whether AI was going to take over every job ever. Now, about that. Overall, Ben, like many economists, is bullish on AI.

He says that automation is generally good for workers. Sure, it causes some job loss and dislocation in the short term. But in the long term, it raises the quality of life for everybody, in part because it lowers the price of things we need and want. But that doesn’t mean people don’t get harmed along the way.

Ben Jones: So, one of the fears people have is that these automation technologies take jobs.

Jess Love: Yes.

Ben Jones: And of course, this goes back to the early years of the Industrial Revolution and the Luddites. You know, when you have handcrafts people who are silversmiths and blacksmiths, and you’re making clothes by hand, and then suddenly you have, say, the loom, you’re going to put out of work a lot of hand clothes makers.

But they’re going to migrate into other jobs, right? Over the long run, we’re going to see people do many other things, right? 

Think about the real economy. Think about farming. I don’t know if anyone in the audience has ever had a chance to ride around in a combine harvester.

But, that’s one of the most incredible pieces of machinery you can imagine. I think a single combine harvester with a bunch of trucks driving up and around it, was able to harvest something like four or five million pounds of corn in twelve hours. 

Jess Love: That’s insane. 

Ben Jones: Right? And certainly, agricultural technology is one of those areas where we’ve had massive productivity growth.

We have better seeds and irrigation systems. We have a lot more knowledge. 

Jess Love: Fertilizer. 

Ben Jones: Fertilizers… All sorts of pesticides, depending how you feel about that. And those together have made for very high productivity, meaning, one hour of labor working with those machines can produce an enormous amount of food.

And the end result of that is that instead of having half our workforce in the US, you know, be in agriculture- 

Jess Love: And that’s what it would have been, like- two hundred years ago.

Ben Jones:  And that’s what it is in many very low-income countries today, which still struggle to put in place even more basic agricultural technologies.

So, if fifty percent of the American workforce was in agriculture, now it’s, say, two percent of the American workforce working with an enormous amount of machines. But of course, that’s why many people have gone on to find other careers in other parts of the country. And overall, that process, while it’s not incidental, it can be quite painful. This process of what we call creative destruction, in the long run is what allows us to be far more productive, have higher standards of living.

Jess Love: Coming up, Ben Jones explains why he wants AI to be really good at every job it automates, and why the bigger risk is mediocre AI.

I do want to point out a very important difference that I think a lot of people are feeling, which is that what is happening with artificial intelligence right now, it’s that the most desirable jobs seem to be the ones that are most at risk. So, I don’t think many people grow up and think, “Boy, I really want to spend all day turning a knob again and again and again.

But a lot of people really do want to do things like animate, or people want to become journalists. People are accepting giant salary reductions already, to position themselves in these fields. And it seems like those jobs are also at risk. So, there does almost seem to be an emotional difference with the jobs that are being lost right now, that puts kids growing up today in a really hard spot in terms of, follow your dreams.

Ben Jones: Well, so, I mean, this is a bit too soon to tell, I think. But, you’re right that if you think of a job as not a task, but a bundle of tasks, that doesn’t mean that you don’t like some of the tasks more than others. 

Jess Love: Yes. 

Ben Jones: If you’re a chef, I presume you think about the menu and, and some of the cooking is more fun than washing the dishes, right? 

Jess Love: Yes, that’s exactly right.

Ben Jones: Maybe I’m wrong. And I don’t think people are the same. So, I think people have different preferences, even in the same job, about different kinds of tasks that they have to do. But to the extent that, you know, AI replaces, it’s just lower cost for how good it is at the tasks that you like, I think that’s not gonna feel so great.

You know, what if it’s really good at creativity and then all you’re doing is editing? But you’d rather do the creative part. 

Jess Love: Exactly. 

Ben Jones: But on the other hand, you know, it’s also the case that when you bring in… Like, let’s go back to regressions. I like to do statistical analysis. I think it’s really interesting.

But I’m mostly interested in the output. I have no interest in the calculation, and it’s boring. So, this makes me get to the interesting part much faster. And maybe when you’re writing, it’s a collaborator, right? I mean, maybe it’s not a replacement. It’s like another voice on your team.

And suddenly you’re getting a higher quality product, and it could be that you’re making a screenplay or writing an article, you’re writing a song, and suddenly it’s better for your collaboration with the AI, but they’re not actually taking kind of full control. 

Jess Love: That’s the optimistic version here.

That at least for a lot of jobs, some tasks will be automated, but other tasks won’t. And those tasks, those bottlenecks, will keep us humans busy and employed and maybe even fulfilled. And looking for these bottlenecks in our own jobs or industries, that is something that feels actionable. 

Ben Jones: Naturally, if there’s some bottleneck tasks that are slowing up the whole system, those are gonna have a lot of value.

If you can continue to do those or do them relatively well, people will value that. You’ll get a lot of income, you’ll get a sustained job, based on those kinds of tasks. So, the question I think is, are there tasks that are bottleneck tasks that computers can’t do very well yet? That’s really where you should focus.

Jess Love: And in Ben’s view, assuming there will always be some role for humans, we should actually be rooting for AI to get really, really good at whatever it automates. To understand why, we first need to understand two opposing forces. 

Ben Jones: So, the first is automation. Automation would be when a machine starts doing a task that we used to do as labor, humans.

Jess Love: Yeah, and I think this is the force that a lot of people think of. They think about, okay, this machine can do more and more and more work, and so there’s obviously going to be just, like, less and less and less of the income or the share of the economy that’s going to be available to us non-machines.

Ben Jones: Correct. That’s the sort of obvious force that most people get. But there’s another force which is a little bit more subtle, and it has to do with the fact that if machines are taking over tasks, and they are really good at those tasks, right… And at some level they have to be good at those tasks, otherwise they wouldn’t be taking them over from labor.

But if they get very, very good at those tasks, then what happens is those tasks become very cheap, in the sense that whenever you flood a market with something, right? If I made anything abundant, its price will fall. 

And so what’s gonna happen if machines are really, really good at those tasks is the kinds of services they produce will become very inexpensive.

And so, what we really want if AI is gonna take over a lot of tasks, it’s very important that it not just slightly beat labor at those tasks. We want AI, if it’s gonna take over those tasks, to be really great at it, because then we’re gonna have an abundance of that output, and actually the prices fall and people end up being much better off in a real income sense.

Jess Love: So, I was trying to come up with an example, and instead of it being, quite as theoretical, I was trying to think of something that maybe is already happening in the economy today so that we could sort of play out this idea, this thought experiment, and really understand why even us mere laborers really ought to be working, rooting for AI to get quite good at the tasks of ours that it takes over.

So, I’m gonna read an example to you, and I am hoping that you can respond to it. Give me a sense of whether this is correct, and if so, what it means. If not, where I went wrong. 

‘Imagine that companies start replacing customer service professionals at call centers with AI. This is something that we already know is happening, like when you call your health insurance company to ask about coverage, or you ask your bank why a deposit hasn’t gone through, what have you.

The person you talk to gets automated. They’re replaced by AI, and even if the AI isn’t cheap, they still cost a lot less than paying the salaries of fluent English speakers trained in customer service. So, one scenario is the AI is just ‘okay’ at this. It provides worse customer service than humans. People don’t always get their problems resolved.

Maybe they get frustrated, maybe they even get angry at the company. But the AI is still good enough that even though the company’s reputation might suffer a bit, the money they’re saving kind of makes it worth it. Maybe they come out a bit ahead. And this is your nightmare scenario as I understand it.

So, all these call center jobs get eliminated. Nobody really benefits except for the companies that might save a little money every year. Is that a fair depiction of the nightmare scenario?’

Ben Jones: Yes, because if the companies are only just saving a little bit of money, that’s a sign that the computers aren’t that much better at the task.

We get bad services. People lose their jobs. That’s an accurate version of the kind of nightmare scenario one would have in mind. 

Jess Love: Okay. Now, if the AI is really, really good at customer service, then eventually most of us do benefit. So, some people do lose their jobs, but they are able to move into others.

And meanwhile, the rest of us get our questions answered really quickly, really easily. We have more time in our day to do other things, and hopefully the products and the services from that company become less expensive because the firm is able to pass along those savings. 

Ben Jones: Correct. And as long as those workers who are displaced are able to find other kinds of jobs in this new economy, everyone’s income will eventually, be going up.

Jess Love: So, if I want to go back to this farming example, it’s very easy for me to see that, you know, in the last two hundred years, let’s say, since half of America was a farmer, it’s very easy for me to see that life is a lot better now.  You’re gonna live if you are born — you stand a much higher likelihood, you’re just going to enjoy a lot of things that our ancestors would not have been able to. But I think right now, a lot of people are feeling like life in general is not getting better. They’re kind of feeling like, the chances that my kid is going to have a better life than me are not that great.

You know, houses are getting more expensive. It’s more expensive to go to college. Like, I think a lot of people are feeling like they are just treading water. And so how does that square with this idea of incremental progress? Is it just that we are imagining the past to be better than it was? Or is it that maybe some of the things that are real pain points are because of these bottlenecks or something else?

Ben Jones: I would say that’s more to do with inequality, and where the benefits have been felt in this economy. And this goes way before artificial intelligence. 

Jess Love: Okay. 

Ben Jones: Going back to the ’70s for quite a long time, the median real household income was kind of stagnating in the United States.

And so, we had a lot of growth, but as I think people know, a lot of that increased value added in the economy every year has been captured by people who are pretty well off already. One of the concerns with AI, as we see, these billionaires, trillionaires starting to emerge, is that too much of the added value will be captured by those who own the companies or own the capital. So, I think that’s a real question. The other kind of inequality, which is actually the major kind of inequality that’s been at work in the United States, is within labor. 

Jess Love: Okay. 

Ben Jones: And what we’ve seen over time since the ’70s is a spreading where people who are in the upper educated professional classes have been capturing a lot of the growth, in income per capita for themselves. 

Jess Love: Oh, interesting.

Ben Jones: Whereas the median household has been struggling. Compared to history, we’ve seen relatively slow growth in real standards of living among the median household. I think when you think about why, Americans maybe not necessarily fully seeing these facts, but feeling them right? Feeling that maybe they put it on globalization, but that’s maybe not the right… But there’s some globalization, there’s a lot of automation, there’s a lot of technological change within the country. But feeling that somehow, these advances or this participation in the world economy, doesn’t seem like it’s making their lives better. And I think for many people, at least going back to the ’70s, there’s some truth in that.

Jess Love: In Ben’s view, much of the pessimism Americans feel about technological progress comes not from the technology itself but from how its gains have been distributed. But he thinks that at least some of this inequality, not the kind driven by tech billionaires, but the inequality within workers might actually decrease as AI automates white collar work. And to be very blunt, he does not share the average American’s concerns about AI making human labor obsolete.

Jess Love: So, I mean, a stat I ran across earlier this year, just 6% of Americans say workplace AI use will lead to more jobs in the long run, and you would just wholeheartedly disagree with that and say that, this might just be kind of a failure of imagination.

Ben Jones: Correct. I mean, you wouldn’t have imagined there was an ATM technician or a satellite engineer or a pilot before those things were created. And, I think it’s very hard to predict what’s gonna become easier, what’s gonna become hard, and that’s for the future to tell us. It’s gonna be unexpected and pretty interesting.

But there’s nothing so far in history that would suggest that suddenly all labor could be replaced. And even if it were, I mean, just to be clear, I mean, if it were, right?  

Jess Love: Yes, let’s go there! 

Ben Jones: Then the amount of output would be enormous, right? So, if you actually replaced all human labor and you had these efficient machines that could do everything, the amount of consumption you could have would be incredibly high. It would at that point become entirely an issue of equity and access to that income. 

Jess Love: Right. 

Ben Jones: And, you know, that’s where governments can step in, right? And we can have redistributive policies. I mean, do you wanna just let a few capitalists own everything? I doubt that is a political equilibrium- with everyone starving and, like, three people are, you know, quadrillionaires. Maybe, this gets into dystopian scenarios —

Jess Love: Yes, it does

Ben Jones: And it gets into various other things. And I’m not saying that’s not possible. We certainly see authoritarian behavior around the world that is problematic. It’s ultimately a question of our political process and the representation of voices who are being left out, or doing much worse because of technological change.

And, you know, if you’re making the pie much bigger through this kind of churn and technological advance, there’s more resources around. You should be able to compensate those who lost in some way and not have them live a life that is much harder for virtue of having been replaced by machines.

And I think, you know, in the United States, you have to be careful in both directions. Technological progress is inevitably about creative destruction, right? I mean, think of what Amazon has done, right? E-tail destroyed bricks-and-mortar bookstores first and a whole lot of other kinds of bricks and mortar, right?

But it is more efficient for a lot of people. Consumers like it. They get things delivered. It’s cheaper. You know, they get things faster. But if you owned a bookstore, it was terrible, right? And, so, how do we think about that as a society? Well, on one hand, if we let the people who are already kind of in charge resist creative destruction, then we don’t grow and we don’t advance, and then those people are kind of holding us back, right?

On the other hand, if you just let creative destruction go with no kind of willy-nilly, with no compensation, which is pretty close to the American model, right? You got a lotta progress. You got a lotta churn. You get a lotta technological advance, but you’re gonna displace a lotta people. And so, at that point, you definitely need a very, at minimum, a very flexible labor market which allows people to reallocate into other jobs, and there are ways to have policy to facilitate that.

Retraining, et cetera. But that’s not to sugarcoat anything here. I mean, if you’re in a rural area and the main employer goes down, reallocating to a new job probably means reallocating to a new place, leaving your family, leaving the region you know and all the culture that goes with that.

That’s an extraordinarily painful experience. And whether that’s because of technological change or trade, you know, I think it’s on all of us and as citizens in a country to take that extremely seriously.

Jess Love: I think this question of how we as a society deploy and adapt to automation is one of the most important questions today. But before we can adapt, we need to understand. That’s what this show is trying to do, help more people, including me, make sense of things. So we have a lot more great conversations coming up with economists, scientists, experts in design and media, people whose work can help us understand this moment and maybe even peek into the future.

I’m Jess Love. Life, Automated is a project of the Ryan Institute on Complexity at the Kellogg School of Management at Northwestern University. We’re distributed by KQED. Our show is produced by Jesse Dukes with help from Nathan Ray and Steven Jackson. Music by Steven Jackson, Q Shop, and Neutral Propulsion Laboratory.

Recording help from Will Feeney and George Christensen. Administrative support, recording, and wise counsel from Stacia Sliger.