Will a robot replace you? With automation growing increasingly sophisticated, it is beginning to touch all corners of the American workplace, including redoubts of white-collar stability. On the latest Pivot podcast, New York Times columnist Kevin Roose, who has written a book on the future of automation, explains that the best way to make oneself impervious to technological upheaval is to, well, not be so robotic.
Kara Swisher: Kevin Roose is a tech columnist for The New York Times. He has a new book out called Futureproof: Nine Rules for Humans in the Age of Automation. Kevin, welcome to the show.
Kevin Roose: Thank you for having me.
Swisher: Scott was talking about the future of online school, but you’re talking about automation, mostly.
Roose: Well, I think a lot of that is connected. The reason I wrote this book is because we’ve seen this huge influx of AI and automation into industry, and higher education, and journalism, and it’s changing all of our jobs and requiring us all to adapt. I was freaked out because I was looking at my own future and thinking, what can I actually do to prepare for this? I think as Scott said, I think we’ve realized now that some parts of what we do are likely to be sort of automated or taken online or disaggregated in some way, and so it’s up to us to figure out how to deal with that.
Swisher: Name some of the rules.
Roose: Well, there are nine rules in the book.
Swisher: Not ten? Interesting …
Roose: I want it to be more efficient than the ten commandments. But, yeah, there are basically three buckets of rules, and I won’t list them all because that’ll take forever, but basically there, there are three buckets. One is how to futureproof your life at home, your brain, your personal life, your family life. Then there’s how to futureproof your career, and how to futureproof your community.
Swisher: All right, give us one from each.
Roose:The one that is for your own self is that I think you need to basically find things to do that are not going to be replaced by machines. I think we’ve been training people for the future entirely wrong. We’ve been teaching them to become more machinelike — to major in STEM, to become super efficient, to optimize and life hack their way to success. And I think we really need to focus on the more human skills that machines can’t replace.
Swisher: Creative ones.
Roose: Yeah. Creative ones, compassion. I have three categories in the book of work that I think is unlikely to be automated soon: “surprising,” “social,” and “scarce.” Those are the types of work.
Roose: One thing that I think you both have done very well is what I call “leave hand prints,” which is to make yourself less of a sort of cog in a machine. To make it clear that you are a human creating human work. One of the things I got from interviewing AI experts and economists from this book, is that in the future, things that are done by machines will become very cheap and things that are done by humans will become more valuable.
Swisher: Artisanal. We leave our dirty hand prints. You know that, that’s our thing.
Roose: There’s a lot of literature showing that we actually value things more when we think that people had a real hand in creating them, and so that’s something that people can do for their careers, and for your community. I think what we teach children and what we teach ourselves, as adults, is really important here. I think that we need to kind of update the curriculum that we’ve been teaching people now for a hundred years, which is all about making people into effective workers in an economy that rewards things like output, and productivity, and efficiency — rather than the more human traits that I think people are going to end up moving to as a result of AI automation.
Galloway: I love that: Leave hand prints. First off, and this is the most important question, although it’s a bit of a digression. Do you meditate?
Roose: I try, yeah. I don’t always succeed, but I try.
Galloway: Because I can just hear it in your voice — you have nice chi and center. We need to roll. We need to hang-
Galloway: Because my testosterone therapy has totally made me jumpy, so we’ll talk after the show, but anyways, kudos to you. I can tell you’re a centered person. I love that — hands-on. Let me ask you the following. My sense is we need a new approach where capital doesn’t see labor that doesn’t have a double leaf from MIT as a cost and a negative — as opposed to looking at people and trying to figure out how automation enhances human capital instead of replaces it. Don’t we need a different mindset around robotics? Isn’t the problem us, not the robots?
Roose: Absolutely. Yeah. I mean, one of the key lessons of the book is that robots don’t do anything on their own right now. I mean, people say robots are coming for the jobs, but it’s really the executives at the Fortune 500 companies who are saying, “I want to shrink the accounting department by 200 heads, or I need to squeeze out some more margin in next quarter’s numbers, so I’m going to automate these jobs away.” It’s a really simplistic, really substitutive kind of automation that we’re seeing a lot now. As I put it in a story yesterday for the Times, it’s about replacing fill in accounting, rather than becoming a market leader, doing new dynamic things, developing new products. The economists call this so-so automation. It’s the kind that kind of sucks. It’s like the automated customer service line that you’re like, “I just want to press zero and get to a human. Please put me through to a human.” That’s the kind of automation we’re seeing a lot in the corporate world right now, and that’s the part that’s really dangerous, because that’s not actually adding to human capability. That’s not empowering workers. That’s not developing new tools that are going to move the economy forward. It’s purely substituting a machine for a human.
Galloway: It was Intel that drove the computing revolution, you could argue. What are the technologies that are driving robotics right now?
Roose: Well, machine learning is the big one. That’s the thing that is transforming automation from something that can do rote and repetitive work to something that can do more kinds of creative, sort of cognitive work. People, I think, still conceive of AI and automation as this thing that’s going to replace people in factories. The thing is that the factories are largely automated already. That happened a long time ago. Now, the robots are doing more kind of managerial tasks. There’s some great research out from the Brookings Institution in Stanford about the fact that actually the people at most risk of replacement from AI are white collar professionals in big metro areas who are doing sales projections and data analytics, and those kinds of sort of cognitive tasks that are done by people with college degrees who make a lot of money. That’s not safe the way we thought it was.
Swisher: The debate right now is over the $15 minimum wage, and what happens to those jobs. Do you feel that they’re at risk?
Roose: Certainly. I mean, retail is a big target of companies doing automation. You know, we’ve already started to see some of that happening with the self-serve kiosks at fast food restaurants. I think what worries me is less the sort of threat of displacement for those workers, because I think some of that stuff is going to end up being surprisingly hard to automate. It’s more using automation and AI to turn workers into human robots, essentially. This what we see.
Swisher: It’s push-button, like at Amazon warehouses.
Roose: Yeah. Exactly. If you work at an Amazon warehouse, you are taking instructions from one algorithm, you’re putting things into a box, you’re wearing a bracelet that tracks your productivity. You can be fired if you miss your packing target. I mean, it’s essentially these jobs are kind of human robots, and I think that that’s one of the cautionary tales.
Swisher: And they have the arms to do it right now until they figure out arms that can do it better.
Roose: Exactly. Those people should be concerned. But there are also ways that this can improve workers’ lives. People can be freed up from mundane sort of bullshit tasks, and we’re just not seeing enough of that right now. We’re seeing, “I want to replace 10 people in accounts payable, so I’m going to buy this off the rack AI solution that can do that for me.”
Swisher: Should we work at all? I mean, this idea of … There’s all those movies where we just sit in chairs and eat, and then robots take care of you. Do you predict that’s what’s going to happen with automation? Is there any need for humans?
Roose: There’s definitely a need for humans. I’m not full dystopian, but I think that our jobs are going to change a lot. I think that right now, a lot of people are employed in jobs that involve making things, and I think that the future of jobs is going to be jobs that involve making people feel things. It’s going to be the kind of things that bring about human connection. It’s not going to be enough to be a really good radiologist. You’re going to have to have a good bedside manner too, so that people have a reason to come to you rather than going to an AI.
Galloway: We were just talking about education. Do you see any thoughts around robotics as it relates to education, and also health care?
Roose: Yeah. I mean, these are the areas that have been sort of resistant to automation. We don’t have robots teaching college classes, most of them.
Roose: Yet. We don’t have many robot doctors yet, but that’s coming. I think we need to stop educating people and telling them that they need to take on the machines head on, to sort of compete with them like the old John Henry thing. We need to start telling people what they can do that is not going to be disrupted by these forces, and that’s a journey that I’ve been on too. I mean, I’m a journalist at a newspaper — that’s not the most futureproof job in the world.
Swisher: But wait, wait. Why not? How are you getting replaced?
Roose: Well, for example, in one of my first jobs in journalism, I covered finance and I wrote corporate earnings stories.
Swisher: Yeah. Those were stupid, though.
Roose: “Alcoa made this much money last quarter in their smelting division,” that kind of thing. And that’s largely been automated in just the last 10 years. Those jobs are few and far between now. I think there’s another way in which automation has changed our industry, which is something that we don’t talk about as much. We’ve replaced these people who we used to call editors and ad salespeople with algorithms, run by Google and Facebook, that now choose what to show people and sell the ads around it. So there’s been a lot of automation happening in our industry. It just hasn’t been happening at the news organizations, it’s been happening in Silicon Valley.
Galloway: The notion that any activity that’s rote — that’s the opposite of artisanship, right? I mean, what you do, and I think, most of the journalists other than Kara Swisher at the New York Times do, it’s artisanal. They actually come up with original ideas that no AI for decades is going to be able to do. So how do you prepare a new generation of people coming up and maybe the ones that don’t have access to college — how do you encourage it, or how does our education system instill artisanship?
Roose: Well, I think it has to start from instilling a sense that this stuff is valuable, that you’re not going to be unemployable if you’re a musician, or a philosopher, or a sociologist. I mean, there was this really strong thread coming out of Silicon Valley for many years about the fact that STEM education was all that mattered.
I think Marc Andreessen said, “English majors are going to end up working in shoe stores,” and stuff like that, and so we really defunded and devalued those humanities programs. I think we’re starting to see the results of that. I had a tech CEO tell me recently, “I can hire tons of engineers, but I’m having a lot of trouble hiring salespeople, because no one in the Bay Area, has the people skills to be able to go to a company and sell them software.” That is a real missing piece in the economy today: the people with those kinds of empathetic human communication skills. And I think we need to start reorienting the curriculum that we teach to kids around that.
Swisher: People will be health-care workers, they’ll be social workers. What else?
Roose: Well, I think the sort of categories that are sort of most automation-resistant right now are in things like health care. But I actually think that the job-based sort of taxonomy, where one job is going to be totally safe from automation, and one job isn’t — I think that’s the wrong way to look at it. I think it’s about the way you do that job. Are you doing it in a rote and repetitive way, or are you doing it in a creative and human way? That’s what the book is trying to sort of guide people to. Even if your job is not being an artist, even if you’re an accountant, there are ways to do that job that are more human and less easily automated, and so pushing people toward those parts of their jobs is one of the goals here.
Swisher: But when do computers get the ability to do that? Or are they on that?
Roose: You’re already starting to see sort of AI do what I’m calling emotional work, like analyzing emotions. I mean, they’ve implemented bots that are sort of therapy bots for people who need some sort of assistance there. In education, you’re seeing AI being used to personalize learning for people. But I think this sort of job-based framework is not the right one. I think we need to be thinking less about what you’re supposed to be doing to make yourself futureproof and more about how you’re supposed to be doing it.
Swisher: Right. But there are certain jobs that really have a big red target on them, for sure.
Roose: Sure. But it’s not to say that those jobs are going to go extinct. I mean, even though the corporate earnings reports journalists are not doing so well these days, people who are creative and flexible and very human are finding ways to make money off of it. So, I think that’s where the value is going, is to the things that are sort of deeply human and away from the things that are more mechanistic and based on sort of productivity and output.
Galloway: Whenever you write a book, you go in with a set of kind of predetermined thesis around what the book is going to be about. You think, Okay, I want to write about X, Y, and Z, and then you do the research. You write the book, and typically you discover a couple of things that sort of change your view. After writing this book, what changed in your view of automation and robotics?
Roose: I started off as basically an optimist about AI and automation, and I still am, although I’ve tempered it somewhat. I now call myself a sub-optimist, because I think this technology can be amazing if we do it right and thoughtfully. It could free us from our worst tasks. It could solve world hunger, and climate change, and cure diseases. The technology is not the problem. It’s the people who are implementing the technology right now that’s the problem, so I think what’s changed in my view of AI and automation, is that it’s not a guarantee that all this is going to improve people’s lives. We actually have to work hard to make that happen. People need to prepare themselves. It’s not enough to attend a coding boot camp anymore. You really have to work on yourself so that you’re ready, and you’re not as replaceable as you might be right now.
Swisher: One of the things you do write about, and I love your comment at the end of this, you write a lot about misinformation, and you’ve sort of looked at the past year or two. These hearings are coming up Capitol Hill around the impact of social media networks on the attack on The Capitol. What is your assessment on that right now, given you’re a sub optimist?
Roose: I think the misinformation conversation is related to the automation conversation. I think a lot of what we talk about when we talk about automation is kind of external automation, like robots coming into factories and stuff like that, but there’s a kind of internal automation that’s happening to a lot of people, that is really worrisome, and I include myself in that. Every day, we wake up and we look at our phones and we are just fed algorithmic feeds and machine generated recommendations, and our phones and our devices are telling us what to think about.
Misinformation has always been present, but now it’s sort of hyper present because of these algorithms and the platforms that use them. So, I think it’s not only enough to get the right job to future-proof your career. It’s also about detaching yourself from some of the technologies that exist to sway your opinion, to persuade you of something that maybe you don’t want to be persuaded of, and to ultimately sort of confuse you about who you are. I mean, you have to know yourself better than the algorithms know yourself or else you’re going to be replaced.
Swisher: How do you look at these companies now, how they influence us?
Roose: Well, I think they exert tremendous influence, and I think they don’t want to admit it, but they d. . That’s another thing that I think I’ve become much more skeptical of, is I think that AI is never going to make for an effective sort of editing function. What we’ve seen at Facebook — they’ve been promising for years, we’re going to automate content moderation, AI is going to take over, it’s going to filter out all the hate speech and all the misinformation. I think what they’ve realized, belatedly, is that humans are just better at that. So, you have to bring in a bunch of people, train them and put them to work in content moderation divisions, and that’s sort of a mistake of over automation and underinvestment in human potential.
Swisher:On that note, Kevin, your book is called Futureproof. It’s an excellent book. I’ve read it. Thank you so much.
Galloway: Let’s roll, Kevin.
Swisher: Don’t roll with him.
Pivot is produced by Rebecca Sananes.
This transcript has been edited for length and clarity.