Talking To Teens: Expert Tips for Parenting Teenagers

Andy:
So the book is Futureproof: 9 Rules for Humans in the Age of Automation. And it's about your journey to understand what is the future going to hold, and is AI going to destroy humanity or is it going to be a beautiful rosy existence and we live hand in hand with technology, or where are we going to end up? You have been writing about technology related topics for a long time now, is that how you got interested in AI? Or where did this come from?

Kevin:
Yeah. So I started looking into AI and automation seriously maybe three or four years ago. And at the time, this was what everyone in Silicon Valley was talking about. Giant companies were spending billions of dollars developing new AI. People were splitting into two camps, this optimists who thought AI was going to make the world better and transform everything and we would become enlightened humans who just spent all our time making art and playing video games and stuff because the robots would be doing everything for us.

Andy:
Right.

Kevin:
And then there were pessimists who thought AI is going to kill all the jobs and we'll be obsolete and we'll just become the slaves to the robots. And I was just interested in figuring out which of those was true, or if the truth was somewhere in the middle. So I started talking to researchers, going to meetings, trying to learn as much as I could about what was actually happening in AI. And also to, more importantly, figure out what we should do about it. Because that was the piece that seemed to be missing from the discussion was, what do we actually do about this on a human response level? We need to adapt to whatever's happening, and so I wanted to try to help people do that.

Andy:
And so did you find an answer or are you more confused after doing all this research? Or where did you end up?

Kevin:
Well, I found a lot of answers. And the second half of the book is nine rules that I think will help people adapt and adjust to the age of AI and automation. But the biggest thing that I found was that we've been preparing people for this future all wrong. For many years, we've been telling people, kids, adults, students, that in order to be prepared for the future, we needed to be more machine-like ourselves. We needed to study engineering or computer science. We needed to work super, super hard all the time and grind and hustle and optimize our productivity and life hack all of the waste and inefficiency out of our daily routines.

Kevin:
And what experts in this field told me is essentially the opposite. That we need to become better at the uniquely human things that only we can do. And that instead of trying to train people to be essentially competing with machines, we need to figure out our own niches and figure out what we can do that is going to be very hard to automate.

Andy:
There's one theory that I've heard a number of times, which is that humans and AI will end up collaborating and work together in teams. And that when the computer and a person play chess together, they are actually even better than when either one plays on their own. So there's maybe a happy symbiosis that happens between us. Does that seem possible?

Kevin:
I thought that, and I love that story of the humans and computers playing chess together-

Andy:
That seems like such a happy picture…

Kevin:
Oh, it's amazing. It's like, we're all going to live together in perfect harmony and life will get better and robots will be our friends. Unfortunately, it's not true. I mean, it was true for a while, humans and computers could beat computers alone at chess. But within the last decade, that's become totally obsolete. I mean, now computers just wall up us every time and even wall up us when humans are paired with computers. And actually, that's happening in many, many domains, not just chess. There've been some studies that have found that in a lot of areas, whether it's making decisions about who should get a loan and who shouldn't, or predicting certain types of risk, modeling certain types of outcomes, AI is better without us. We are essentially the dead weight in those relationships.

Andy:
It's like the two people riding a two person bicycle, and one is not pedaling at all, just slowing the whole thing down.

Kevin:
Exactly. And the person who's not pedaling thinks, wow, this is great, such teamwork, this is amazing. And in the meantime, the AI is just carrying us around. So it's, I'm being a little flipped, I think there are some areas where humans can still add value, especially when the AIs encounter unfamiliar information. I mean, humans are still much better than computers at coping with unexpected setbacks, learning from new and novel situations, what is called zero-shot learning in computer science. Taking something that you've never encountered before and figuring out what to do. That's something that humans are still much better at than AI. So humans are not totally obsolete.

Andy:
Okay. So they did a survey, a Gallup survey, and they surveyed people and found out that everybody seems to think that, at the same time, AI is super powerful, it's going to really take over a lot of people's jobs in the future, but not my job because I do something that just could never be taken over by computers. Does that just mean we're all probably fooling ourselves, or how is it that everybody has convinced themselves that they do something that won't be able to be taken over?

Kevin:
Yeah. This is one of the most amazing things I've found while researching the book, is that we are very overconfident in our own irreplaceability. And I include myself in this too. But if the survey you mentioned, I think found that three quarters of Americans think that AI is going to destroy more jobs than it creates, but only one quarter think that AI is going to come for their job. And this is true throughout history. I mean, one of my favorite articles that I found when I was researching this book was from 1984. There was a story in the New York Times about these new ticket machines that were appearing at airports, where people could just, instead of talking to a travel agent, you could just go up to the machine and buy a plane ticket using your credit card. And they interviewed some ... the reporter who wrote the story interviewed some travel agents and they were incredulous.

Kevin:
They said, no way will people ever order a plane ticket through a machine. I mean, can you even imagine that? I mean, who would trust a machine to book a plane ticket for them? It's much too important. And now obviously, we all use the machines to buy our plane tickets. And travel agents still exist, but there are many fewer of them. And so I think we make this mistake again and again, of thinking, well, automation is someone else's problem, this is never going to happen to me. And I see that attitude among a lot of my colleagues in journalism right now, who say, oh, there's no way a machine could ever do what I do. And meanwhile, startups are working on automating their jobs right now.

Andy:
Yeah. So does that just mean that no jobs are safe?

Kevin:
Well, it means that no job is automatically safe. No job is, by virtue of the job title, totally immune from automation. Every job can be made more safe. And that's what the second half of the book is about, is I think there are some jobs that are more human than others and that have more human elements to them than others. So if you're a factory worker or a truck driver or a retail cashier, those jobs are generally easier to automate than jobs like artists and musicians and journalists. But even the artists and musicians and journalists contain a range. I mean, my first job in journalism was writing corporate earning stories. So every quarter a company released their earnings and I would sit there and look at the press release and go, okay, Alcoa made $9 million in its smelting division last quarter. And I would just write that out.

Kevin:
And now those jobs have been automated. Software is being used to generate those stories in a lot of major news organizations. So every job contains very robotic tasks and very human tasks. And what we need to do is figure out which are which, and do more of the human things, because the rest is going to be going away.

Andy:
So is it the same thing like in college, or if you're trying to choose what to major in, is it less about what topic you're studying and more about finding those types of things within whatever field it is that are not easy to automate?

Kevin:
Yeah. I get asked this question a lot by parents who say, what should my kid be studying? Or even I get it from students who say, what should I be studying? If it's not computer science, what is it? And my general take on this is that skills and traits make you more resilient than subject matter knowledge. So if you are a computer programmer, that's a job that is being automated in some cases. They are not safe from automation, even though they are the ones creating a lot of the automation.

Kevin:
And so it matters if you have human skills to layer on top of those technical skills. If you're a computer programmer, you need to be good at communicating with people who aren't as technical. You need to be able to be good at coming up with new and creative solutions to problems. There are these skills that even if the programming language or the subject matter becomes obsolete, you have these transferable human skills that can help you make the jump to another profession. So that's what I'd focus on if I were a student today. It's not necessarily what major I had or what classes I took, but how do I build these skills, these deeply human skills that are going to be helpful in making me hard to replace no matter what I end up doing as my career.

Andy:
And one of those things you talked about briefly earlier is dealing with unexpected things or things that don't match the previous patterns, the surprising things. And you talk about a study in here or 2018 experiment, they're trying to teach AI to recognize objects in a room and they teach it how to recognize all of these different objects in the living room, which it gets really good at. But then once they put an elephant into the living room, it totally starts to confuse the software to where it just can't figure out what's going on. What does that tell us, I guess, about how it works and what things it's not good at doing?

Kevin:
Right. Well, this is an example of AI's limitations. Right now, AI likes things that are really structured and regular and repetitive. And so things like chess, where if you're an AI trying to learn how to play chess, you can play 10 million games against yourself, getting a little bit better each time, and eventually you'll beat even the best humans. But a lot of things in life are not that repetitive or that iterative. We don't have 10 million chances to get something right. So I think that's where humans have the advantage, is these sort of situations and jobs in which there is a lot of change and a lot of chaos and very few regular predictable events. Which is why an AI could beat a human in chess, but it would be very bad if you asked it to teach a kindergarten class, because that's not a job that involves a lot of regularity of any kind.

Kevin:
So I think that's the kind of thing that I'm trying to shift the conversation around AI and automation, from what jobs should I get to, how should I do that job, what kind of traits and skills are important no matter what I'm doing.

Andy:
And so one of those that you talk about that I found really interesting is combinatorial creativity. What the heck is that, and why is it so important and why is it uniquely human?

Kevin:
Yeah. So combinatorial creativity is a word that Maria Popova of Brain Pickings, I first heard that phrase from her. And it's basically this idea that we take things from one area of our lives and combine it with something else and come up with some new ideas. So there are examples of this through history, like famous inventors who play musical instruments and have big breakthroughs while they're playing musical instruments, or-

Andy:
When an apple falls on their head.

Kevin:
Exactly, when an apple falls on their head. But I mean, humans are very good at this. We take something that we learned 10 years ago in a totally different context, and we apply it to a problem we're trying to solve now.

Kevin:
But machines are not very good at doing that right. There's this thing called transfer learning, which is basically where you're building an algorithm to do one thing and then trying to have it learn to do something totally different. And it's not very good. So we still have the advantage when it comes to that combinatorial type of expertise, these rare combinations of skills, these things that we take from one area of our lives and apply somewhere else. That's something that humans are very good at that AI is not very good at.

Andy:
So then is there something that we should be doing to get better at that, or to use that to our advantage somehow, to do our jobs and live in a better way?

Kevin:
Yeah. Well, I think we can embrace the diversity of experiences that give rise to that combinational creativity. So one thing I tell writers who ask me for advice, people who are journalism students who are coming up in the writing world is, is develop multiple niches. So one of my favorite writers is James Fallows at the Atlantic. He's also a pilot in his spare time. He flies his own planes. And that expertise you would think is like, what does that have to do with being a writer? But it has actually made him an extraordinary storyteller. It's given him the ability to ... when things like a big plane crash happens and it's all over the news, he's able to actually understand the intricacies of what it means to fly a plane and explain how that happened.

Kevin:
Just with this big container ship that was stuck in the Suez Canal for the last week or so, there are a couple of journalists who are really maritime shipping experts. And so that expertise has become very valuable all of a sudden. And so I think building up these different skills in different parts of our lives that we can combine together when the opportunity strikes is really important.

Andy:
Yeah. Because then you'll just ... situations will come along where you can see how there's overlap in different skill sets that you have that can be used in a new way. But it's like you need to have lots of those different perspectives and skill sets to draw from, and you also have to be in a lot of different situations.

Kevin:
Mm-hmm. Yeah, absolutely. And I think that we have taught people about the value of being specialists and I think there's certainly something to that, that expertise is good and people who have it are generally well-positioned. But there's also the having multi-specialties, having different things that you can go deep on, I think is really an underrated skill.