Support for this podcast comes from Canva. When you look good, you feel good. But when your presentations look great, it can feel like you're walking on a cloud. You can design stunning work presentations, docs, whiteboards, and videos with Canva. Start with a designer-made template. Use it as a springboard for your design. Add images, graphics, charts, and more from Canva's massive media library. Start designing today at canva. Com. Design for work. Episode 296. The Ferrari 296 is a sports car known as the first real Ferrari with just six cylinders on it. In 1996, the Fox News channel launched. True story. My son goes to sleep with one of those white noise machines. It has three settings: low, medium, and Fox News.

That's good. Go, go, go.

Welcome to the 296th episode of the Prop G-Pod. In today's episode, we continue our special three-part series covering the future of work. In last week's episode, we answered your questions and offered some valuable advice. I didn't really see my kids much before the age of three or four. I was just working all the goddamn time, but either did my wife. She was working at Goldman Sachs at the time. We decided we made the conscious decision to trade off time with them then so we could have more time with them now. What is your niche? Start writing about it, start posting about it, start doing videos, whatever it is. You're in PR, you understand this shit. But start building a brand and a profile for yourself across all of these different platforms. Before you collect dogs and kids, oh, my God, get back into the office. Friends, mentors, potential. One of three relationships begin at work, for God's sakes. Where are young people supposed to meet partners, wives, other founders, great friends? It's about hitting the pub after work. It's about grabbing a coffee and getting advice from someone. Nothing wonderful is ever going to happen to you over Zoom, in my view.

Today, we're covering productivity, the importance of establishing work-life balance, and the disruption of AI on work and higher Ed with Cal Newport, a professor of computer science at Georgetown University and author of a number of books, including Slow Productivity. Cal is like a rock star. He's one of those guys I look up to, mostly because he sells more books than me. Don't know him, but he sells a shit ton of books. He's one of these academics that has found the intersection between retail success and academic credibility. I've got some of the former and none of the latter. Anyways, we'll get to that after our first break. But first, what's happening? The dog is on his way to Canada. That's right. Little dog, little maple syrup with that dog. How do you get 100 drunk Canadian fraternity brothers out of your pool? Hey, guys, could you please out of the pool. That's good. Anyways, going to Canada. I love Canada. I was conceived in Toronto. My dad, every year, gets something off his bucket list as he's old and will be dead soon. Sorry, dad. It's turning 94 this August. Can you get over that? 94?

By the way, I saw William Shatner on Bill Maher. That guy's 93. Jesus Christ, give the guy a fucking medal in a spaceship. That guy is sharp and funny at 93. And let me tell you, that is not easy. That is an unbelievable accomplishment. And congratulations and kudos to the Bill Maher show for bringing on a 93-year-old. Anyways, I thought it was really inspiring. So headed to Toronto. Wait, that's not true. I'm headed to Vancouver. Jesus Christ. Speaking of 93, headed to Vancouver. I was talking about Canada, Where am I? Okay, it was over here, Canada. My dad, every year, gets something off his bucket list, and he chooses the same thing every year. He wants to go see the Leifs, play of the Haves, and the opening night at the Air Canada Center in Toronto. And I have never seen my dad cry, except when the Scottish bag When the bagpipes come out on the ice at the opening of the game, he starts to cry. I have no idea why there are bagpipers in Toronto at the least game. Finally, my dad cries because he just can't handle bagpipers on ice. Well, okay, dad.

Anyways, and then he takes me to this old apartment complex and points up, and we've done this four or five times, and he keeps forgetting that I've actually seen it. He points to an apartment and he says, That's where you were conceived, which is both awkward and weird. Anyways, my parents were met. They were both immigrants, to Canada from Glasgow and London, respectively, met in Toronto, where my mom was a court reporter. She used to write using shorthand the text or what was going on in a courtroom. Can you imagine that? But it was a well-paying job. Canada embraced my parents with open arms. My dad was, I think, a candle salesman, and they couldn't handle the weather. So they read an article that said that the place in North America, the city of North America, had the best weather in all of North America. So they decided, this was their analysis, they would load up my parents Austin Mini Metro and my seven-month pregnant mom in drive for nine days to San Diego. That's why I was born in San Diego, because it's supposed Actually, according to the Toronto Globe and Mail, has the best weather in North America.

Anyway, it's a little background. But I am not headed there. I'm headed to Vancouver to speak at Ted, which I joke, is the widest place on Earth. And I've always made fun of it, but of course, they've asked me to speak, and I said, Of course, I'll speak. And I said, of course I'll speak, and I'm excited to go there. I really like the founder, Chris Anderson. I think he's a thoughtful, funny, interesting guy. I think he's actually been a great steward of the brand. I'm excited to go. My buddy Greg Shove is going to be there who is Canadian, which means he's very nice and very thoughtful and very handsome. I'm excited about it. Anyways, I will report live from Vancouver. Okay, moving on. Speaking of Ted and speaking of speaking, what is the one skill, the one skill, and I was asked this the other day, that I would want my kids to develop in high school? I was asked question about curriculum development or curriculum change in high schools. I said, if there was one skill, it wouldn't be computer programming or foreign languages or stem even. It would be storytelling. Why?

Why is storytelling so important? Our competitive The majority of advantage as a species is essentially communication and cooperation. That cuts a swath right down the middle of storytelling. That is, the majority of animals, they do mirror their parents, if you will, or the the herd or whatever you want or the school. But a superpower is to be able to communicate outside or beyond just instinct or beyond just visible cues and say, Okay, this is where you plant the crops. You don't go over the hill. Our ability to tell stories that could communicate to people outside of our immediate tribe was our superpower. We could paint these stories on cave walls and said, Okay, this is when you plant the crops or harvest the crops based on when the sun, the location of the sun in the sky, or you do not go over the hill because they're mean over there and they will kill us. That ability to communicate and tell stories via captivating stories that captured our attention around the campfire or through song or through poetry or through art was incredibly important to our ability to advance much faster and become the apex predator/species across all species.

Let's be clear, to an unhealthy extent, we rule the world. We have this crazy stat that humans and their pets now have greater mass than all mammals in the world. Isn't that nuts? Whenever I hear these stats and also having a big, big dog, a bit of a digression, having a big, big dog turns you off meat because I have a great day, and I just can't imagine that she is that different than a cow, and she is so lovely and feeling. Anyways, I'm going to try and eat more salad. Don't know where I was going with that. Okay, storytelling. Your ability to tell a story is incredibly important. As a result, Mick Jagger gets to bang a woman 50 years younger than him. What's the connection? Mick Jagger is an incredibly powerful storyteller, and we have learned instinctively that storytelling is key to the species's success. So your ability to tell stories will not only garner you more influence, more money, but also be more attractive to mates, specifically for men. And yes, this is sexist. Okay, I have evidence showing that people born as one gender predisposed to certain strengths over others. And by the way, I don't think men are anymore predisposed to storytelling.

Is that true? I don't know. Anyways. But women, basically, men get turned on with their eyes. Women are unfairly objectified and evaluated based on their looks. By the way, everyone is giving Jeff Bezos and Lauren Sanchez so much shit for essentially how she decided to dress to the state dinner and making comments about her looks. She's 54, for God's sakes. Personally, I think she looks great. I like it when women flaunt their hotness or their sexiness over the age 25. It feels as if the world is saying, Oh, you can only be sexy if you're under the age of 25. Fuck that. Live out loud. I think there's absolutely nothing wrong with that. And by the way, men are not objectified or evaluated on their looks. Just a pro tip to all the men, specifically the young men out there, there are only two things, only two things you should ever say regarding a woman's appearance. One, she looks great, or two, nothing, nothing. By By the way, when you're on a date or when you see someone, it's always a good idea to kick off the date by saying, You look really nice or you look great.

You might find that sexist or objectifying. It's not. Women are unfairly evaluated based on their looks, and men are unfairly evaluated based on their economic success. But you would never say to a man, Oh, I get the sense you're not going to be very economically successful, not be a good protector or provider. You would never say that because that cuts to a man's being or purpose. Yet we feel entirely comfortable He's disparaging a woman's looks. What the fuck is up with that? Anyways, storytelling. Back to Mick Jagger. He gets to communicate really well. Because he communicates really well, he gets a lot of money and has a much broader selection set of mates than he deserves. What would be the skill that you would want to develop or want your kids to develop? Absolutely storytelling. It doesn't have to be in any one medium. I started out presenting. That was how I made my living, was speaking in front of corporations as a consultant. Then I found writing. Then I got decent at creating PowerPoint presentations, or basically, I hired people that could take my words and put them into PowerPoint presentations. I've always overinvested in design because I realized the importance of creativity in storytelling.

These things are now podcasts. If I could give my boys Any competence or skill, it would be storytelling. What does that mean? It means give your kids opportunities to speak, give them opportunities or force them to write and to read. If they want to download Firefly from Adobe and get into AI in terms of visuals and video, by all means, embrace that stuff. Let them tell stories, ask them questions. I love that at Gulfstream School, their eighth-grade project is they have to do a bunch of research and then speak in front of the whole school. Is that nerve-wracking? Is it upsetting? Yeah, it is, and it's important. It's important because your ability to communicate your ideas are really more important than your ideas. This is who you don't want to be. You don't want to be the guy or the gal in the back of the room speaking under your breath. It's always right, but it doesn't move people to action because you don't know how to tell stories. Your ability to move people to action through creativity, through art, through inflection, through presence, through the way you dress, through your humor, is not only key to your professional success, it's key to your personal success.

Your ability to make people feel good, your ability to articulate ideas, to show your commitment, to move people to action. This is the key. Tell stories. We'll be right back for our conversation with Cal Newport. When your work presentations and docs look good, you look good. You can design stunning work presentations, docs, whiteboards, and videos with Canva. You can start with a designer-made template, then use that as a springboard for your design. Add images, graphics charts, and more from Canva's massive media library. Or get a huge head start with AI-powered Canva presentations and docs. Just describe what you want with a few words, and Canva will generate amazing slides and text in seconds. It's AI that anybody can use, no matter what department you work in or whatever work task you need to get done. Look, we all need to visually communicate at work. Canva makes it easy to get your point across while looking professional. And at the end of it all, that stunning Canva presentation is going to make you look good. Wow any audience and finish your work faster. Start designing today at canva. Com. Design for work. Welcome back. Here's our conversation with Cal Newport, a professor of computer science at Georgetown University and author of Slow Productivity: The Lost Art of Accomplishment Without Burnout.

Cal, where does this podcast find you?

I'm in Tacoma Park. It's a small town outside Washington, DC. I'm in my studio.

Good for you. Let's buy straight into it. In your new book, Slow Productivity: The Lost Art of Accomplishment Without Burnout, you share your philosophy for creating great work at a sustainable pace. What do you mean by slow productivity?

Well, it's meant to be an alternative to what we're currently implicitly using as our understanding of productivity and knowledge work, which is what I call pseudo-productivity. I have this argument in the first half of the book that there was actually a large issue, an economics issue, when knowledge work emerged as a major sector because the notions we had for measuring productivity that were developed in agriculture and then carried over to industrial manufacturing, those notions didn't apply anymore. We couldn't count Model T's per paid labor hour input. We didn't have clearly defined production systems that we could tweak and see their impact. I argue, we fell back on a heuristic that I call pseudo productivity, which says we'll use visible activity as a crude proxy for useful effort. I want to see you work, more works better than less. You want to be more productive, stay late, come early. Slow productivity is an alternative to that because my argument is that might have been okay for a while, but it really began to falter when combined with the front office IT revolution that picked up steam in the late '90s, pseudo productivity became less sustainable. So slow productivity is an alternative that's based more on outcomes than activity and is more grounded in how humans actually operate, what's actually going to be sustainable for humans.

I struggle with this or I worry about this with my son, and that is he gets anxious about grades. He says, I'm going to study all week in. And I ask him, I'm like, We'll do this. Get some real studying done for a couple of hours and then take the rest of the day off. Aren't you basically arguing that FaceTime does not necessarily translate our effort or effort, or anxiety doesn't necessarily translate to productivity?

Yeah, I think that's true. And in fact, that specific example with your son I have a lot of grounding in. When I got started in writing, I wrote my first books when I was in my young 20s. They were books for college students about how to study. And the big idea idea of those books is most college students are really bad at study. They use studying as an ambiguous verb. It's, I'm just doing stuff. I'm in the library. I'm stressed. They hope to alchemize the effort and distress of this vague type of studying the better grades. I wrote this book that said, You know what? The best students, they have really specific ways they study that works really well. It takes them a few hours. They nail the test. They move on with their lives. I was actually grounded in that idea from my days as a student writing for students.

And then bridging it to another a personal experience in the work world. My first company was a company called Profit, and we couldn't attract the quality or caliber of consultants for a small firm started by a 26-year-old that we needed to get Fortune 500 clients. So I tapped into this untapped vein of the labor force, and that was recent mothers who felt like they couldn't commit to being in the office nine hours a day, five days a week. And if you gave them a little bit of flexibility, they would take the job. And so I didn't care if they worked from home as long as they were in a couple of times a week. And I found, I immediately registered right away, they were the most productive employees because come hell or high water, they had to be on the train back home at a certain point. Yeah.

Well, I think this is one of the negative externalities of activity-based productivity is you take these types of options off the table. So if it's, look, activity is what we look for, more activities, better than less, you're basically biasing towards 24-year-olds right out of college with no families and nothing else to do. Because if activity is what we value, you're narrowing that lens. It's one of the advantages you get when you begin to value more things like expertise, like experience, like leadership, like know-how, like a long track record of things you've produced over time that have been very valuable. It allows a lot of different people into the workforce, a lot of different configurations of work into the workforce as well. In a pseudo productivity regime, we look with suspicion when someone wants to work, let's say, flexible hours, or they want to work less than full-time, or they to take a leave because of things happening within their family situation. In a pseudo productivity regime, that's suspicious. Well, that's just producing less value. I'm therefore going to devalue you and your contribution. In more of a slower productivity regime, a lot of that becomes a lot more acceptable.

What we care about is What you can do? What have you done? What skills you have to do something cool in the future and interesting in the future? I think this is absolutely right. When we focus on activity, we really narrow what work can really mean, what successful work can really mean.

You have three principles of slow productivity. Break them down for us.

Do fewer things. This is really an argument about workload management. I think it's one of the biggest unforced errors in modern knowledge work is that we allow individuals workloads to spiral out of control. This doesn't make people more productive in any objective sense of the term because everything you agree to brings with it administrative overhead. When you agree to 10 things versus five, you're doubling the administrative overhead, which means you're having the amount of time available to actually do work. I think a workload management being something that we ignore and allow a list of tasks and commitments to go out of control, it's a real unforced error. The second principle is work at a natural pace. I We're seeing this idea that we're trying to mimic a factory set up in knowledge work. Come to an office eight hours a day, full intensity. We're suspicious if at any point during that day, you're not doing full intensity, year round, no change. That's not very natural. Humans aren't good at that. We need more variation. We We need sprints and rest. We need seasons that are more busy than others. We need to play the long game, what's going to produce the most value.

Then the final principle is obsess over quality. The biggest bulwark we have towards the allure of busy is to get good at things. Craft is like an antidote to the poison of excessive busyness. As you care more and more about doing something well, the more the idea of just being very busy, being very visibly just displaying the performance of work, the more that becomes anathema. So the final principle is this idea of let's make craft what we focus on, not activity.

I have a a boomer mentality around this, and that is, and I'm very open to pushback here where I have this wrong or a I love your color here, but I've always thought that if you want to be economically in the top 10%, much less the top 1%, that it just requires a certain level of commitment around not only productivity, but just sheer time. That output is going to be a function, I think, of productivity times the amount of time you apply to something. And then unless you're smart enough to be born rich, if you want to attain real economic security, there is going to be a certain or extended periods of your life where you're working a great deal. Where do I have that wrong?

I don't disagree with that. What I think matters is what you're aiming that effort towards. So I think a trap that people get in is that during this period of busyness, they focus it on things that aren't actually pushing forward their value to the market. I am in on every meeting. I'm quick to respond to the emails. I'm volunteering for every small thing that comes forward, as opposed to, I'm really locked in on trying to get very good at this thing that's unambiguously valuable to the market. I agree with you that if you want to get to the top 1%, there's a huge amount of hours involved in that. There's a huge amount of grinding. I think that's just a property of the jobs that get you to the top 1% of income. This is just how major law firms are structured. This is just how finance is structured, for example. But if you want to be in a situation where you're financially secure and you have a lot of autonomy in the shape of your working life and how it fits into the other aspects of your life, building a skill that's unambiguously valuable, that's the number one lever you have to pull.

That's what's going to allow you to say, This is where I live. I work these months and not this month. I'm taking a sabbatical. I don't do meetings. You want that type of autonomy, that really slow productivity. That's all going to be based on unambiguous skill. This thing I do is valuable. Other people can't do it. Now I'm going to start dictating how I want to actually apply it.

I love that term unambiguous skill. I think a lot of young people think, Okay, I have the context for that. Maybe I have the certification. They think they have the core character and attributes to develop an unambiguous skill that's highly differentiated, can't be replaced easily. I often hear people talk about your 20th century for workshopping. Have you thought about steps and the atmospherics for a young person to try and figure out, identify, and then develop what is ultimately that unambiguous skill?

A decade ago, when I was first entering the job market, out of grad school, entering the academic job market, I wrote a book all about this because I really cared about this question of how do you engineer the right career. I wrote this book called So Good They Can't Ignore You. It was all about this idea, building up what I called career capital. One of the tricky things that came out of that book is that It's difficult to figure out what's valuable, and it takes a lot longer than you think when you're in your 20s to actually get good at those things. I ended up advising young people after this book came out to spend even more time, that I emphasize, almost treating their own field like you're a business journalist. I am studying my own field. I really want to understand who is getting ahead in this field. Who's the stars in this field? Why are they the stars in this field? What was it that they did that other people didn't do? What really mattered for them? Often, these answers aren't what you think they're going to be or what you want them to be.

I also emphasize it takes longer. I came out of college moving hot. I was like, This is what I want to do, computer science and writing. This is what I want to do. I want to be a professor and I want to write. It took me about a decade before I got anywhere notable in either of those. A decade of really solid building my life around computer science and writing, and this is all I do, and I'm not going to be distracted by this. It's head down, systematically trying to get better. It It wasn't until I hit my young 30s, after a decade of work, that I had really made what I would say are publicly visible, notable steps forward. It took me a decade to get my first major hardcover idea book out. It took me a decade to get hired as a tenure track professor. Often, this is the advice I tell young people, spend more time than you think to figure out what matters. Spend more time than you want to get good at it.

Yeah, you're getting no sympathy from me. You're talking to a guy who wrote his first book at 52. I think it's just so funny to listen to you because you're a young man. You can't objectively evaluate time. You came out of the gates rocking and rolling. I don't care what you think about yourself. I can tell you, someone who's older than you and has seen a lot of people, you came out hot at a very early age. What were the things or the attributes of the people or the surroundings that helped you get on the tenure track so quickly and have a best-selling book in your 30s? Most people just don't accomplish that at that age. That is ridiculously young. You're just You're just not old enough to realize how young you are, so to speak. But in your 20s, what do you think were the primary influences and decisions you made that gave you the ability to do that stuff in your 30s?

Well, one thing I think that influenced me is that I had been an entrepreneur when I was a teenager. I was a teenager in the late '90s. This was the first dot com boom. We weren't very smart in that boom. You remember we had this moment where we thought, 16-year-olds, they must know a lot about technology because they're young. Let's sign contracts with 16-year-olds to come redevelop our web presence. That's right. They are mentors. Yeah. Exactly. I had a preposterous business in the '90s, but that had exposed me a lot to the entrepreneurship and business literature, the pragmatic non-fiction literature. I was steeped in this stuff at a very young age. I came to college with a different attitude. I came to college saying, this is a problem to solve, just like any other entrepreneurial problem. In fact, when I went looking for the right books as a college student, okay, how do I win at this game? How do I study? How does this work? I was so disappointed in what I saw because in the late '90s, early 2000s, the sentiment in the publishing industry is to write a book for a student, you had to be cool.

It had to be like, hey, college is fun. Don't talk too much about success because you're going to scare off the kids. The number one best-selling college guide in the year 2000 was The Naked Roomate. It was like an other tales of wellness. I had this idea, this is nonsense. I I'm going to write a book about how to do well in college exactly like a business book. I was bringing a entrepreneur business mindset that would be completely non-exceptional to someone who is mid-career. I was bringing that to the world of college. The thing is, it worked really well. I got very systematic about what's the right way to study, what's the right way to write papers. That experience to me, oh, by just getting serious about how you do this, the mechanics of work, that made such a big difference. I had to write a book about that. I wanted to bring that mindset. I had this interesting dual experience in college where I was there in college, but I also felt like I had already graduated 10 years earlier, this impatience to get out there. That's how I got started writing. I was like, I'm going to write the college book I didn't have.

That's what I was like. I came out of the gates hot because I was there in college already ready to be 30, if that makes sense. I couldn't wait to get out and start doing things.

How has your domain expertise and your study of productivity changed the way you allocate your time in personal life? How has your personal life changed?

Well, one of the things I figured out pretty early in grad school is just strict separation. This was an anxiety diffusing technique, I guess, if we're going to call it anything, was, okay, I had to have a strict separation between working and non-working. And that's what I did. In grad school, I just established a standard. I'm going to work standard work hours, which is unusual in grad school because you don't have to work standard work hours in grad school. But I got married young and my wife was working standard hours. So great, I'm just going to match my work to when she's working. And outside of that, no work. I just work backwards from that standard, and I've more or less kept that ever since. Work fits in work hours. When you're done, you need to shut it down. There needs to be this strict separation between when you're working and when you're not. If your work doesn't fit, then you need to scale back or get more efficient. But these are the lines in the sand. It's like, these are the hours I work. I innovated that in grad school when I had a lot of flexibility and I've stuck with it ever since.

Do you have kids, Kyle?

I have three.

How has it changed your approach to fathering?

Well, I think it helps this idea that when I'm done, I'm done, because I can synchronize that pretty closely with I'll start after they go to school and I can finish pretty soon after they get home. But the three kids, they influence slow productivity in particular because of the timing of when I wrote that book. It was when I have all boys. When they got elementary school age, as they all approach elementary school age, is where there is this switch where they seem to need maximal bad time, which was different than before. It was this every minute I could possibly give them was important in a way that when your kids are toddlers or babies, it's much more of a survival play. Hey, let's just keep these things alive and prevent either of me or my wife from burning out. When they got to elementary school age, they needed maximal bad time. That led to this interesting reappraisal of my own work. It helped really influence slow productivity where I was really thinking, I need to stretch out what I'm doing. Maybe I'm at a peak right now in my capabilities, but I got to slow down the pace at which I'm doing this because this is the time in my kids' lives where they need as much as me as possible.

There's a lot of slow productivity. There was a reaction to my kids getting to a certain age.

We'll be right back.

In a lot of ways, the leader of Canada is not like the leader of the United States. Justin is 52, Joe is 81, but the two do share some struggles. Trudeau has been in office for a while. He's being challenged by a conservative populist, and he's struggling in the polls, especially with young people. But now he's got a plan to fix it, all of it. Among the $52.9 billion in new spending- Proposing a $23 billion program to get nearly 4 million homes built between now and 2031. He's got a budget that speaks directly to millennials, and Gen Z, or Gen Z, as he calls them, and he's hoping it'll keep populism at bay in Canada, too. Prime Minister Justin Trudeau is going to join us on Today Explained to tell us all about it. Today Explained, Vox's Daily News podcast, Monday through Friday in your feeds.

Let's revisit AI. On your Deep Questions podcast, you recently spoke about AI and productivity and how to get ahead while others Panic. I love that. How to get ahead while others panic. You mentioned that a pressing question about AI is the following, When will AI be able to empty my inbox on my behalf? Say more.

I think that's a good stand-in for the crucial future development to watch an AI. This came out of a New Yorker piece I published a couple of months ago that was looking at planning an AI. I'm increasingly coming to this point of view is that we have too much emphasis right now on language models. I think this idea that the the right way to increase the reasoning capability of artificial intelligence is to build larger and larger language models and just hope that the emergent reasoning properties that come out of trying to do token prediction effectively, like this is the right way to build reasoning, I don't think that's going to be the future of AI. I think the future of AI is going to be much more streamlined language models that are good with language, but that we are going to more and more explicitly externalize reasoning and cognition into separate models. Let's go to this example of emptying my inbox. A large language model can never do this. I don't care how many parameters you put into GPT-4, it is not going to be able to empty my inbox because in particular, that requires future planning.

You have to actually simulate possible futures based on the novel input that you're looking at right now, not something that you pretrained on. Static feed forward models can't do this. On the other hand, if we look at a model like Cicero, the model that was developed at Meta that plays the board game Diplomacy and plays that game really well, the way that AI system works is that it has a language model to interact directly with the human players. But a future simulation reasoning engine to simulate the possible moves, to try to much more explicitly understand, Okay, what's the right thing to do here? It has an actual explicit theory of mind. Here is the other players. Well, if this player secretly thinks this, even though they're telling me this, then would it make sense to make this move, or could they betray me in a way that's going to hurt my position on the board? Or what if player really thinks this? And it actually explicitly thinks through future scenarios with explicit theories of minds for the other people involved. Then when the planning engine figures out what makes sense, it tells the language model, All right, go say this to the other people in whatever way is natural for this game.

That's the type of system that can empty our inbox. I'm increasingly convinced that's the future AI play. We see this in the move towards agent systems. I think we underplay how hard it is to build these agent systems. We've come back full circle It's like 1950s Dartmouth College, original conference on AI, that in the end, we're going to end up having to build explicit reasoning systems that correspond and collaborate with the language models. Those are going to be really hard, and they're going to be bespoke. We're back to where we started in some sense, except for now, we have language models to attach them with. I'm developing this theory of AI, where the totality of language models is a dead end. It's going to be language models plus explicit other types of models. The ensemble ensembles of these is really, I think, going to be the future of more useful AI.

What do you mean by ensemble?

Well, you might have a language model to talk to other humans and to take what they said and translate it into language that other models can understand. You might have a future simulation model to try to simulate the impact of various decisions to help plan what might make sense. You might have a goal model. The whole thing there is to try to look at possibilities and help tell the planning engine what is aligning with what we're trying to do here or not. You might have a world state model. Language models are static. They can't do this. But you could have an explicit world state model that evolves its understanding of what's happening in the world around it, who people are, what they believe, or what's going on in the physical space. All these models can work together to create something like a functional artificial intelligence. This is what I think is going to give us artificial intelligence agents that are going to be much more like the how 9000. It's something that reasons and thinks and acts and plans. Ensemble I think, is the future. We also see this move within industry. Openai hired away from Metta, the engineer, Noah Brown, who was the leader of this project to build that ensemble model that won at Diplomacy.

Noah Brown's big thing. His big claim to Fame, was that he earlier built the first AI system that could beat professional poker players at no limit text at Hold'em Poker. And his real innovation in that was figuring out how to simulate future moves that had a theory of mind. Openai hired away, probably to run their Q-Star initiative, which almost certainly is a major initiative to add more explicit planning engines into their AI products. I think this is really the direction AI is going, even though right now, all the focus is just on the language models themselves.

You're right at the intersection of technology and HR culture within the organization. Do you think the promise of AI, where the performance is going to live up to the promise, do you think it's underhyped, overhyped, and where do you see it having the most impact on cultures?

I think there's going to be areas where there's really big impact, where there's activities that can be largely automated. I'm trying to carefully track the potential impact of AI on administrative overhead. This is a big part of my thinking about digital knowledge work, is the degree to which the humans have to keep switching our attention back and forth between many different administrative overhead tasks, so emails and chats and meetings and forms. I'm interested in what the impact of AI is going to have there, because there is an optimistic scenario in which it could take a significant amount of that off the plate of the average knowledge worker. That would be a huge impact. I think we underestimate the degree to which we are reducing the cognitive productive capacity of knowledge workers by forcing them to be in this continual state of constantly switching our attention back and forth to servicing all these different administrative tasks. If you could offload that, That could be a major productivity boom. I mean, productivity in the economic macro sense of just looking at the whole sector itself and dollars produced per worker in the sector. That could be a huge boon.

On the other hand, I've been arguing for a long time, we We're trying to get those same advantages by just changing the way we actually organize work. We don't actually have to have a computer program take over our email. We could just reorganize work so that we don't have to check email once every five minutes. I think there's more proximate solutions that require less of a breakthrough. But they do these solutions. This rethinking of how we do knowledge work is really difficult. There's a lot of friction there, so maybe it is going to have to be an AI breakthrough that gets there. But that's what I'm looking at. It's a metaphor, but also reality. Ai emptying my inbox could be a substantial transformation of the experience of creative or non-entry-level knowledge work.

We're both academics. I've been predicting the disruption in higher Ed for a good decade, and that's where I've been wrong. Where do you think AI or disruption is with respect to higher Ed?

Yeah. I mean, a lot of people have predicted the disruption of higher ed. There's two ways to see it. I go back and forth. One way to see it is we've been teaching the same way, basically, a human in front of other humans in a room facing them for 500 years. That means, by God, we really have to change. Or you could flip that around and say, by God, there must be something very fundamental about this mode of instruction if it has survived 500 years. I mean, the biggest possible disruption to higher Ed was the codex. All right, so we have a way of taking the knowledge of the very best thinkers on a particular topic and putting it into a portable, in flexible, completely accessible type of artifact that can then be widely distributed. If anything was going to undermine the model of people gathering and looking at someone at the front of the room, it would have been the codex, and it didn't. Because of that, I'm always a little bit skeptical about maybe major transformation. We thought COVID was going to transform the academy because once we get used to being able to do things virtually, why would we go back?

But it turns out people hated doing it virtually, and we couldn't wait to get back. When it comes to AI, I'm watching carefully. I'm on the task force at Georgetown for the pedagogical uses of AI. I'm working on a big New Yorker piece right now about writing papers with AI. I'm talking to students who do this, trying to understand how students use it. I think it's going to change things, but I think potentially that change is going to be comparable to, for example, the arrival of the internet and Google, which did change a lot of things. Teachers had to adapt. You could now look up information very quickly. The solutions, the problem sets could be googled and they could find them. It might end up We're seeing something comparable to that. Certain types of classes, we have to move more evaluation to in class. When it comes to writing-related classes, we're going to have to get a little bit more explicit about the role of AI in that writing and teach the process of where to integrate it or where not to integrate it. It's possible that it's the next Google, which was a big disruption for higher ed, but it didn't make higher ed, for example, unrecognizable to someone who was around pre-Google.

I would give that the middle percentile, the middle quartile, possibilities would be that.

For anybody who thinks, Okay, I recognize AI is going to be big, or how would you suggest that somebody starting from almost zero gets somewhat AI capable?

It's a hard It's a hard question. I've been thinking about this recently. It's a hard question because what that means is changing. Right now, becoming AI capable essentially means become a good prompt engineer for a current generation of chatbot-based interfaces to language models. I don't know if that is what being AI capable two years from now, once AI actually has some economic traction and is really having some real impact on the way we run our economy, I don't know that prompt engineering to GPT-4 is going to be all that relevant. I think this is the type of thing that's going to be improved on the other end of it. I think a watchful waiting right now is not incautious. Okay, what's going on with language? Maybe I'll keep an eye on it. But I don't know if I think it's urgent right now for everyone to, for example, be aggressively experimenting with chat interfaces to language models, because I just don't know if that is going to be the first major footprint that these tools are going to have on the economic world. It It makes sense to keep up with it. But I just think when the tools have the big impact, we've seen this before.

I study a lot the impacts of technology on economics and business systems. The tools that often have the real tsunami effect, completely changed the way business is done overnight, tend to be incredibly easy to pick up. I think an email is a great example of this. You didn't have to train on email or keep up with what was going on with email and the pine email systems in 1980s universities. When they turned on email at your office, this makes sense. I I type in the name and I hit send, and it completely transformed work. Google was similar. Once it was out there, it was very easy to learn how to use. You just type in what you're looking for, oh, my God, it finds it. This is really great. We've seen this historically that often when tools come through that transform business in this really big way, they're not that hard to pick up. We don't yet know what the first major form factor of AI is going to be that's going to have really large impact. Again, I'm not sure that it's chat interface to language model, so keep an eye on But I'm not going to advocate right now that we need to be teaching prompt engineering to school students.

I think that's just going to be out of date too quickly.

Just as we wrap up here, what would the advice be to your 25-year-old self? I almost feel like you're so far ahead of the game, and you seem so guided and focused from such a young age. What would you add it or add more or less color to? What advice to your 25-year-old self?

One of the things I actually, I wrote about this when I left grad school, and I said, What did I wish I had done earlier? I left grad school at over 28. The big thing I looked back and said, I wish I had done, is when I had first entered grad school, talk to the students who were leaving, the students who had just graduated, and in particular, the students who were getting the really interesting professorships, and really interview them about what matters, how do you succeed in this? Because I felt like from an academic perspective, I wandered too much in grad school. In other words, I worked on problems I thought were interesting. I invented problems and solved them, and I published a lot of papers. But when it came time to enter the academic job market, it's not like I had a huge head of steam. I wasn't on the hottest topics. I wasn't doing the work that was really moving the scale. I remember thinking, Man, I should have actually figured out at the beginning, how do you succeed in this? Because I was at MIT, and I still could succeed in it, but I wasn't coming out of there at the level I could be.

That was My advice to the 22-year-old Cal would be, talk to the 28-year-olds who are just graduating and find out what really matters when you go in the job market, what's going to make you successful versus not, and then starting right now, aim for that. I think I probably wandered more than I needed to in grad school.

Yeah, we always talk about the importance of a kitchen cabinet. Cal Newport is a professor of computer science at Georgetown University, where he's also a founding member of the Center for Digital Ethics He is the author of eight books, including A World Without Email, Digital Minimalism, and Deep Work, which have sold millions of copies and been translated into over 40 languages. Professor Newport is also a contributor to The New Yorker and hosts the Deep Question Christian's podcast. His latest book, Slow Productivity: The Lost Art of Accomplishment Without Burnout, is out now. He joins us from his home just outside of our nation's great capital. Cal, I've been a big fan for a while. You're just such an impressive young man. And your books really have... You bring this, I don't know, you thread the needle between substance and what I'll call retail. Your stuff is very popular, and yet it has that academic heft. Anyways, congratulations on all your success.

Yeah. Well, thank you, Scott. I enjoyed talking with you.

So Algebra of Happiness, quick dad hack. A dack, a dack, de-hack. I have started... My father got me working out when I was, I think, almost 10 or 11. After my parents got divorced, he and I used to run together on the beach in Laguna N'Aguel, and I have very fond memories of it. We would run, and then he would keep running and meet me back, and then he'd have a race with me where we'd sprint. It was just a ton of fun. I have this memory of me. My dad's very old now and has a walker, but I have this memory of he and I sprinting on the wet sand in Laguna N'Aguel, and he would look at me and laugh. Anyway, a nice image. I think one of the images I'll always have in my brain with my dad. Then we'd do this Royal Navy workout together He had this little manual that was all tattered up. I wish I still had it that he had from when he was in the Royal Navy. My dad served on a ship. He was a frog man and repaired aircraft. We do burpies and pull-ups and push-ups, and it got me into working out very early.

Anyways, I'm trying to do the same thing with my boys. What I do is I'm on the road a lot. I'm about to head to Vancouver for Ted. I'll be Vancouver, then Orlando, then New York, then back to LA for Bill Maher. I'm doing a lot of name dropping right now. Then back to Florida, Key Largo, and then home. But every night, I FaceTime with both my boys, and I take them through this mini workout, this mini CrossFit workout, where they do some Arnold presses, maybe some burpees, maybe some sumo deadlift high pulls. They have a and we do it. It's only about 10 or 15 minutes, and I do it with them on FaceTime. It's very bonding. It keeps us connected. Especially young boys who I think have a lot of energy that can sometimes be unproductive, I think it's just really good for them. It's just so much fun at the end. For the last round, I scream beast mode, and they go hard, and I feel closer to them. I think it's going to be great for them. It's something I'm passing along down with my father. Anyways, FaceTime, workouts. Get your boys strong, get them mentally tough.

Find something that you can pass on from your dad to them that makes them feel better about themselves, and quite frankly, just calms their nerves during the day and helps them sleep a little bit better. Anyways, my dad hack, FaceTime workouts with your boys while you're on the road. This episode is produced by Caroline Shagrin. Jennifer Sanchez is our Associate Producer, and Drew Burrows is our Technical Director. Thank you for listening to the Propety Pod from the Vox Media Podcast Network. We will catch you on Saturday for No Mercy, No Malice, as read by George Han, and on Monday with our weekly market show. Oh, Daddy. Hello, Fuego. Can I send Fuego, Son El Pedro. Si. Thanks to Canva for their support. You're busy, there's no denying that, and we all wish for just a little more time in the day. So why not let Canva help you get your work done faster and more efficiently? You can get started with their AI-powered presentations. Just describe what you want with a few words, and Canva will generate amazing slides in seconds. It's AI that anybody can use, no matter what department you work in or whatever task you need to get done.

Finish your deck faster. Start designing today at canva. Com. Designed for work.