By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
Great conversations run long. Their best ideas don't have to.
uListen is free on iPhone, iPad, Android, Chrome, and Claude.
See howTranscript
70 min read · 13,532 wordsBeware of “efficiency” theater
Cal Newport 0:00 So we're at a confusing moment of the AI revolution right now. On the one hand, it seems like every week the major AI companies have some sort of super impressive new announcement to make of a, of a new capability that their tools are able to master. And yet, at the same time, we're getting, like the recent report from Deloitte, which said, uh, when they studied companies deploying AI tools, while they found that most companies had their workers speed up by using AI tools, the return on investment was mixed, which might explain, for example, why Goldman Sachs recently actually reduced the number of their employees who were given access to advanced coding tools like Claude Code, citing both cost and productivity hits. So what's going on here? How can we have, at the same time, tools that are demonstrably becoming more impressive in their capabilities, and yet we're not seeing this translate in the workplace as cleanly as we would hope? It's not just simply making us all much better at our jobs. Well, today I wanna suggest both an explanation for what's going on and a suggestion for how to fix the problems. And in making these arguments, I'm gonna draw from an unlikely source: the history of an incredibly fundamental notion in the modern world, that of efficiency. This is something that we've all heard before, but very few of us know the actual scientific and mathematical history. But if we learn it, we can actually gain a lot of insight about what's going on right now with AI in the workplace and how to make things better. Uh, it's a... Trust me, it works. So stick with me here. Let's get into it. [gentle music] All right, so I'm gonna have to get into professor mode here because the history of efficiency is something that's going to unlock a lot of insight for us into what is actually happening with AI in the workplace. Let me tell you the source that got me started down this path. It was a 2009 paper, I have it on the screen here for people who are watching. Uh, it was written by the historian of science, Jennifer K. Alexander, and the title of the paper was The Concept of Efficiency: A Historical Analysis. All right? So we can take that back down, Jesse. But that is the paper, I just wanna give credit where credit is due, that brought me into the world of the history of the notion of efficiency, and all the ideas we're gonna talk about started from this paper. All right, so what is efficiency? Well, this feels like common sense to all of us. Yeah, you wanna be efficient versus not efficient, but it's actually a very, uh, complicated concept that has evolved over time. As Alexander points out, it goes all the way back to Aristotle. He didn't use that term, but they had notions of efficiency back in Aristotle that then got adopted into scholastic scholarships. So you have Aquinas, among others, who are taking, uh, Aristotelian thoughts and combining it with Christian theology, and so then you have this notion of God as, uh, uh, an efficient first mover. And so we, we have efficiency in an almost a both philosophical and theological lens that's sort of bouncing around. It means a bunch of things. But then we get to the Industrial Revolution, and this is where the modern notions of efficiency arose. Now, what I discovered reading Alexander is that if we go back and look carefully at what this term used to mean and then how that changed over time as the Industrial Revolution progressed, this is where we're gonna get our deep insight into why AI is having issues with work and how to fix it. All right? So we're gonna go into history mode here. Um, I wanna start by going back to the 1750s when a British engineer named John Smeaton, which, uh, Jesse, I think you would agree is what you would name a British engineer from the 18th century if you were, like, doing a Simpsons episode about it. John Smeaton, who ran a series of critical experiments on water wheels that began a revolution in what we thought when it came to the notion of efficiency. You know, water wheels are where you have wheels, where moving water turns them, so you can take moving, flowing water and turn it into, uh, mechanical energy for doing things like grinding grain or running belts in simple factories. Okay, so what did Smeaton do that was important? I'm gonna put a picture up here. This is from a, a paper that he wrote. Uh, I have the date. It's in the Philosophical Transactions in 1759 to 1760. Uh, so I have this paper. You can see the old fonts in it, but I wanted to show this one diagram up here. All right, this is from his paper. For those who are listening, you see, like, a small water wheel. It's in a carefully controlled box. There's a tower behind it and there is a pulley system that leads to a weight. So the water wheel has this tower behind it, and it's connected to a pulley system with a weight that is on it. This apparatus revolutionized the way we thought about efficiency. All right, we can take that down, Jesse. Here's what Smeaton was up to. He said, "I want to quantify how well a water wheel is working, and I wanna do this with numbers," and that's what he built this, uh, machine to do. And so what would happen is, as water would pour, uh, down from this tower and it would move the water wheel, the water wheel would lift this, uh, pan with weights in it, and it would lift and you could measure how high it was lifted. And what Smeaton's, uh, figured out or his, his observation was, we can use the amount of weight and how high it is lifted to measure how much work that water wheel just did. So I'm actually gonna put this on the screen here. We'll do a little... It's like my chalkboard here as the professor. But the first equation we're gonna look at, which comes from Smeaton, is when he was looking at water wheels, he said, "Useful work So I'll write the word useful work, beautiful handwriting, you can tell, like an 18th century calligrapher, equals... I'll shorthand this as weight times height. So let's think about this for a second. Think about that water wheel apparatus. How much weight did the water wheel, uh, lift during the experiment, and how high did it lift it? So that he can quantify then, this is how much work the water wheel did in a particular experiment where I ran a certain amount of water over the water wheel. Okay, what he wanted to then compare is how close did that come to the maximum amount of work that water wheel could have done in that same experiment. Now, this is where things get tricky. How do you measure, theoretically speaking, the most work that water wheel could have done in that situation? Well, it depends on the type of water wheel. Um, the type that's easier is what's known as an overshot water wheel, and that's where the water pours onto the water wheel, uh, at the top and fills a bucket, and that bucket tips the water wheel forward. You see, you know what I mean, Jesse? Where you, you come in at the top and it fills and then it moves forward. You fill another bucket and it moves forward.
Speaker 7:24 Yeah.
Cal Newport 7:24 Yeah, as opposed to an undershot wheel where it's more like a paddle wheel or where the water is moving at the bottom of the wheel and pushing it. So with an overshot wheel, um, he figured out, and I'll, I'll put this on this, our next equation here. For an overshot wheel, he could say, "Great, we can measure potential work pretty easily as potential work equals," what I wrote, "weight of water times height of wheel." All right, so what do I mean by that? So he said, "Look, we have the water. The wheel is this tall. We have this much water up here, and the water is this high off the, the, the ground at the bottom of the wheel." Right? So the water's at the top of the wheel. We're gonna have it fill it from the top. How h- And it's this high from the bottom of the wheel. This is the, the, where the water, the stream below it. So he says, "Let's just take the weight of the water up here that we're gonna pour on the wheel and multiply it by the height that water is gonna fall by the time it gets to the bottom after it's gone through the wheel." That's the theoretical maximum. Getting that whole weight times the whole distance it can travel if we had a perfectly efficient water wheel, that's the, that's the, uh, max amount of work that it could do. All right, this is where it got interesting. One last equation we'll put up here, Jesse, is he said, "We will now have the idea that efficiency," so now we're getting this term having a technical definition, "is gonna equal useful work over potential work." So just divide those two equations. So now what efficiency was measuring in Smeaton's 1750 paper is what fraction of the potential work this wheel could have done with that much water moving that distance, what fraction of it did it actually realize? So the more efficient your wheel, the closer you get to 100%. And the more inefficient your wheel, so maybe it's creaky and it's leaking and it has a lot of, like, friction that makes it hard for the wheel to move, um, that, that fraction's gonna get much smaller. The percentage will get much smaller, that you're gonna be wasting a lot of the potential energy that was in that water that was sitting there up top. This was a huge idea. It's the work you produce by a machine divided by the maximum, theoretical maximum work that machine could have done with the, the amount of energy you had available. That is how we're gonna measure efficiency. So the whole game then was just how do you measure these quantities? The, the output is usually pretty easy because you, you-- typically these machines produce mechanical motion, right? So a wheel that moves, that's a mechanical motion. We can have it lift the weight. The hard part was always gonna be then how do we measure the potential work? And for the overshot wheel, that was easy. We have this much water falling this distance. That's the maximum potential work. For an undershot wheel, he-- Smeaton dealt with this in his paper. It was more complicated. He looked at the velocity of the water and its weight and tried to get some sense of the, the work. It, it gets into some sort of, um, proto-physics there. But now we had an idea. Efficiency is what fraction of the maximum potential work that your system and this fuel source could do did your machine actually realize? I wanna take a quick break from our discussion here to hear from some of the sponsors that make this show possible. Now, longtime listeners know that I'm fascinated by the idea of augmented reality, in which you can put on special glasses that add digital information or digital elements into, uh, your view of the real world that is around you. I have done podcast episodes and newsletters about this for many years, about how I think, uh, this technology is poised to break out big. This is why I'm excited by the new RayNeo Io smart glasses. These retro minimalist eyewear use advanced micro optics connected to cutting-edge AI models to seamlessly bring the data that you need into your world. Imagine, for example, that you're traveling overseas and trying to order in a restaurant. If you're wearing the, the Io glasses, you can activate the built-in AI assistant by simply saying, "Hey, RayNeo," and then ask a question about the menu or ask for a translation, and then the translation can show up in your field of view like closed captioning as the waiter is talking to you. You know, right under it. Or maybe you're on a walk wearing the glasses, and you have a great idea. Again, you could be like, "Hey, RayNeo," and then explain the thought, which the AI assistant transcribes, summarizes on your behalf, stores for you without you ever having to pull out a phone or look at an artificial screen. You can even use the RayNeo app to switch between ChatGPT and Gemini AI models, depending on your preference. Um, with Black Friday upon us- This is the perfect time to get a pair of the I/O glasses for the tech enthusiast, traveler, creator, student, or busy professional in your life. Here's the details. RayNeo is offering 15% off all orders that are placed from November 19th through the 30th, along with a 30-day free trial, free return shipping, and a free six-month ChatGPT subscription. Best of all, if you order on Black Friday itself, which is November 27th, you will also receive a free custom prescription lenses with your purchase. So take advantage of these Black Friday deals and learn more about the RayNeo I/O smart glasses at rayneo.com. That's R-A-Y-N-E-O.com. So something I've learned running my own business is that you often get started thinking that you're gonna be working, you know, all of your energy on producing whatever it is that your company does, but you end up having to wear all of these other hats that you never anticipated. I'm talking about like payroll or benefits or HR. Like these are things we worry about and have to deal with a lot here, uh, at my media company, and it's not what you thought about. This is where a product like Gusto enters the scene because it can bring together all of these sort of behind-the-scenes activities into one tool that makes everything easier. Gusto brings payroll, benefits, HR, and time tracking together in one system so small business owners can spend less time managing their workforce and more time growing their business. Gusto automatically calculates files and pays your payroll taxes, and on average, it can, uh, s- require just eight minutes to run your regular payroll. Actually, our ad agency that we work with uses Gusto to deal, you know, back and forth with us, and it's a breeze. Uh, it's like this is fantastic. It makes everything clear and simple. So I am a fan. Payroll, benefits, compliance all in one place. Get started at gusto.com/deep and you will get three months free when you run your first payroll. That's three months of free payroll at gusto.com/deep. One more time. That's gusto.com/deep. All right, let's get back to the show. Okay, so Smeaton introduces that idea. This is an incredibly important idea. So now people are thinking we can measure how well machines work and try to make machines better by getting that percentage closer to 100. Where this idea of efficiency, this ratio definition of efficiency, where it really took off was with the steam engine. Because the steam engine was at the core of the Industrial Revolution. We're gonna burn some fuel. It's gonna boil water. That boiled water is gonna create steam. That steam is gonna be transferred into mechanical motion. We're gonna go from heat to mechanical motion, and that's what was gonna power the Industrial Revolution, at least until we got to electricity. Efficiency was a critical notion for developing the steam engines because it became really non-intuitive. I don't know if this is a good design or not. I don't know. Steam engines are-- they all look different. There's completely different ways of doing it with active vacuums or just atmospheric pressure, different types of cams and running bars and valves, and they all look different. They're all different sizes, and they're all complicated to look at. And who knows? Is that good? Is that good? I don't know. So how do we measure the efficiency of steam engines? We needed to answer this question for the Industrial Revolution to move forward. Luckily, a new type of physics began to get developed in the 19th century that helped us here, which was thermodynamics. Thermodynamics allowed us to bridge the gap between heat, like produced by fuel, and mechanical motion. We, we can give credit to this entire sort of trajectory of steam engine efficiency to, uh, James Joule, who built a sort of complicated apparatus that allowed him to figure out an equivalency between mechanical motion and heat. It had a, a water with a thermometer and something that turned, and it's a complicated apparatus, but he, he came up with a unit of, of work called the joule, named after himself, which you could apply to both, um, you know, heat and mechanical motion. It could unify it. How many joules of potential energy is in a piece of coal? How much joules of work did, uh, this steam engine then do? Um, we needed to solidify those relationships mathematically. It's not as easy as I just said it there, but the laws of thermodynamics, which really got developed in the 19th century, allowed you to have this very nice formal relationship between like the potential energy in a fuel source and mechanical energy, and we could then come up with an efficiency ratio. I'll write it here on the screen. Very similar to with the water wheel. We, we got to this idea of engine efficiency now equals energy of output, which I'll put in parentheses, is mechanical, divided by energy that was inherent in the input, which is measured in, uh, the heat that can be produced by it. So how much heat, how much heat could be produced by a piece of coal? How much actual mechanical energy was output on-- by the engine that's being powered by this coal? All right? So using thermodynamics and these new measures of energy, we could then talk about the efficiency of steam engines. But again, it's a ratio. It can't be bigger than 100%. So a perfectly efficient steam engine is one that transfers every bit of energy latent in the fuel into the mechanical motion on the other side. That's what efficiency meant. That was the definition from the 18th and 19th century that was used to help define machines that did mechanical work. And so it was a very sort of like, uh, a, a very precise motion, notion, very mathematically defined. Okay. So how are we gonna connect this now to knowledge work? Because that's not really how we think about efficiency, right? When we're thinking about are you an efficient employee, there's no thermodynamics. There's no, like, ratio of, like, 100% to go between zero and 100%. So how do we get to the way we think about efficiency today, which is much more about, like, more generically, like, I don't know, I don't waste, I move quickly. Like, it's this much more sort of squirrely notion. How do we get from this nice precise world of thermodynamics to this sort of more squirrely notion of efficiency that we, we think about today? Well, this begins to happen in the late 1900s and in, or the late 1800s rather, and into the early 1900s. 'Cause here's what happens. So many different thinkers out there, social scientists, progressive politicians, managers in companies, all these different thinkers are looking at the, the sort of economic miral- miracle of the Industrial Revolution where these efficiency measurements, like, wow, it's mathematical. We're, we, we use science, and it was precise. And, and we, we could build better machines, and they made us more money. They said, "We want to apply this to other things." Like, social reformers were like, "We wanna use efficiency to measure, like, government programs. So how do we, we could be precise, and is this government program more efficient at delivering goods for the money we give it?" And managers were like, "Can we apply this to people? You know, what's the, the number of forms filed per employee, or the number of Model Ts produced per, uh, per worker on the assembly line?" Like, we, we wanna think about efficiency, but for things more than just the, the transfer of fuel from one form to another, from potential energy to mechanical energy. We wanna use it to all these other places. So what arose was a different way of thinking about efficiency, and this brings us to an interesting character who, uh, comes up in the Jennifer Alexander piece. He's been largely forgotten, but he's a critical character for understanding our current moment. And in fact, I think, let me pull it up here, I think I even have, uh, his book. Yes, this his book. Yeah, here we go. Um, the character I'm talking about here is Walter Polakoff. Walter Polakoff was a, a thinker and a theorist, an interesting character because he was a Marxist, straight up Marxist, but also was sort of in the circle and was carefully watching and trying to learn from Frederick Winslow Taylor, who you would think the, the, the Marxists would hate. Like, the guy with the stopwatch to say, like, "I wanna make workers' lives miserable," doing time motion studies, trying to get workers to be, to move as, as, uh, as optimally as possible, uh, the scientific management. He was a Marxist who liked scientific management. He was always sort of an interesting character, but here's why he's important to our story, is that he looked around and said, "Hmm, all of these people are talking about efficiency today." He, and he wrote a book about this. Here, I'll put it on the screen here. Mastering Power Production. This came out in 1921. Look, I'll show you the contents. Jesse, you'll appreciate this. The contents actually look like a flowchart. I think it's pretty cool. So he writes this book in 1921, Mastering Power Production, um, building on some ideas from a 1909 paper, where he said, "There's a different notion of efficiency we're talking about." He says, "We have to be clear about our terms." This is very important. He said, "The efficiency that we know, that we use for steam engines and water wheels, we're gonna call that bounded efficiency, and it gives you a value from zero to one." There's a theoretical maximum. Converting 100% of potential energy into ma- uh, into output, right? That's bounded efficiency. He said, "What everyone else is doing today in the early 20th century is different. We're choosing two perhaps loosely connected measures and just putting them into a ratio," right? Forms filed per worker, um, you know, uh, jobs filled per dollar put into this government jobs program. We're taking two things, and we're putting them into a, into a ratio, and we're measuring that ratio because we like the fact that there is a, um, distinctive number. But he said, "But, but this is different. Why? Well, first of all, there's no limit to it," right? There's no 100% here. Forms filed per worker. There's no limit. What's the maximum number of forms a worker could ever, uh, file? We don't have a mathematical law of thermodynamics to tell us. We're not talking about conversion of energy. So we're, we've lost a world of, of, of limits. We're not talking about fractions of potential realized anymore. It's just a ratio. And he said, you know, um, "We no longer have a, a, a, like a strict physical s- connection between the measures as well. We're choosing the two things we wanna measure in ratios." He said, "I'm gonna call that arbitrary efficiency." So we've gone from a regime of bounded efficiency that helped us build the machines of the Industrial Revolution, and now we're in a world of arbitrary efficiency, where we're taking the mood of efficiency and trying to apply it to other things, but we're losing the sort of mathematical structure that actually made efficiency made sense. We're in a different world here. Now, here's what's important. Polakoff didn't say this is definitely bad, but he said it's different, and because of that, we need to be a little bit careful. So he said, "There's clearly advantages to arbitrary efficiency. Uh, it gives you a way of measuring something about complicated systems that might be otherwise inscrutable, and if the ratios are well-chosen, it can really help," right? I mean, I wrote about this in several of my books. When Ford, Henry Ford started measuring Model Ts produced per paid worker hour, that arbitrary efficiency metric gave him something to look at. So now if he tinkered and changed something in their assembly process, he could look at what it did to that number and say, "Look, I think if that number gets better, that means we're building cars better." He had to make that claim himself. It's intuitive. It's logical. But it's still a claim he came up with. But he says, "If we believe my claim-" That more Model Ts per paid worker hour is better, then we now have a target. And if we mess around with the assembly line and it makes that number go up, that's a better way to build cars. So it does have a way of allowing you to cut through complexity and complex systems and have a single number to see how it moves. And in the case of Henry Ford, that really worked because it helped him develop and refine the continuous motion assembly line, which increased that metric by, uh, a factor of ten to one hundred. That clearly turned out to be the right metric because it became the biggest company in the world and he became the richest man in the world, at least for a little while. But there are also disadvantages of arbitrary efficiency that you have to be careful about. Problem number one, because you're coming up with arbitrary ratios, again, it's not a, a, a physical relationship between these two entities, like fuel energy and mechanical output, right? Uh, there's not-- they're not actually-- there's not a, a physics connection between them. If you choose those incorrectly, you might be measuring spurious connections, or things you think are improvements might not actually be measuring real improvements that are making your system better. So once you go to these arbitrary ratios, you can get arbitrary results. Let me give you an example I was thinking about before. Let's say, for example, you say, "Okay, I'm running a university, um, and what I'm gonna measure, my arbitrary efficiency metric is gonna be graduates per faculty member." So how many, um, I don't know, how many graduate students are graduating, like getting their degree, um, divided by the, the number of faculty member we have at the university, because maybe that, like, makes sense. Like, our, our goal is to, to spread knowledge and to spread new thinkers, and so that ratio makes sense to me as administrator. So as we mess around with policies, what we're looking for is policies that makes that number go up because we're a more efficient university if we can graduate more graduates, right? So this is, like, an example of a, a metric we might, um, come up with. But what if in reality, in this particular university, only about half of the faculty members actually supervise graduate students? Maybe the other half are teaching faculty, and they don't, um, graduate graduate students. So what they do is important, but it's not captured in that particular arbitrary ratio. All right. So now let's say, why is that a problem? Well, now let's say, you know, you put in place a... Some dean comes up with a policy they put in place, um, that, uh, begins to drive, I think I have this backwards, that begins to drive some of those teaching faculty out of the university, right? Like, uh, whatever. It's a policy change that has this as a side effect, is that, like, you're losing teaching faculty. It makes their life worse in a way. Like, "I'm gonna go, um, I'm gonna go somewhere else." If you're really looking at this university, you're like, well, this has not... And, and, oh, at the same time, let's say nothing else changes, and the, the, the faculty supervising students, this policy changes nothing about what they're doing, and the number of students graduating stays the same. The ratio says this policy was great because the denominator just got smaller. The number of faculty got smaller. So students graduated per faculty member just jumped up. Wow, that policy was great. But in reality, you actually had major university much worse. You did not get better at, you know, producing students, and the, uh, teaching quality just plummeted because you just lost a lot of teaching faculty. So the point of that example is while it can be useful to find a ratio to quantify an otherwise complicated system, it can also be fool's gold. A ratio that might make sense might not be measuring the right things, or the relationship between the two factors might actually be spurious. And so, you know, you're, you're, um, you're measuring the ratio of, uh, people with blue eyes and Model Ts produced, and these aren't actually connected at all. And so, you know, things that are changing the number of people with blue eyes or the, the Model Ts produced, like none of this might have anything to do with how effectively you're producing cars, but you're chasing this metric, and it can lead you to arbitrary places. The second problem, and this is the problem that Polakoff really focused on in his book and his paper with arbitrary efficiency, is that there's no limit, and that's a problem. It's, it-- How do you use a metric where the, it, there's no maximum? It could just keep getting, in theory, larger. He uses the example of potato yields and saying like, uh, he gave some example, like potato yields, I forgot where, you know, um, some state or country or whatever. And he's like, "This, this metric we have of potatoes per acre is," he's like, "arbitrary. Like you could have two hundred percent yield, three hundred percent yield. So what does this even mean?" And this is a problem because if there's no upper end to a metric, it can create unrealistic expectations or like a, a relentless drive to continuing to like push and do more and more because there's always... If the number can always get larger, there's no end to it. And it could drive you to places that aren't great, um, for the people actually involved in it. Okay. That's our history of efficiency. How does it connect to modern knowledge work? Modern knowledge work is deeply intertwined with the concept of arbitrary efficiency. It's implicit, right? We don't use this terminology so explicitly, but Polakoff's analysis is straight on because implicitly in the minds of managers in the knowledge work environment are these arbitrary efficiency ratios. And probably the most common one, um, is going to be, uh, in fact, in the spirit of having written other equations, I'll even write it on the screen here. But probably like the most common implicit arbitrary efficiency ratio in knowledge work is something like KW, knowledge work efficiency, equals, and again, we don't say this, but we kind of just know this, number of tasks completed divided by, you know, employees. It's, it's, it's something like this. In my book, Slow Productivity, I call this pseudo productivity, but I think the efficiency framework gives us a even more precise way of talking about this. That we have this sort of arbitrary efficiency ratio of, like, how much stuff do I see you doing or finishing, or activity, you know, divided by, like, employees. So, like, the more things I see being done as a manager, the more efficient I think my operation is running. So we're really managing knowledge workers with an arbitrary efficiency ratio. Well, both of the problems that Polakoff looked to or pointed towards or implied about arbitrary efficiency really do apply to this particular efficiency ratio that we're using in knowledge work. So for example, the first problem we said about arbitrary efficiency in general is that what you're measuring might not be actually capturing the real dynamics of the system or might not actually be capturing what you care about. I think this is definitely the case when it comes to knowledge work because it's often, if you look at it, the, the, the visible tasks, the things that a manager can observe are being completed, many times these can be only loosely connected to the actual outputs that create value for the company. Right? There's a loose connection. Sometimes there's a negative correlation that if you spend more time doing, like, answering emails, sending out, uh, your sort of, you know, elaborate decks that you, you, you created, um, summarizing calls, jumping off and on Zooms. Like, if you're doing more and more of this, as I document in my book Slow Productivity, you could actually be getting worse and worse at producing the things that people pay for. So visible tasks actually can be loosely connected at best to the actual ROI, to the actual, uh, underlying economic health of an or- organization. So we have that issue. The second problem with arbitrary efficiency is what happens if you have no upper bound? Well, that clearly is a problem with knowledge work as well because a ratio of tasks per employee, there's no one hundred percent. It's just a numerator can keep getting bigger and bigger. And so that creates pressure to do more, work harder, keep your intensity pegged at ten longer. And because we have mobile computing and phones and smartphones and laptops or whatever, spend more and more of your time, if you're in office or not, continuing to task, to, to rack up more visibly completed tasks. So because there's no upper end to it, the opportunities to work become endless, and any moment of recharge, relaxation, or downgrading of intensity is hurting that ratio. It's hurting th- this ratio that you can make larger. I call this situation that this arbitrary efficiency metric has pushed knowledge work efficiency theater. We're all playing the role of doing visible things. It's a theater. It's a game. We're pantomiming work. And the more of this sort of stuff we do, the more efficient we think our operation is running. This all connects back to technology like AI because what we have seen in the last twenty-five years that's reaching a sort of reductio ad absurdum apex with AI is this idea of creating digital tools in the office that are made to increase that arbitrary efficiency ratio, not the production of value for the company. We got this with, like, the super streamlined digital asynchronous and synchronous communication tools like Slack, for example. What is that good for? Increasing that ratio, not necessarily producing the things that actually produce value for the company, right? This happened with, um, meeting invites and digital meetings and Zooms and everyone shooting invites left and right to everyone and jumping in and out of meetings because, man, that's visible. Can't say that you ain't doing stuff because you see me on that screen six times today. And when we got the... made that easier and easier to do, suddenly our lives were filled with many more of these sort of digital meetings and digital invites fractioning our schedules because we're trying to increase that arbitrary efficiency ratio, not trying to create more actual value for the company. AI makes this almost absurd because with AI you basically have a machine where you turn a crank and it spits out some visible task was done. You know, however many seconds it takes for the model to do it. Well, if we're in an arbitrary efficiency world, you're like, "I'm gonna start turning this crank more and more. Hey, look, if I turn this crank, I've produced as many visible tasks as it used to take me a whole day to do." But there's no bound to that ratio, and it keep getting bigger. So you start turning faster, and you start turning longer, and then you're like, "I'm gonna actually create a, uh, a bigger crank, and that crank is gonna have ten other crank arms, and each of those is turning another smaller crank." Talking about agents here. And now we're having, like, I, I can... I'm automating the completion of task and automating things that manage things that manage things that manage things to complete tasks. And oh my God, all I'm doing is running around with all of, like, monitoring all of my systems where everything's connected together so I can make this ratio get as big as anyone's ever seen. And then you actually do the ROI and say, "We're not making any more money." Right? We're not, we're not shipping more of the things that are actually valuable. Our product isn't better. We're not producing more books. We're writing a lot of code, but we don't actually have new, more new software that lots of people love. And so I think AI is so accelerating the flaws and the shortcomings of the arbitrary efficiency approach. It has so accelerated them that it's becoming impossible to ignore. I think that's what's going on now. So that's how we connect. We went all the way from the water wheel, trying to make the water wheel take more advantage of water, to today, where I have an agent orchestrator who's organizing sub-agents that are trying to make sure that the color coding of my slide deck is gonna match what was in the website from this client for a thing that some other AI is gonna summarize, and that summary is gonna get put into bullet points that someone else is going to read during a conference call that no one's listening to because they're looking at agents, but we'll have another note-taking agent summarize them for again. Tasks, tasks, tasks, tasks, tasks, tasks, tasks. Ratio is huge. Very little is actually getting done. That's what we're seeing in these reports. So if we're gonna bring this all together, here's what I'm arguing. Arbitrary efficiency is a tool to use carefully. Ratios of arbitrary, uh, elements that, like, make intuitive sense to you are a good thing to measure sometimes are, sometimes aren't, and they have no upper end, so you have to apply them with care, and you have to put your own structures and frameworks around them so they don't get out of control. Even Henry Ford did this. He said, "Oh, this assembly line, continuous motion assembly line, makes my ratio I cared about much bigger." But he also said, "We're gonna, we're gonna pay the workers more. We're gonna have to have standardized workdays." Like, he realized, sure, I could, I could, uh, double the hours the same workers did. I can cut their pay, and I could get the, the, the dollar paid to a w- the Model T's per dollar paid to a worker ratio super, super high, but, like, they're gonna come get me with pitchforks eventually. They're not... They're gonna quit the job. Like, it's not, this is not sustainable, right? So even Henry Ford realized you gotta be careful with arbitrary efficiency because there is no limit to it. And he actually cared, am I measuring the right thing? Well, he had a very better, uh, ultimate proxy if the company's growing, so we must be doing something right. We're not doing this in knowledge work, so it's getting out of control. So what is the alternative here? Here's what's interesting. In the Jennifer Alexander paper, she talked about an alternative way of talking about machines and processes that efficiency supplanted during that period of thermodynamics and steam engines. And that alternative was effectiveness. And the way that Alexander defined effectiveness is she said it is a, a measure of the amount of... and I'm paraphrasing here, but essentially, like, the m- a measure of the amount of valuable things you've produced. That's it. And when you're, you produce this office, this person, this whatever, if they produced more valuable things, they're more effective. The key about effectiveness, she noted, was you don't care about how they did it. You're n- there's no, you have no notion of, like, I don't wanna waste things. I don't wanna waste fuel. I don't wanna waste potential. It's just you produced a bunch of good stuff. That's good. You're effective. You produced even more good stuff. You signed 10 clients that are each bringing in money. That's really good. This person signed five clients, didn't bring in as much money. You're more effective than that person. I don't care how you did it. I don't, I don't care if you, y- if you jump on a pogo stick across the parking lot and back before you make every call to a client. I don't care how you did it. The results speak for themselves. That was called effectiveness, and that fell out of favor for efficiency because, yeah, if you're trying to make an engine better, y- it matters. Like, I, I wanna... otherwise, I'm gonna run out. I'm gonna spend too much money on coal. Like, I want my steam engine to be efficient. But when we got out of the, the, the, the energy world, when we got out of machine efficiency, the effectiveness mindset starts to make more sense. So this is my, the, the, this emerging idea I have is that in knowledge work in particular, arbitrary efficiency matric- metrics have screwed us. Like, we're not handling them well. We're, w- 'cause it's hard to measure stuff. We're measuring the wrong things. The lack of the bounds make us crazy. We're inventing tools to serve that metric instead of actual output. You know, like, let's... that's dangerous. But what can we do instead? Turn back to effectiveness. Now, how do you do that? Well, I can tell you. It's actually pretty simple. All right. Two situations, you have a boss, you don't. If you have a boss, how does a boss measure effectiveness? What, how should this work? On a semi-regular basis, like once a month, you know, you have the people you supervise. One by one, they come in, and you say, in a narrative form, "Make the case to me. What did you produce that was valuable to us? What did you, what was your value to us? Like, what did you produce?" And they're like, "Okay, let me make, let me make the case. I worked on this new client portal. Here was the problem we faced. This new client portal, I think it solved all those problems. Let me show you how. I have evidence. The, the clients have said this, and this client, we have all this praise from them. I think these two clients signed because we have this. Here's a client. I'm gonna give you a narrative. I'm gonna tell you a story about the things I did and why they were valuable." And if you, and if you're the boss, you're like, "Okay, if this value seems to justify, like, your position, yay. If you're blowing me away, then maybe I wanna give you a raise. And if you really don't have much to report, I'm gonna be like, 'I don't know what you're doing here.'" Work for yourself, you get to have that same conversation with yourself. I don't care what I did Tuesday at 12. I don't care how many emails I sent. I don't care, um, about my agentic-powered AI personal OS. What value did I produce for my comp... Like, what things did I do? Like, what new products did I put out? How successful were they? That's effectiveness mindset. This, I think, is the way forward 'cause it's, uh, knowledge work is complicated and messy. Getting the human brain to do symbolic processing and de novo idea creation is messy and weird, and it requires us to take time sometimes and, and to disappear and come back and, and, and sit with our thoughts and then be frenetic. Like, I don't know. I can't pin down what you should be doing right now to be good at your job. But I can come to you a month from now, and you can tell me and make your case for why you did your job well, and it'll be very clear to me. I don't need to hear about how many emails you answered, how many meetings you did. What did you do that was valuable for us? That's effectiveness. That's probably where we need to go. But if you head that direction, it changes our relationship with digital tools because digital tools that are called productivity tools in the workplace really have been designed to make that arbitrary efficiency ratio higher. We need a whole different way of thinking about tools when it comes to, uh, thinking about effectiveness. I'll just give you, like, a quick example. Like, in the world of AI, uh, the computer scientist Jaron Lanier last week had an interesting op-ed in The New York Times about AI companies are not giving us what we need. They're giving us tales of doom and their weird obsessions with their apocalyptic cults and, and also tools that are just making people miserable. Like, they need to give us things we need. And he gave an example because he's a computer scientist and innovator, um, and he's working in some material science stuff recently. He's, he was talking about this, he was writing about this AI tool he created for himself. That was built to help him navigate the sort of idea space for this type of innovations he works on, and he could report narratively, "When I use this tool, I'm getting to better ideas that are producing better solutions. This tool may be more effective." That's very different than the sort of Claude cowork mindset or the Claude code mindset of, "I burned more tokens. I generated more lines of code. God knows what they do. I automated every last bit of writing that I could possibly automate. I made my slide decks into these, like, monstrosities of, of, of, like, pages and takeaways, and I..." It just, it's like a... No, he produced a tool that made him produce better solutions. So anyways, this was all very clarifying to me, 'cause I've talked about a lot of these issues before, but I thought this efficiency framework pulled it together. Arbitrary efficiency is a dangerous approach. It can work, but you gotta be careful. We were not careful in knowledge work. It's led us to a bad place. Digital tools had us double, triple down on that bad decision, has made work way less tolerable without making us that much more effective at actually producing results. AI has pushed us to an abducto ad absurdum endpoint where we have to make a change, and I think that change is effectiveness. I don't care how you did it. I don't care what you did yesterday. What have you been producing that makes us better? That's the way we have to start thinking. All right. That's my history lesson. You can tell I'm on sabbatical, Jesse, because I don't have a chance to lecture to people.
Speaker 44:19 [laughs]
Cal Newport 44:20 Efficiency. I read that article. Um, anyways, there we go. Efficiency, ratios, et cetera, et cetera. All right, I want to take another quick break to hear from some of the sponsors that makes this show possible. You all know how much I prioritize diligent focus on the things at work that really move the needle. That's why I... That's why when I learned that sales professionals lose roughly half of their working hours to administrative clutter instead of focusing on actually building relationships and making deals, it blew me away. I knew there had to be a better way. That is where today's sponsor, Pipedrive, enters the scene. Pipedrive is an intelligent AI-powered sales CRM that is loved by growing sales teams. Now, they have a new feature in particular that I wanted to emphasize, 'cause I thought it sounded immensely useful. It's called their pre-meeting brief. It automatically pulls deal history, email records, and previous conversations into one brief that's ready for you before every call, so you know exactly where things left off, so you can jump right in to trying to close that deal. So switch to a CRM built by salespeople for salespeople, and join the over 100,000 companies already using Pipedrive. Now, my link gets you an exclusive 30 days free instead of the usual 14-day trial. No credit card or payment needed. Just head to pipedrive.com/deep to get started. That's pipedrive.com/deep, and you can be up and running in minutes. I also want to talk about our friends at MyBodyTutor. Do you want to get in better shape but struggle to follow through? Here's the solution. MyBodyTutor, T-U-T-O-R, is an online coaching program that actually delivers results. Now, how does it do this? When you sign up, you're assigned a coach, and they help you devise a nutrition and exercise plan custom fit to your specific goals and your circumstances. And then, and this is the key part, you quickly check in each day on the MyBodyTutor app to let your coach know how things went. This is the accountability that really helps you follow through. It also puts an advocate in your corner who can help you adjust your plan to whatever life throws your way, or notice where you're struggling and offer to help. If you want to get healthier, go to MyBodyTutor today, and if you do, mention that I sent you, and they will give you $50 off your first month. That's mybodytutor.com. All right, let's get back to the show. I mean, often in our second act, we like to react to things that you send us and answer questions or et cetera, and I thought we're gonna, we're gonna focus the universe of things we react to today to actually news stories. There's a lot of news out there, um, that's very relevant to both what we just talked about or the things in general we talk about. So we're gonna kind of focus in here. It'll be like an all news reaction, and I'll move quick here, but an all news reaction version, um, of our second act. Remember, if there's an article you want us to look at or you have a question for us or reactions, send it to podcast@calnewport.com. But what we're gonna do here is, uh, start with a couple stories that are relevant, um, that are relevant to what we were just talking about here. Chrome browser doesn't like me on your... I got it fixed, but it doesn't like me. You know what it tries to do? Actually, it's not Chrome. I'm gonna tell you what ha- behind the scenes, everyone. Every time I go to click on the Chrome browser on the tablet we use here in the studio, Perplexity jumps over.
Speaker 47:46 Oh.
Cal Newport 47:46 So it, it's like this is AI coming alive. It tries to get me to click on it. All right. Um, let's, let's react to some stories here.
Speaker 47:54 Our first story is from the AP, which looks at the phenomenon of AI boomeranging.
Cal Newport 48:00 All right, this seems relevant to what we were just talking about here. Um, I'll put it on the screen. We got what to consider if a company that made AI-related job cuts wants you back. Ooh, this was from a couple weeks ago. I'll read a couple lines here. Uh, it says, "Companies around the world are increasingly
The phenomenon of “AI Boomeranging”
Cal Newport 48:18 citing artificial intelligence when slashing jobs, fueling anxiety that a lot more workers may lose their livelihood to machines. Uh, Microsoft, Barclay Bank, and supermarket chain Kroger are among the latest entities on a growing list of companies that have announced AI-related workforce reductions this year, including the elimination of jobs deemed redundant or shed to free up money for investing in technology." Before we get to the, like, contrarian hook of this article, I just want to point out something that you have to be very careful about in AI job reporting. Look at what they said at the end of this sentence here. They listed some companies that said they cited AI saying, um, for job reductions due to elimination of redundancy or to shed or free up money for investing in the technology. That is actually at least, uh, in the last year, often especially with tech companies, when you- they say, like, "We've cut this many jobs due to AI," what they really mean is, "We've cut these many jobs not to replace with AI or because AI has made these jobs redundant, but because we want to take those salaries and invest them in AI to build... We want to get part of the, the land grab, try to make some money on AI products. But it's expensive to buy data centers and build data centers, so we're gonna take salaries and use that to invest in data centers so we can try to make money with new AI plays." So that's, that's just a careful inversion. Um, that's not always the case, but that, that, uh, especially the last couple of years, you know, Microsoft and others tried to spin job cuts where they're really just like, "We wanna shift money from this sector to AI" to make it sound like, "Oh, we let these people go because we're using AI." So you gotta be careful about it. But anyways, here's the contrarian hook of this article. But the job cuts of the AI, uh, area, uh, the AI era, which are affecting tens of thousands of people occasionally have an epilogue. Some companies, after discovering that AI models, applications, and agents don't perform as well as expected, end up rehiring laid-off workers to fill eliminated positions. They call this the AI boomerang. All right. So, um, this is... My reaction to this is, yeah, we would expect that because of the issues we just talked about. Because of AI, when AI is used to accelerate an arbitrary efficiency ratio, doesn't necessarily make your company produce more. And if you, you lay off someone that had something they did that was useful, and then to compensate for that, you took someone else who does, works on something related but something somewhat different, and then just increased his arbitrary efficiency ratio, that doesn't actually replace this thing that this person did necessarily. So yeah, this person is completing more tasks, but it doesn't mean that the core thing that this person did or their core expertise has been replicated. So I think this gets to the core difference between speed, like the Deloitte study found. There's a difference between speed and actual concrete effective things that are being completed. So, you know, it's not so simple of we have more tasks completed, you know, so we can reduce people and keep the ratio the same. Not all tasks are made equal. But these sort of arbitrary efficiency ratios don't try to differentiate. They're just like stuff is better than less stuff. All right, what do we got, what do we got next?
Speaker 51:45 Our second story is from The Washington Post, and it's about the new job numbers from the Bureau of Labor Statistics.
Cal Newport 51:51 Let's see what we got here. The title here is, "Is AI Taking White-Collar Jobs? Latest Hiring Report Suggest Not Yet."
Job numbers from the Bureau of Labor Statistics
Cal Newport 52:03 Economists have been asking whether AI is starting to take white-collar jobs, but the latest numbers out this week point the other way. Employers in professional and business services, the part of the economy that includes lawyers, accountants, and consultants, added seventy thousand jobs last month, the Labor Department said Friday. That's the biggest gain in several years. All right, so that's interesting. There's two ways to take this. I mean, the first way is just to say, "Hey, this is what we're talking about, is, um, uh, arbitrary efficiency doesn't actually measure things very well. AI tools are making that better, so just putting more AI tools into the workforce wasn't actually creating more value, so we still are hiring. It's not actually replacing people." And I think there's some truth to that. There's a, uh, another way to think about this. Let me throw in, like, a contrarian take. If we weren't using arbitrary efficiency, I think actually those job numbers would be worse. So in, in other words, there is a buffer that the white-collar worker has right now because we're actually really bad at, uh, measuring, like, what are we measuring? We're actually really bad at measuring that. We have these, like, tasks per person. It's, like, a very bad thing to measure. And if we're building tools to improve the wrong thing, it's actually, like, a little bit of a safety buffer to people because it's hard for a lot of these white-collar firms to cut down headcount because having the remaining people just complete task factor doesn't necessarily allow them to operate like they did before. If we actually were more effectiveness focused, I think we would be better at finding ways to use AI to complete specific value-producing things, you know, um, with less resources or less people. And that actually we would see more of a reduction in the numbers than we're seeing now. So this is sort of my, my new take is that I don't know how much reduction LLM power tools are gonna actually be able to cause in various sectors. It really depends on the sector. Um, I, it, there, there's no indication looking at these tools that it's going to be the sort of wipe-out that, um, the, the apocalypse have talked about. But used properly, if you're really focused on, like, this thing produces value and I have, like, worked on a custom system to, to make me better at producing that thing, and now I can produce, um, I can produce this thing twice as fast or do, p-produce as much as, like, two people used to produce, you would lose jobs. And so actually we're s- we're, we're, we're buffering partially job impacts because we're, before AI came around, we were already so bad at thinking about productivity in the workspace that we, we, of course, when we got this new tool, we're often using it for the things, the wrong things. And so if we come out of our stupor of Claude coding, you know, email summaries and actually start thinking about effectiveness and how AI can help it, there may be we're gonna see more probably, uh, instability in the job numbers. So it's interesting. Our own- Our own, uh, incompetence in terms of how do we manage knowledge work is actually protecting, ironically, is protecting knowledge workers perhaps. All right, what else do we have, Jesse?
Speaker 55:07 Next, we have a Stanford report about whether school phone bans make a difference.
Cal Newport 55:12 Yeah, this is an interesting one. Let's look at this together. This is from November 2nd. "Do school phone bans make a difference? A national study found benefits from school, school phone bans, but no simple fix.
Do school phone bans make a difference?
Cal Newport 55:24 Stanford researchers and collaborators explain how educators shaped the research and what questions remain." Let me just look at the brief bullet points here, and then we can talk about them. Um, they examined school phone bans that used lockable pouches, and they found that lower phone use, uh, led to higher reported well-being but had little overall change in test scores or attendance numbers. All right, so I want to push back on the framing from the, the headline that banning school phones is no simple fix. Simple fix? What are they trying to fix here? It seems like the standard is if this intervention didn't fix all things in schools, I guess that's a, a problem. I mean, how much-- Or do we also still have, like, body odor problems with teenage boys? Phones didn't fix that either, right? Is there still, um, do we, do we still have bad cafeteria food? I mean, the phones fix that either. I'm being a little bit facetious here, but this is a common framing when we're talking about techno harms and intervention. There's a common framing that happens because it... I don't know. It's a very common sort of academic commentator type framing where you're like, "Well, yeah, it did make these things better, but, you know, other things didn't get better, and so, like, what are we gonna do?" What's the metric there we're looking for? That 100% of things that you care about, if they don't get better, then we can say, "See?" We're kind of shrugging our shoulders on the one hand, on the other hand for an intervention. And I think we're kind of falling into that a little bit when it comes to the sort of very recent, uh, school smartphone bans. I think the fact that we're seeing well-being data jump up is exactly what we would hope and we should celebrate, that it was i-incredibly distracting to be on the phones. It makes, uh... You have all sorts of complicated emotional back-and-forth stuff happening during the school day, bullying, social posturing and positioning, all these issues. Take that out and just let the kids just be in school. School is better for them. They're happier. That's got to be the number one thing we care about. Now, what about test scores? Well, test scores depend on a lot of things. Like, I don't actually... I guess some people claim this, but I don't think the claim was, um, the fact that kids were looking at their phones a lot at various points during the school day was, like, directly affecting their performance on tests. There's a lot of things that goes into test scores and where they are. I mean, in particular, the kids are on their phone as soon as they get out the door. They're on the phones when they're studying. They're hurting their attention span with the phones and AI. They're doing less homework. You have a countervailing fact because if you're in a big school that actually can't correct for AI-based cheating, you basically just have large school districts now just like, "What can we do?" If kids want to cheat, they can get a perfect score on anything, any homework assignment. So, like, all sorts of things are also depressing test scores. You can't pull that apart. But we know the direct experience of being in school is better, so that is a, that is a great first start. If you want to make everyone smarter or make all test scores go up, that's always been very hard. We've never really found a good intervention, uh, that has done that outside of economic growth, right? If you're richer, you tend to have more resources and a co- and do better on tests. But it-- That is a white whale of intervention. So I'm actually very happy with those findings. We're making school more tolerable, and you're setting a cultural standard that's starting to make it easier for parents to keep phones and distractions away from kids longer. It's no longer this atomized thing where it's, it's me versus the world, and your kid says, "If you tell me I can't be on, uh, you know, TikTok when I'm 12, I'll be the only one. It'll be devastating." And you as a parent are like, "I just... This feels bad, but I'm just fighting this battle against you." When you have things like school smartphone bans, you can say, "I've got coverage here. Like, these things have such an issue that our school won't even let you take them out of the school. That's why I don't want you to have them. You're too young for it." It gives you cultural cover. It's the same thing with the social media bans, like in Australia. There's this whole line of argument that's basically saying, "But people are getting around the bans, so why are we doing them?" It's not because people can get around them. Hey, guess what? Uh, kids also found ways to find cigarettes even when you couldn't buy them in the store, right? The point was, if it's banned for kids, you have now given the ultimate trump card to the parents who care about this and worry about the impact of this on their kids. 'Cause when their kid says, "Why can't I have a phone at 14?" They can say, "'Cause it's illegal. Next question." No argument needs to happen. You've just made the kid's role now to try to argue for the parent, "We should break the law." Right? That's a hard argument the, the... That's a hard argument to, to win. So, you know, I think there's a lot of good that comes out of this, and I'm really excited to see already after just a year, we're almost immediately seeing kids' experience in school is much better, that they're happier, and that should just tell us, man, the more you can keep kids away from these and the longer you can keep kids away from these, it's got to be the better. 'Cause even in just this controlled piece of the day, taking the phones out of that piece of the day is landing strongly in the data, in classically incredibly messy type of data is we're already getting a strong signal. I think that's all good news, um, for cell phones, even if they haven't, uh, school bans, even if they haven't, like, magically fixed all problems with schools. All right, let's, um, let's do one more. What, what else do we got here, Jesse?
Speaker 1:00:43 Along these same lines, a new essay in Nature reviews early data on social media bans for kids.
Cal Newport 1:00:49 All right, so this one's a spotlight, so it's sort of a, a review essay, not a new scientific paper. It's titled, "Is social media changing children's brains? Here's what the science says." I read through
Early data on social media bans for kids
Cal Newport 1:01:01 this. Um, I won't read through the, the whole thing here also because it's probably firewalled on this tablet. Um, I'll give you the very short summary. It-- The way it opened at first, I got a little nervous because here's how it opened. "In 1972, sociologist and criminologist Stanley Cohen published a book called Folk Devils and Moral Panics, which explored the concept of moral panic, the idea that societies periodically fixate on a newly emerged condition, behavior, identity, or community as being a threat to the public." Uh, I get nervous when we casually apply the term moral panic when we're talking about something like, um, tech-induced harms. If you look closer at that book, when you're talking about moral panics, the examples given in that book are almost always things that are sort of clear, not just in hindsight, but to almost sort of any interested observer of the time, that the, the thing, the con-- source of the concern was illusory, and that's where the word panic comes from. Like, this is... If you really look back at that book, we're talking about things like, um, the use of that term is the, the, the moral panic around, um, satanic preschools, where we convinced ourselves at this preschool they were doing all these satanic rituals, or the, the moral panic a-around, um, razor blades in candy. Like, this-- We, we all have to care about this, oh my God, because, like, one news report, one place, we all then had to, like, x-ray our candy for a few years there in the late 1980s or whatever. That's not the right term to apply when we're talking about we have a sociological determinant here. We have, uh, observed harm that we're very concerned about, and we have evidence of a source, and so we want to investigate that and be careful. If this source is causing observed harm, and that harm is really concerning, then we want to pull it back. So when you actually have the widespread harm, it's hard to apply the moral panic framework because there weren't a lot of razor blades in candy. We kind of, uh, invented that. They weren't doing satanic rituals in the preschools. This was, um, the sort of repressed memory psychology that was, uh, unfounded, inappropriately applied. But teenagers are unhappy. They are having major social health crisis. They are decreasing in test scores. They are... Like, these things are actually happening. Now, we can argue about the various, uh, sources. There's different factors. We could argue about the various interventions. But the moral panic, uh, to me, this is like saying, "Look, heart disease exploded in the twentieth century," especially in the second half of the twentieth century, and we begin to really try to figure out and care how do we make heart disease better. We don't say that's a moral panic because it's a new thing that lots of people cared about. It's not a moral panic. Heart disease went up. That's bad. Lots of people were dying. We wanted to find out why. We found that diet and exercise played a big role. It led to some interventions. Also led to medical interventions that did a good job of reducing the death and, uh, the mortality that came out of, uh, heart disease, right? So I, I, I get nervous when I see moral panic thrown around when we're talking about something where we have a, a widespread measurable sort of self-evidently negative determinant. It's not a moral panic. It's a, it's a social problem that maybe has complicated sources and solutions. But this article redeems itself because it, it doesn't accept a priori that, yeah, this feels like a moral panic. It goes on to say, "Today, concerns about the impact of digital and social media on children and adolescents are never far from the headlines, but there is growing evidence that this isn't simply another case of moral panic. Politicians, religious leaders, and others have been speaking out with increasing intensity about how social media harms young people, and a growing volume of scientific literature is exploring the nature and prevalence of those harms and what might be done about it." Later on, it points out what I think is the absolute fundamental point in this debate. Overwhelmingly, in survey after survey throughout multiple different countries, right? So not just, like, the American context. The teens themselves are saying, "This is making me unhappy, this phone, these apps." And when you get the actual group itself saying, "Save me from this. I don't like this," I think we're out of the realm of razor blades in candy. So anyway, so the article, I think, lays out that, no, this is a, a serious worthy, uh, a topic seriously worthy of discussion. Um, I won't get into the science. It's a survey article. There's good science in there. Um, there was a, a... They note a study, so it's reviewing various studies. One interesting one that caught my attention was of, uh, eighty-three hundred students. It was longitudinal, so they, they, they actually studied them over time going forward, so not just, like, a retrospective study on, uh, data sets. And they found an association with social media use and inattention. More importantly, to my mind, half of the adolescents in this study self-reported that they believed they were addicted to these apps, right? So there was good science, and there's other papers in there. They're looking... I, I would say if I really had to summarize this, it's when it comes to self-reporting or psychological impacts, like you can measure, like, um, inattention or something like this, we're getting lots of signals from lots of different ways and lots of different experiments. Messy, but they're-- we get them from all different ways. When it comes to the neurological correlates, trying to study how the brain works, like what's actually happening in the brain, that's all new. We don't know yet. That's all still inconclusive. That's still new. So that's like the science takeaway is that the, the, the social psych is way ahead of the neurology and neuroscience right now. Uh, that's often the case. It's hard to study a brain. It's easier just to ask someone how they feel. Um, but I, I just thought it was worth discussing there is, like, be wary of being casual about calling things moral panic, and I think this paper did a good job of saying that's probably not the... This is probably not the right... That doesn't feel right. Like, we do need to look at the science, and it does, and it's mixed, but it's mixed in the way that, like, emergent literatures often are. Very strong signals there of what the students themselves are actually saying. This is a problem. I like to end with the third act of the show update with what's going on with me. Um I'll get in the books next week. I have been reading. I actually, Jesse, got to, um... I was at, like, you know, I do five books a month. I was, I'd finished four by, like, the 6th of November.
What Cal is up to
Speaker 1:07:18 Really?
Cal Newport 1:07:18 Because I just had a lot of ones I've been working on, and things just start, uh, some things just finished, and a couple things were easy to read. Uh, so I don't know, books are going fast this month. Um, two things I want to talk about. One is my new book. I, I'm gonna do an official announcement on my email newsletter at calnewport.com if you want to find out more. I might actually release a little short podcast episode in the feed about it. But I am sort of announcing the release date for my new book, The Deep Life: A Structured Approach. March 2nd is the release date. There is a cover. You can find out more at thedeeplife.com. I'll put this on the screen for people who are still watching us at this point. This is the official book website, uh, The Deep Life: A Structured Approach. This is what the book itself... That's the cover we've got, Jesse. Um, it's all about applying a systematic Cal Newport mindset to this sort of otherwise, uh, inspirational, aspirational, subjective idea of trying to create a meaningful life, and so that's why it's called a structured approach, as, like, what are the lifestyle algorithms? Like, how do you actually transform your life into something more meaningful? It's a book that does not focus on content in the sense of here's what your life needs to be meaningful. It focuses on process. How do you master the art of actually successfully changing your life for the better, whatever that actually means for you? Like, how do you figure out what that is? How do you break it down? How do you clarify it? How do you actually schedule and integrate into your life the work necessary to start making those changes? What changes do you make? What traps should you avoid? It's about the art of transforming your life, not a disquisition on the specific things you need in your life or the specific things that are holding you back. Why did I write that book? Well, I actually sold this concept during the pandemic when I began to notice it is difficult to talk to people about the deprivations of the digital if their analog is a mess. The actually, the best foundation from which to resist what Big Tech is trying to push on us is to make your own life so intentional and compelling and appealing that what they're offering no longer is that interesting to you. 'Cause if your life is a mess, and it's not intentional, and it's all over the place, why not lose yourself on TikTok? Why not have, like, this always friendly, reassuring, sort of semi-Elizabeth Holmes creepy companion on ChatGPT? Why not? 'Cause, uh, the world outside of that is, like, even more worse and chaotic. It's numbing. It's, it's reward hacking. But if you make your life intentional, you take the reins and say, "I want to live on purpose," completely changes your relationship to Big Tech. You say, "I... What I... Here's what I want out of you, Big Tech. Where you can make my life better, I'll use you. Otherwise, get out of my way. I've got stuff to do." And so I realized all these books I wrote about the dangers of the digital needed to be counterbalanced or supported by a book about the wonders of the analog done right. So I'll, we'll talk a lot more about this book, um, as we get closer. We'll do some episodes on it. But I just wanted to say, hey, I'm officially talking about it now. You can pre-order it, uh, at thedeeplife.com. You get the links. You can pre-order it on Amazon, Amazon UK, bookshop.org. If you want a signed copy, there's a link there to buy a signed copy from my local bookstore, People Books, here in Takoma Park, Maryland. I'm ready to start signing. Um, so we'll talk about it more, but thank you for your support. The other thing we want to talk about, Jesse, is the backdrop. We are, we are working on a new backdrop. It, I like the black, but it makes it really hard to do almost anything else in the studio. Because if I'm on TV, or I'm on someone else's podcast, or I'm doing, like, a Zoom meeting, it looks really weird when you're completely black. Uh, on those, it just looks like a little face in blackness while everyone else is in these well-lit rooms. We need a different backdrop. Um, the time has come. So we're working on it. I'm gonna reveal, Jesse, an early... For those who are watching, and if you're listening, you better go back to the YouTube version. Hidden behind a curtain is the, the early version of the backdrop that we have been building, and by we, it's, it's Javier.
Speaker 1:11:07 [laughs]
Cal Newport 1:11:08 Professional carpenter. I mean, it's we in the sense that we asked him, but I can't take any credit. I'll reveal some of this. All right, here we go, Jesse. Let's see here. All right, so it's not really in the shot. Ah, there we go. Uh, there's a peek. There's stained paneled wood panel shelves. We're gonna put my books on it. We're gonna have a light on there. I don't know what else I'm gonna put on there. But we are moving towards a, a beautifully built custom stained wooden backdrop with shelves and a, a small amount of things going on. Um, so stay tuned. We're getting there. Things be a-changing, Jesse. Uh, I was gonna do a Joe Rogan style neon, but with my name instead, just, like, two feet behind me, just vvv, like aggressively, aggressively glowing. So things are up. All right, that's enough time for today. Uh, thank you for listening. We'll be back next Monday with another flagship episode. We've been doing a bunch of Thursday shorter episodes too, so stay tuned for that as well. Check out the book at thedeeplife.com. And until next time, as always, stay deep. Today's episode was produced and mastered by Jesse Miller. Nate Mechler provided research and newsletter support. The video was edited by Ben Crane and Rob Monks. The theme music is adapted from an original composition by Jay Kerstens. For weekly ideas from me to show up in your inbox, sign up for my newsletter at calnewport.com.
Episode duration: 1:12:31
In the uListen app, the transcript lights up word by word as you listen. Get uListen
Transcript of episode WzkQUKmR1Ug
