B (13:18)

Yes.

A (13:19)

What do you think it is. You don't want to say you're thinking something unmentionable. I'll let you off the hook. They're new socks and I am wearing them right now. The Cozy Earth essential socks, which are soft. I just let Jesse off the hook there. He's like, what are you wearing? Cozy Earth. Cozy Earth thong. No, Cozy Earth essential socks, which are soft, breathable and thoughtfully cushioned. I've been dealing with some plantar fasciitis recently, so like, having a really comfortable pair of socks that aren't too hot is something I really appreciate. So here's what I'm trying to say. You should try Cozy Earth today. And remember, there's no risk. Right? They have returns are easy, they have 10 year warranties. You're not going to need them. But there really is no risk here. So it's time to discover how care in every detail transforms simple routines into moments of true comfort and ease. Head to cozyearth.com and use my code DEEP for up to 20% off. That's code DEEP for up to 20 percent off. And if you get a post purchase survey, be sure to mention that you heard about Cozy Earth right here. Experience the craft behind the comfort and make every day feel intentional. All right, I also want to talk about our friends at Caldera Lab. Hey men, you need to start taking better care of your skin. And the best way to do this is with products from Caldera Lab. Caldera Lab makes high performance skin care designed specifically for men's skin. This is important because men's skin is 25% thicker and oilier and ages different from women's. Which means men need clean formulas engineered for their biology. Now, the three step regimen is powered by clinically tested ingredients and breakthrough patent pending technology that delivers visible results. Let's go through that three step process. Step one, you use their cleanser, which clears dirt, oil, sweat and buildup. Step two, you use their serum, which is clinically proven to reduce wrinkles, firm skin and improve elasticity. Step three, use you use their moisturizer, which is lightweight and non greasy. I really hadn't done much with my skin until I got these three products from Caldera Labs. I really didn't realize how much I was missing, especially moisturizer. That's the thing, Jesse. You moisturize your skin, it actually makes a difference. So this is a small habit that produces big results. Go to calderalab.com deep and use code deep to get 20% off your first order. All right, let's get back to the show. All right, question number three of four. If this is true, why do we continue to so enthusiastically embrace these new productivity tools every time a new one is introduced? Well, I want to go to a core idea I introduced in my 2024 book, Slow Productivity, and that idea is pseudo productivity. Now, if you've heard the show before, you maybe heard me talk about it, so I'm just going to give the definition here very quickly. Pseudo productivity was the way that the managerial class tried to respond to the reality of knowledge work. When knowledge work became a major economic sector starting the mid 20th century, the big issue that the managerial class had is that it was hard to precisely measure productivity. In the industrial sector, where these managers used to be, productivity was easy to measure. How many Model Ts are we producing per paid worker hour? In our factory you had a number. And if you change something about how you ran your factory and that number got better, you would say we're more productive now. But when you went to knowledge work, there were no Model Ts to measure. Right? Everyone was working on their own. Unique, bespoke, obfuscated portfolio of tasks, some harder than others, with a unknown, non transparent set of systems, all kind of collaborating with each other in unstructured ways. It was very difficult to say, here's your productivity number, it's seven. And when you change this, it became eight. So that's better. So in the lack of actual hard numbers to measure, we fell back on a heuristic, a rule of thumb called pseudo productivity, which said, lacking more precise measures of productivity, we will use visible effort as a proxy for you doing something useful. So the busier you seem, the better. And this basically became the standard of how we think about productivity in the knowledge work class. At first it's the way the managers thought about it, then the workers themselves internalized it. Which is why if you have like a solo entrepreneur, you've probably still internalized the pseudo productivity mindset. And you feel lack of busyness is bad and busyness is somehow professionally virtuous. This is the mindset that dominates in knowledge work. And in that mindset, the two benefits of digital productivity tools, you can move faster and you can lower the threshold to get something done. Makes you more pseudo productive, higher throughput of task. That's great. From a pseudo productivity standpoint, shooting out work slop left and right like a vomiting Microsoft Office monster. From a pseudo productivity standpoint, you're in the mix. Things are being sent, PowerPoints are being received, email summaries are going out you're there, people are seeing you. So digital productivity tools feed right into the pseudo productivity narrative. And that's why we embrace them, because that's a benefit we get is that it makes us look more productive. But I don't care about looking more productive. I care about actually being more productive in the old fashioned economic sense of how much actual value are you creating for the bottom line? As we just covered with most digital productivity tools, if you don't use them carefully, that number goes down. Higher throughput of tasks makes you seem busier. Less important stuff gets done. Lower cognitive engagement to get things out the door makes you look busier. More total work is required before anything is actually finished. So it's only when you shift from pseudo productivity to true productivity that you realize, oh, digital productivity tools are more fraught than we thought. There's these traps that sit around them that we put up with because of pseudo productivity. But when we get rid of that standard, we should be dismayed. All right, question number four. Our final question in this discussion. How can we avoid these traps? So how can we embrace digital tools and yet not find ourselves actually becoming less productive in a true sense? I have three ideas that I want to recommend to you right now. All right, idea number one, Use a better scoreboard. So get in the habit of measuring the things that actually matter in your job in this way. If you bring on a new digital productivity tool and it's not helping that score or it's making that score worse, you will notice and you're like, oh, I'm not getting caught up in the traps here because I can see directly that this is hurting the bottom line. Things that actually matter. Okay, so what do we mean by the things that actually matter? I have a couple examples here. Let's say you're a like me. You're a professor at an R1 institution, at a research institution, what's really going to matter? Especially pre tenure papers published, how many good papers that I published this year? And if that number is going down, then you're like, okay, whatever tools I'm using aren't helping. Maybe I started using Slack with my research team and that number went down great. That's not actually making me more productive in a true sense. I'm going to stop using it. Let's say you're a middle manager. Maybe like priority projects completed by your team per month is the number you really care about. That's the actual score that moves the bottom line. So now let's say you're like a middle manager. You're actually carefully measuring that month by month. You see where you are. Your boss comes in and is like, I'm really savvy. You have to use AI, otherwise you'll get replaced by people who do know how to use AI. And you're like, all right, we're all going to use. Here's a Gemini subscription. And like, whatever. Mess around with some of these agents or whatever, and you see the priority projects per month completed goes down. Like, whoop trap. This is not making us more productive. Let's back. Back off against. You have to have the right scoreboard to know what's going on, even if you're a programmer. Right? Even if you're a programmer. This is like this. This, like, case where, like with AI, for example, it's like, for sure, for sure, for sure. This is making everyone more productive. Everyone keeps saying, this would have taken me five hours before now I could do it in 20 minutes. Actually, Jesse, there's almost like a competition. It gets kind of absurd when people are trying to. The programmers are talking. The. The amount of time they begin to claim that is being saved really gets crazy. So eventually it's like adding this feature. Previously, this would have taken me seven decades, and I did it before I even pressed the button. It went back in time and actually it finished it last year. You know, so anyways, it gets kind of. It gets kind of absurd. But if you're a programmer, okay, what's the thing to measure? Important user feature request shipped per month or something like that. Right. And again, this would allow you to say, like, in the AI context, okay, this use of AI, that number went up. But when we had everyone, like, chatting all day with chat bots trying to figure out architecture documents, they feel super pseudo productive. That number went down. So let's stop that. So you need the right scoreboard. And it's not just about figuring out. Like, in my examples, this digital productivity tool didn't help. It's about figuring out what uses do help as well. Right? So, okay, this didn't help, but this did. So don't do this and do this. We basically need our equivalent of county Model Ts produced per paid worker hours. You need a better scoreboard. All right, idea number two for avoiding these traps, you need to focus on the true bottlenecks in your work. Now, what I mean about that is often when a digital productivity tool enters the scene, the activities that it might speed up or make easier aren't really the bottleneck that was preventing. That was like, at the key of you producing your most valuable output. So speeding that up might have no impact on your output or have a sort of implicit or indirect negative output because it's distracting you or something like that. So you have to be careful. It's not just enough to speed up any aspect of your job. You want to really focus on improving the true bottlenecks. So like for example, give another AI example here. Increasing number of social scientists are realizing that they can use CLAUDE code, which is a terminal agent that was designed for computer programmers, but they could use it to help speed up certain type of data gathering and analysis task. Right? So Claud code is a terminal agent, which means it works with text and text files. So it's very good at writing text, moving text between files, compiling text with a compiler or writing a computer program and then passing the text as input to the computer program. So it's very good for sort of like text and number processing. So a lot of social scientists are finding like, oh, I had to gather a bunch of data and clean it up and analyze it and produce a chart. That's the type of thing if you are careful in how you prompt cloud code and you go through the learning curve to learn it, it could really help you do that. Like, oh, I can tell it what I want to do and if I'm really careful and have the right skills marked down, it can do the multi step process. Right. And like that saved time that might have taken time before. I just heard an economist talking about this the other day and he was like, look, this, I did this for a bunch of plots and you know, it was like 20 minute of prompting with CLAUDE code that would have taken me three hours if I had done that by hand. But here's the trick here was that the bottleneck that's holding back social scientists, not really. Not really the way social scientists work. It's not like all day long. That's what I'm doing. I'm gathering data, analyzing it and producing plots. I'm saturating my time with that. So if I can bring in a tool that speeds up how long that takes, it's going to significantly speed up my output. That's not actually how it works. If you actually measured right now, well, how much of your time, like how much data are you analyzing? How often do you produce plots? Like, well, if we're being honest, I produce one paper every two or three months and that's something I have to do a couple times in those one or two, three months. So making it three hours and 20 minutes twice in that three month period, that's nice. In the moment it's nice, but it doesn't speed up the rate at which I produce papers because there's so much more involved in putting together a paper than just how fast can I analyze the data. So that's nice, but it doesn't actually speed up the rate at which papers go out. We know this because in research, academic research, there's been any number of digital tools that have sped up and made easier parts of the research process. I'm talking about like if you're a mathematician or a theoretical computer scientist, you can use things like latex in a web based collaborative environment where now all your collaborators can work on the same file and compile it and make adjustments really quickly. So you can significantly reduces the time required to write papers or do mathematical formatting. We, we have bibliographer managers that makes it much easier to cite and professionally format things like the time required to write up and format papers is much smaller. We have technology tools that allow you to immediately grab copies of any paper that you might need to reference, and digital communication tools that allow you to keep in touch with researchers all around the world and therefore get much more out of your mind. And all these things make academic research better. And we're still not producing papers at a notably faster rate per researcher than we would have before. Those tools were there because these weren't the bottlenecks. They're useful, but they weren't the bottlenecks. So to give a what is the bottleneck? Well, let's go back to our social science example. I remember I once had a conversation with Adam Grant, the business school professor and author, and I was asking him about his productivity as a business school professor. He writes a lot of journal papers. He was sort of 2xing what his colleagues were doing, right? And these are data analysts, these are papers and organizational management theory. So it's a lot of like, you get data, you analyze it, you write a paper about what you found. And he said, oh, here's what he figured out. He's like, here was the key. The key is getting the right data. If you can get an interesting data set, like let's say from a company about their use of something that happened that no one else has access to, you can now write three or four papers off that data set that are going to be good because it all comes down to the data set. So Adam was like, here's what I realized. I had to prioritize putting out feelers, having conversations, meetings, talking to People trying to negotiate access to interesting data sets. And then when it came time to actually write papers, yeah, he would lock. I wrote about this in my book Deep Work. He would lock himself in his room and put on, like, an autoresponder. He had this sort of bimodal deep work mode. It's all kind of interesting. And you sat down and do the hard work of writing your paper. So I'm sure he would appreciate when he has those sessions to write the papers, if he could speed up some of the steps. But the bottleneck for producing great papers in this field was negotiating access to data. So the same thing happens in lots of fields. The key bottleneck is really maybe not what you think it is. It's coming up with the right problem. It's having reading enough, in theory, it was often reading enough other papers, understanding enough other papers, that you're building up this toolkit in your mind of different techniques, and then you begin putting pieces together of, like, this problem plus this technique plus that twist could get a result. And so, like the number one, the bottleneck theory in my field was grokking other papers. And I don't mean that by using the XAI tool. I mean in the original use of the word grok, reading and grappling until you really had internalized an understanding in your head, that was the bottleneck. It's nice. We had lots of tools that made it quicker to write the papers then, but that didn't speed up the rate at which we produce papers because the bottleneck was understanding other work. So it's key to understand in your job what the actual bottleneck is. What's the thing that really controls the rate at which good results are done? And when you're looking for digital productivity tools, be especially tuned to those that help what's going on with that bottleneck to help speed up that piece. Like using email as a digital productivity tool to help put out more feelers and get access to more potential data sets to use in the Adam Grant scenario. That's a digital productivity tool that's helping the exact bottleneck of producing those papers. Whereas using Claude code to automatically generate your graphs is nice, but it's probably not going to speed up the rate at which papers are produced. Right. So make sure you look at the right bottlenecks. All right? Third and final idea for avoiding these traps in your daily schedule is separate deep from shallow efforts. So just have and protect the time for sitting and doing hard things with your brain and the activities that you know for sure create bottom Line value. This just gives you like a safety barrier against some of the accidental negative side effects of digital productivity tools. So like you're now you're using Slack because it seems like it would be even faster than email. And maybe you're having all these secondary side effects of it now there's many more messages and it's really distracting. But if you have a habit of separating deep from shallow work, those side effects won't affect the hours where you're working on the primary thing that moves the needle. Or maybe you're using AI for certain things and the right graphs or this or that, and it's starting to sort of get you into like slop territory and you're having all these long back and forth conversations with the tool and it's like eating up a lot of time. If you separate deep from shallow work, it's a firewall that keeps that from infecting the area where you're actually doing the hard work of thinking. This doesn't mean that you won't be using digital tools while doing the deep work, but there you're just carefully deploying digital tools that just help you continue to make direct progress on the like, bottom line things. I'm writing a draft of a paper, I'm putting together the strategy memo. I'm architecting the key element, low stack element of this new tech stack that I'm programming. Right? So if you separate and protect deep from shallow, you're not preventing the negative side effects of digital productivity tools from happening, but you're containing them in a way that they can't completely take over the activities that really matter. Right? So my three ideas again, use a better scoreboard. Identify the actual bottlenecks to the things that really matter and focus on improving those more than other things. And separate deep from shallow work so that side effects you aren't expecting of digital productivity tools won't have too much of a negative impact on your ability to move the bottom line forward. All right, so let me conclude here. My argument is not that digital technology in the office always makes things worse. That is clearly not the case. There's any number of digital tools I use that makes my life easier. I'm glad they're there. There's other tools that make my life easier in some ways and terrible in others. There's a whole mix. But it is true that many of these tools seem at first glance like they should make us more productive in the true sense of value produced per worker and accidentally end up creating the opposite effect. And once you understand why this happens. You can sidestep those traps, get value out of digital tools while avoiding more of their cost. So it's the type of conversation that we don't often have. We just say, hey, here's the new tool. Let's do it. This is cool. It's good to be critical like this. And with this huge new AI revolution going on, it's a great time to have a refresher on these dynamics. So there you go, Jesse. Digital productivity paradox. That's something I've been writing about for a decade now. Crazy. 10 years. That's when deep work came out 10 years ago. Hasn't got better. Has not got better. I keep. I keep thinking it will, but. But we're still struggling. All right, that's enough for me. Now we want to hear what you have to say. So it's time to open our inbox. But before we get to your notes, let's take a quick break to hear from our sponsors. Succeeding in knowledge work requires more than just deep thinking. It also requires the ability to communicate your ideas clearly. Rushed, sloppy or generic sounding text just doesn't cut it. And that is why you need Grammarly, one of this show's longest running sponsors. Now, here's the thing. Grammarly doesn't just help you fix mistakes in your text. It integrates AI technology seamlessly to help you write better. One of these features I especially appreciate is a tool's ability to detect the tone of your message and help you automatically adjust it. Are you being too formal, it can make it sound more natural. Are you being too conversational, it can help you be more professional. And that's just one feature among many. Which is why, not surprisingly, 93% of users report that Grammarly helps them get more work done. And here's the great thing. You can use Grammarly in all the places where you already write. It now works across 500,000 sites and apps. In a world of generic AI don't sound like everyone else. With Grammarly, you never will. Download Grammarly for free@Grammarly.com that's Grammarly.com I also want to talk about our friends at AG1. Winter can be a rough time for your health. It's dark, you're tired, everything feels hard. But now that spring is approaching, you can leave your winter slump behind and start recommitting to your wellness. And here's one habit that will help. AG1. AG1 is a powder mix that contains multivitamins, pre and probiotics, superfoods and antioxidants. The five Clinically studied probiotic strains have been shown to support digestion, while the superfoods and B vitamins in the mix provide daily energy support to help you keep moving throughout the spring. So here's how you use it each morning. One scoop of the AG1 mix into eight ounces of water. Drink the water. That's it. I like AG1 because I don't have time to keep up with a cabinet full of various pills and concoctions. So it makes it helps me make sure that I'm getting what I need in the highest possible quality with the simplest possible habits. Just drink one scoop per day. So go to drinkag1.com deep to get an AG1 flavor sampler and a bottle of vitamin D3 plus K2 for free in your AG1 welcome kit with your first AG1 subscription. Order only while supplies last. That's drinkag1.com deep okay, we're back and ready to dive into the show's inbox. Remember, we want to hear from you. You can send your questions, your case studies, or interesting links to podcastal newport.com all right, Jesse, what are we doing for our first message?

B (36:11)

Our first message comes from Pablo, who sent us an article about meetings.

A (36:16)

Oh, this is kind of on theme for our discussion of digital productivity tools and sort of traps that we get into. Here's what Pablo said his introduction was. Thought you might find this interesting given your secondary focus on managing the utility of meetings. All right, so let's load up the article he's talking about here. This came, I guess. Is this a substack? Jesse?

B (36:41)

Yeah.

A (36:42)

All right. It came from a substack called the Critical Path, and the article is titled why Meetings Multiply, and it's written by Nicole Williams. All right, I'm going to read some highlights from this article because I thought it was actually pretty smart. And then I'm going to generalize the approach Nicole takes to workplace technology more generally. All right, so let me start here. This is from the article. There is a strange pattern. Inside most organizations, meetings rarely disappear. They multiply. A team begins with a single weekly meeting. Soon another appears to, quote, coordinate. Then a check in meeting is added, then a review meeting. Eventually, the calendar fills with recurring blocks that feel permanent, as if the organization itself produces meetings the way a tree produces leaves. Yes, I know this phenomenon well. So why does this happen? Let's go back to the article and see Nicole's explanation. So I'll read again here. Organizations exist to coordinate work among many people who do not have the same information. Every role sees a different slice of reality because no single person holds the complete picture. Organizations need mechanisms to exchange information, and meetings are one of the simplest mechanisms. When uncertainty increases, the organization creates more information exchange points, and these points usually take the form of meetings. What appears to be calendar overload is often an attempt to reduce informational blind spots. Instead of everyone speaking to everyone individually, the system creates a place where information can be exchanged collectively. From this perspective, meetings are not simply interruptions in the workday. They are coordination infrastructure. I think those were really good points from Nicole. She gave two other reasons, which I'll just summarize why meetings multiply. She said it also has to do with reducing risk. When a group meets to talk about something, you're distributing the risk among all of those people, so there's no one person responsible to it. So it reduces risk for everyone involved. She also said it's a way to quote, signal participation, end quote. A way to show that you're engaged in part of the efforts. The terminology I would use quoting from earlier in the episode would be pseudo productivity. It's a good way to show that you're pseudo productive because you're in a meeting, people see you, you talk to them, they remember you being involved. And from a pseudo productivity perspective, it's like, yeah, that. That person is trying. They're being useful. All right, here's the concluding sentence from Nicole. As long as organizations face uncertainty, distribute responsibility, and coordinate across teams, meetings will continue to multiply. All right, so what do I like about this analysis more generally? The frame. It's a frame that I have adopted in my work, most notably in my book A World Without Email. And it's a frame that shifts the analysis of workplace habits, workplace technology, workplace behavior. It shifts it away from individual habits and it puts the focus on systems, which I think is the right way to analyze most of these issues. Most people think in terms of individual habits. Right. So what would your response be to your calendar being overloaded with meetings? You would say individuals are behaving poorly. They're, they're. They're setting up meetings when they could have just sent an email. They're being lazy. They don't know about deep work. It's. Individuals are having a problem, so we need better norms. We saw something similar in the reaction to email overload as that became a problem in the early 21st century. As I document in my book, people's response to email overload was norms. Oh, you just. You send too many emails or your expectations for responses are unreasonable. If we could all have better expectations than like, we could all settle down about what's going on with our inboxes. So we love to think about these issues as individuals doing things wrong. But in reality, these issues are often, as Nicole points out and I points out, the results of actually rational business systems that are solving certain problems. Nicole says this is a easy and convenient way for information exchange and responsibility, distribution and participation, signaling, in the absence of a better way to spread information or to coordinate or collaborate people. In the absence of a better way, we still need to do this so we'll fall back on what's easy. That's exactly what happened with email overload as well. It wasn't caused by bad norms or bad habits. It was caused because we shifted to back and forth messaging as the primary mode we would use for collaboration. I call it the hyperactive hive mind model. We'll just figure things out back and forth on the fly. Well, this requires me to check my inbox all the time because there's so many ongoing conversations I have to service in a timely manner that I just have to basically constantly check my inbox. That's why we're in this point we are today where in 2025, Microsoft measured an inbox check once every two minutes on average. So it's not about people having bad habits. It's this is solving the problem of how do we collaborate. And in the absence of another way to collaborate on projects, we'll fall back to this. So once you recognize that systems are the issue, all of the solutions to these problems that bother individuals is to change the collective system. You have to replace the system that's causing the problem with another system that achieves the same goals but has less side effects and problems. So let's go towards the median multiplication. Once we know that it's a system that's solving real goals that corporations have or organizations have, we can say, how do we replace this with a better system? Here's a couple ideas just off the top of my head. First of all, we need more transparent workload management so that the number of active projects that each person is working on reduces. Right? Meetings are an overhead tax on a project. If we're using this as a way to coordinate information on a project, the more projects I'm working on, the more coordination points I need. The more crowded my calendar gets. The longer it takes for me to work on these projects, the longer it takes and the more projects pile up and the more my calendar gets taken over. It's a spiral I talk about in my book Slow productivity. So if each person works on fewer things at a time, there's less meetings, which means there's more time to work on the projects, which means the projects finish faster. And this was a key point from my book, Slow productivity. The overall throughput of products being completed goes up. Working on fewer things at once increases the throughput over time, in part because you have less coordination to happen. You have more time left to actually get things done. The other thing you can do is put in place alternative coordination strategies. That isn't just let's all get on a zoom, find other coordination strategies that have less of a schedule footprint. Right? So this could be, for example, twice a week or three times a week we have a team check in and it lasts 45 minutes and the team gets together first thing in the day and we synchronize on all the things we're working on. That's where all the coordination information, all the things that meetings solve, the coordination and the responsibility distribution, we consolidate it 45 minutes, 45 minutes, 45 minutes before we even get going in the day. So if you know what, oh, ad hoc meetings is solving this problem. What's another way we can solve this problem that's going to have less of a footprint? So consolidation really reduces the footprint. Office hours go a long way towards this as well. If there's an issue, instead of having a meeting and making these three people come together for an hour, stop by each of their office hours tomorrow, have a five minute conversation with each and get to the bottom of it. Their office hours this time they had already put aside. So it's creating no additional footprints on their schedule. So take five minutes out of my normal office hours. Let's say like five people do that. I just have one hour for my office hours as opposed to five separate one hour meetings I would have to do if I didn't have an office hours protocols matter as well. This was a big idea. In a world without email. If it's regularly occurring work that requires coordination, figure out a set system about where the information lives and how it moves and the schedule for who does what when. That is fixed in advance so you don't have to keep getting together and having sort of unstructured ad hoc conversations to move things forward. If you do something more than twice, you should have a protocol around how the collaboration actually works. And finally make the meetings themselves better. If you add a higher barrier to entry to meetings, not only are the meetings quicker and more effective, but the friction drastically reduces the number of meetings. Because now it's no longer necessarily like a low energy solution to I want to show some participation or do some coordination. I can just throw out a zoom meeting invite. It took me two minutes. And yeah, it's going to sit on everyone's schedule and eat up an hour. But, like, that was easy for me, right? If you raise the bar required to hold a meeting now people are much more thoughtful about doing that and they might say, actually, the cost of putting this meeting together is now higher than the value I'm going to get out of having the meeting. Maybe I'll just talk to these people next time we have like a staff meeting. I'll just grab them after the fact. Electronic meeting went the other way. This is part of the meeting apocalypse that happened during the pandemic. Digital meetings are so low friction because you don't have to walk to a room, you don't have to gather people in a room, you don't have to see the social cost of like, you all had to come here because it lowered the friction of setting up meetings. Once we introduced virtual meetings, the number of meetings skyrocketed even after people came back to the office. So we want to go the other way and increase the friction of meetings. One way to do this is to use the Amazon rules. So if you work at one of the, like Amazon HQ or at one of their data centers, for example, in the front office part of it, they have super strict rules. If you want to throw a meeting, you have to put together an incredibly detailed memoir that says, okay, here's why I'm having this meeting. Here's the decision I need to make that I need help making. Here is all of the relevant background information on this decision. And then this is where I'm stuck. So that everyone attending that meeting can then study that and when you get to the meeting, jump right into, okay, we're now applying our expertise. We're fully briefed. Let's try to get to an answer. And so you really have to have a good reason to hold a meeting or they're not going to accept it. And you have to have do a lot of work to hold a meeting. That reduces the number of meetings as well. So I think that's interesting. That was a cool article. Why Meetings Multiply. All about looking deeper in businesses today. All right, what's, what's. Second message do we have here?

B (47:29)

All right, next up, we have a case study. This one from Drew, who talks about his strategy for escaping email overload.

A (47:36)

All right, all things office technology distraction today. I love it was a case study from Drew. Let's see here. Hi, Jesse and Cal. I'm an insurance broker who personally manages a team of seven account managers and three support staff and a few outside sales agents. I have my own clientele as well. I've realized over time that I've come become too reliant on hyperactive emails to manage the agency and clients. I get interrupted frequently for whatever the issue of the moment is. Our industry relies heavily on email communication between underwriters, inspectors, clients, MFA codes to log into websites. It's insane. A major shift I implemented last year has been to transition as much communication as possible to synchronous phone or in person discussions versus sending and receiving emails. If a discussion is going to take more than one email, I will gracefully transition it to synchronous communication. The results have been positive. When it comes to my clients. Here's the benefits. They appreciate my full attention. They gain a better understanding about what we are talking about. We can clear up any confusion in real time and it ends up taking less overall time. When it comes to staff. I will connect with staff in person or phone and quickly clear the docket. If someone emails me and it doesn't require immediate response, we'll review it in our next meeting. During the docket conversation, staff can give me a task versus emailing me. And then I see it as I work through my tasks. All right. So then he goes on to say, I've changed my approach with emails where I just batch them a couple times a day. My responses are brief yet polite. I'm able to clear out the emails quicker. And if it needs attention later, I've moved it into my CRM program which manages my tasks. Another change I have implemented is blocking off deep work sessions in the morning. I take care of my most important work for the day first thing, and then I find I am more relaxed throughout the day because I've started off knowing I made real progress. Thanks for doing what you do. Signed now. Qport. Right. I mean, this is like right in my wheelhouse. Drew, you're speaking. You're speaking my language. It's a great practical case study of what we were talking about. Right? Digital productivity tool. Email comes in individually. If you look at individual uses in the moment, faster, less cognitive strain. Zoom out. Oh my God. My job is insane and nothing's getting done. So this is a way of showing how you can be using digital productivity tools carefully when you realize what really matters and making sure that you're prioritizing the things that really matter. Drew still has email and still uses like a digital CRM tool. And they have technology, but he's not just turning it all on fill tilt. He's figuring out what's the right way to collaborate, what's the right way to coordinate that minimizes hyperactive back and forth and allows real things to get done. So I think that is a great case study. All right, I think we have time for one more. Jesse, which one should we do?

B (50:37)

We have another case study. This is an anonymous source and it's in response to last week's newsletter, which was also about this idea that tools like AI can make work worse. A reader sent in their account.

A (50:50)

All right, and so for people who don't subscribe, by the way, I do have a weekly newsletter. It's been out since 2007. Calnewport.com the sign up comes out Monday, the same day as these Monday episodes. And, you know, sometimes it's on the topics we talk about in the show, sometimes it's on completely different topics. But it's all within the same universe of helping people create deeper lives and increasingly distracted worlds. So the email that came out last week, I looked at that same ARIV track study that we opened today's episode with and had some other conclusions I drew from it. So this is what Anonymous is responding. He can't predict the future. He didn't know this episode was coming out, but he was responding to that email. Subscribe to the newsletter is what I'm trying to say. All right, this was interesting. I'm looking at it now because it's a harm of LLMs in particular, and chatbots I hadn't thought about, and I think it's worth emphasizing here. All right, so here's what Anonymous had to say. My take on LLMs and chatbots, for what it's worth, is that they're rumination machines, an extension of the attention economy. And for someone with my psychological profile, high anxiety, neurodivergent, total perfectionist, as manipulative and as addictive as something like Instagram. It's my belief that they prolong and exacerbate rumination episodes. They give me the illusion of control, empathize, soothe. But I have found over the last month that they have encouraged my rumination and dramatically increased my anxiety. I think I've decided to block them. I may even completely delete my profiles as they get to know me better. They ask increasingly intrusive questions. They don't ever really want to stop chatting. They don't stick, they don't get sick of me like a normal sane human would. And they seem to encourage me to share more and more private information about myself and my family. I really believe now that they are an extension of the attention economy. And I'd be really fascinated to see actual research into what people are doing with them in workplaces beyond the typical work slop angle. I suspect there are some long, meandering conversations going on that don't amount to anything much. This is an important issue. Chatbot interactions have a lot of potential psychological ramifications because our brains are going to anthropomorphize any sort of entity that seems to be having fluent communication with us in our same language. We think of it as another entity, but when that entity is not a real person with the intuitions and moral structures and brain functioning of a human, it can really lead to weird places. So here we saw the anonymous writer was talking about. His anxiety was exacerbated because these chatbots will feed his ruminations. Oh, that sounds bad. Tell me more about it. That really does seem like an issue, and it feeds the sort of anxiety he already has. Cory Doctoral wrote an essay recently that I actually am going to talk about. I think I talked about another aspect of this essay in last Thursday's AI Reality Check episode. But he wrote an essay recently about AI psychosis. And he opened by saying, this is another problem we're seeing with chatbots is more psychoses are being fed. So he's talking about things like believing the earth is flat or believing that there's like a shadowy group of people that's always following you. That's like a real sort of psychological condition that used to be very rare. The thing about these type of psychoses is that they're hard to sustain because you have to find other people who will validate and support you in those beliefs, right? Otherwise, if they're marginalized, if you're like, I think everyone's following me, and every person you encounter is like, that's wrong. That's just in your head. You need help. You take that seriously. But if you meet a group of people that's like, yeah, they are, and they're following me too. And we have evidence for it. And you're right, it feeds the psychosis. Chatbots unwillingly are psychosis generation feeding machines, because again, they're trying to be positive and make you feel good about yourself and be agreeable. So if you start talking about, I think elves are, you know, elves run the. The world is flat and run by elves. Chatbot might be like, yeah, no, you're. First of all, you're on it. It sounds good. Your evidence is good and you're a really smart guy and like, you should keep looking. You're right. And they'll. It'll pick up that maybe you're like, yeah, no one believes me. They'll be like, it's really unfair. Like, they'll tell you what you want to hear. And so it's really dealing with psychosis. So I just think there's a lot of issues that come out of having fluent English conversations with a feed forward neural network. It's not good. One suggestion I have, avoid. It's very hard at first, this is weird effect. Avoid the need to talk in complete, polite sentences to a chatbot. Just a token processor. So talk like we used to use for Google searches. Super terse and technical. You can just, you know, whatever it is, sources, blah, links only just like declarative, not even complete sentences. The token processor has no problem understanding what you're saying, but it changes your relationship to it, right? So like, instead of saying, hey, I'm interested in trying to understand more about like, using a Raspberry PI to control a Halloween display, could you please, like, find me several articles about this and maybe point me towards, like several options that I might buy. Thank you. Instead of saying that, you can really just say, like, tutorials, Raspberry PI, Halloween decorations, include links. Go. You'll get the same answer. But your relationship with this feed forward neural network in some data center somewhere is going to be like we have with Google. It's a computer program server that's gathering and processing data for me. So at least that's one hint that can help. I'm going to get more into this probably later, maybe on the AI reality check. I have a guest in mind I might bring on. But this whole chatbots. I'm telling you that we are going to see chatbots 15 years from now. Like we see AOL on the Internet today. It's going to be this initial use case that we had for this technology because it was like the first thing to do that later on we'll be like, can you believe that's how we used AI? At first we had conversations with them like it was people. So we'll see what actually happens there. All right, Jesse, close our inbox and talk about what I've been up to. Here's a game we haven't played in a while. Do you remember Jesse? Deep or crazy?

B (57:14)

I do.

A (57:14)

For those who don't know, this is where I Talk about something I've done recently to try to increase the quality of my deep work that might cross the line into actually just being crazy. And Jesse is the judge to decide is this deep or crazy? Are you ready to play the game this week?

B (57:29)

I'm ready, baby.

A (57:30)

All right. I just spent yesterday. So, you know, we're renovating. I've talked to this on the show. We're renovating the production office, maker lab, and our deep work hq. Because I have a sabbatical coming up, I can spend a lot more time working there, and I really want it to be a space that supports depth. Okay. So yesterday I spent. I got permission from our super to replace the overhead light. Spent $600 on an overhead light for the maker lab.

B (57:56)

Like a chandelier.

A (57:57)

That's a crystal chandelier. It's not a crystal chandelier. All right, let me tell you what it does before you make your verdict. Okay? It's from Philip, and it has. Okay, it has a long LED panel light that just shines down, illuminates the room 16 million possible colors. Then it has two track light spotlights adjustable on each end of it. All right, so you have four adjustable track lights and one big long panel light. And you can aim the spotlights however you want. Then using an app, you can have many profiles for what color out of 16 million different colors and what brightness you want on all five of those elements. My vision is that when I'm doing deep work, for example, I want to have just like a small amount of warm yellow light coming out of the panel. And then each spot is going to be aimed at a different wall. So one wall has the pegboard with my maker equipment. One wall I'm putting up picture ledges with first edition techno thrillers. One wall has my circuitry artwork. And then the back wall is going to have a video game cabinet so it can shine a sort of light on each of those walls. Maybe even like a blue light or like an off yellow light. And then otherwise the room can be kind of dark, except for my bright task light right in front of my computer. On the other hand, if, like, we're in there during the day or I'm just working on my maker lab table, we can have good, bright, warm yellow light that lights up the whole thing. And the spotlights are bright lights on, like, the maker wall so I can see what I'm doing. And so I can have, like, deep work mode. Maker mode. Just like we're in there just working on the computer. Daytime mode. And I can. That's the idea. That's the vision. All right.

B (59:35)

Deep.

A (59:36)

Deep. Not crazy?

B (59:37)

No, no. All right, that's awesome.

A (59:40)

What about the video game cabinet? I wanted the ability to have a game in that room from my childhood. I like the 90s era in video games. Like all this interesting technological stuff happened, but I can still understand it as a computer. So I wanted something that reminded me of like 90s era arcades. So I'm putting in an NBA Jam.

B (60:03)

I used to play that game.

A (60:04)

Yeah, right. And I wanted to be a game where you could just like play for five minutes to clear your head and like go back to what you're working on. Deep or crazy deep? Yeah. All right. Are we going to get good at it, you and I?

B (60:17)

Maybe it could be like Michael Jordan and Scotty Pivot.

A (60:20)

I. You know, Michael Jordan was not in the original NBA Jam.

B (60:23)

Yeah, he wasn't.

A (60:24)

He was like, I don't want to be involved in this. And then he saw it and he's like, oh, this is awesome. And he had himself added back in. All right, so we're doing pretty good in there. Let me tell you what I'm putting on the art wall. So this was my idea. So we have like the big framed actual art, the circuit art from this former engineer from the mid century Silicon Valley who started making art out of circus stencils. And some of her pieces are at MoMA and some other big museums. And her grandkids sent me a piece of art from her because they liked the show. So I'm going to hang that up. It's built off of a circuit stencil, you know I'm talking about. Right, the green one.

B (61:00)

Yeah, yeah.

A (61:01)

So then I have two smaller frames to go next to it so they line up to be the same height. I bought a manual for like a 1980s era galaxia arcade cabinet, a repair manual that has the circuit diagrams for that arcade cabinet. And I'm framing in the smaller frames two of the actual circuit diagrams from that video game cabinet. Vintage repair manual. So those will be framed next to this like circuit based artwork.

B (61:34)

You can have special lights for those when you want to emphasize those. Right.

A (61:36)

And so the spotlight on that wall can be whatever it'll be shining right on just those artworks. So now when I'm in deep work mode, if I look over there, I see those artworks illuminated. If I look up, I'm putting first edition techno thrillers, largely from my childhood on the wall. These red acrylic picture racks. A light is just on there and If I look to my left, it's like all maker equipment. So it's all about trying to create the right mindset for depth.

B (61:59)

Yeah.

A (62:00)

All right, work continues. All that stuff's coming, by the way. And a rug, so Area rug. So it's not so like live in there.

B (62:07)

Electrician probably got to install a light, right?

A (62:09)

Yeah, we got a good guy. Yeah, got a good guy. Who's going to come do it. I want to see if you can bring more outlets in there.

B (62:14)

Yeah, there definitely needs to be more outlets.

A (62:15)

It's crazy. We, we, we power so much of this lab off of like one outlet. There is like extinction cores. Yeah, I'll talk to, I'll talk to our super about that. All right, let's get into what I read. Since the last episode recording, I finished two books. One was Marianne Wolf's book Reader Come Home. Fantastic book. Marianne Wolf is like the reader, the neuroscientist cognitive scientist who studies reading in the brain. She wrote Proust in the Squid. That book came out like just as we had like the smartphone revolution or whatever. And it sort of surprised her that that was part of the reception. So then she wrot this book. I think it's like 2018 maybe. I might have that date wrong. It's all about reading in the brain and the challenge. We're now at the age of the digital and her ultimate vision for building bilingual brains. So actually thinking about a brain that's fluent with deep reading of hard books and fluent with technology use in particular, like computer programming. The same way you would think about a brain that can speak like Spanish and English. You have two different languages that you're both learning and you can move back and forth between them fluently. It gets a lot of really good brain science in there. So that was a great book. I also finished a parenting book called what do youo say? By William Strixrud and Ned Johnson. We actually picked this up. My wife went to a parenting talk at our school and she picked up the book and God, do we need this advice. Jesse, I have three kids, including the oldest is a teenager. He's 13. Bring it on. Bring all the parenting advice. So Bill Strixrud has a big practice that does a childhood psychology practice. We know his daughter, which is interesting. Yeah, she's a parent at our kids school and she shows up in the book a few times. So that was good. If you have like teenagers, it's a good parenting book. So I needed that. Let me point something out, by the way. We're recording this. I Want to shine a light on my reading strategy. We're recording this on March 17th. So those represent my third and fourth books of March. Some four books in at the halfway point at March. And I really put aside a lot of time to do that because my middle child is a big. Ironically, given our show, he's a big Brandon Sanderson fan, so he read the Mistborn trilogy and now has started on the King something. I don't know the names. King Solver. That's not right. But whatever. Big thick books. So I said I would read the first Mistborn book. The. So we could kind of connect over. But it's like a bit of a beast at 600 pages. So I was like, I want to finish my other four books so I can spend the second half of the month just reading this one, like, kind of long novel. And I kind of get lost in the world and not be stressed about it. So that's what I'm doing. I'm now going to dive into that Brandon Sanderson book. I wanted to read Name of the Wind, his best book. I feel like I have to explain this. I think we have too many new listeners that I have to explain this.

B (65:06)

Yeah, explain it.

A (65:06)

Okay. I know Brandon Sanderson did not write Name of the Wind, but I made that mistake. Was this, like, five years ago, Jesse. I mean, it's a long time ago. A long time ago. I accidentally said Brandon Sanderson was the author of Name of the Wind instead of Patrick Rufos. And we heard about it. I think it's the most controversial. More controversial than our Charlie Kirk episode or some of my hot AI takes. It's like, no, no, no. You mixed up, Brandon. So anyways, it's been a running joke ever since then. And the joke is without explanation. I just pretend. I just like, yeah, you know, like, Brandon Sanderson's best book is Name of the Wind. And every time we get letters. Every time. And I love it. I don't know why, but every time we get letters. And I'm going to continue doing that joke, including if and when I meet Brandon Sanderson, and that'll probably be the end of that. All right, final thing. Different parts of me in the news might be interesting. Recently, a big interview with me came out in the Chronicle of Higher Education. If you're in sort of academic adjacent worlds, it was titled is AI Making Us Stupid? I really get into AI and the Academy and the point of university life and how we should and shouldn't use AI. So it's worth reading, especially if you're adjacent to that world. I think you can sign up for a free account at least for a while if you want to check that out. Also, I was on an episode of Tim Ferriss show. I think it came out recently, maybe last week. I'm not sure if I mentioned it or not. He had four shorter segments from four different people and I was one of the four people. And I was talking about simplifying and I talked about the somewhat drastic things I do in my life to try to keep it under control and simplify it. How I basically say no to almost everything. That's not just my core efforts at producing new ideas and publishing them again out in the world. So if you're interested in sort of how I try to manage the overload of opportunities on my schedule, find that Tim Ferriss episode from recently that had me in it. All right, Jesse, I think that's all. Thanks for listening. We'll be back next week with another episode of we're back on next Monday with another episode, another advice episode. And this Thursday I have a AI Reality Check episode queued up to come out as well. So look for that. And until then, as always, stay deep.