The Summer Schedule Reset To Save You From Burnout & Unlock Deep Work

00:00:00.000 | For me, today is, in some sense, the first day of summer. I'm a college professor and my final

00:00:05.980 | grades were submitted last week. I hosted the lunch for the graduating seniors. Commencement

00:00:10.600 | was on Saturday. The campus has just shut down for the summer season. Unlike most professors at

00:00:16.720 | most research universities, however, I don't take summer salary. As soon as the spring semester ends,

00:00:23.340 | I am officially off the clock until the fall comes. I've received no paycheck from the university.

00:00:28.360 | I received no paycheck from the grant agencies during the summer. There is no expectations from

00:00:33.300 | anyone else. My time is my own. Here is the schedule that I more or less try to run

00:00:39.920 | during these summers of no external obligations. I try, if possible, to have no professional

00:00:47.640 | appointments, meetings, or calls on Mondays or Fridays. I want to begin and end my week with

00:00:52.700 | quiet and depth. So the transition into the weekend and from the weekend back into the

00:00:57.680 | week is one that is much more gradual and subtle and cognitively rewarding. You're not going from

00:01:02.580 | a quiet Sunday to a crazy Monday. I try, when possible, to make one of those days, Monday or

00:01:08.020 | Friday, an adventure-thinking day where I'll spend at least three or four hours, usually somewhere

00:01:13.520 | outside in nature, being way too hot because I do live in Washington, D.C., but maybe going to

00:01:17.900 | like the Pactuxic Wildlife Refuge right outside of the Beltway or to Wheaton Regional Park or to the,

00:01:23.800 | the, the, I like the, the Rachel Carson Greenway, to a trail, go for a hike, spend a couple hours

00:01:30.080 | outside, go to Rock Creek Park, bring my notebooks and try to really work through ideas. For Tuesday,

00:01:34.920 | Wednesday, and Thursday, those days start with deep work. I don't look at an email or a computer until

00:01:39.560 | midday. Then in the afternoon, I have an admin block. What I try to do is have 30 minutes on my calendar

00:01:45.260 | midday, Tuesday, Wednesday, Thursday, where like a, a, a furious, a furious task

00:01:50.320 | completing machine. I go through small logistical things that have piled up and try to make as much

00:01:56.620 | progress as possible. In the afternoons on those three days is when I'll have calls, I'll have

00:02:00.880 | interviews, I'll have, you know, professional related, uh, appointments. And I try to end those days

00:02:05.860 | by four. That is my summer schedule and I love it and I can get away with it because I, I have these

00:02:11.220 | really slow summers, but it is a vital part of my success. It is a step away from the frantic demands

00:02:20.940 | of the school year. I get my best thinking done during the summer. I produce on my best right in

00:02:25.560 | the summer. It's when I recharge for the year ahead. So I want to talk to you today about this general

00:02:31.000 | strategy of seasonality. I'm going to argue that it's important for basically everyone, especially if

00:02:37.080 | you have some sort of knowledge or office work job, you might not be able to push your summer schedule

00:02:43.220 | to the same extreme as mine, but I think it should be different. I'll argue why. And then I'll give you

00:02:47.880 | some ideas about how you can do that. Even if you have less flexibility than myself. All right, let's

00:02:53.860 | start with the case for the seasonality, taking one season to be different and less intense than other

00:02:59.500 | seasons. When it comes to your work, I'm going to read something from my, uh, my book here, the Bible

00:03:05.160 | for the show, slow productivity. I wrote about this idea, did some good research on this. So I'm going

00:03:11.020 | to take a second to take advantage of the research I already did. All right. I'm going to read here.

00:03:17.560 | The side-by-side comparison underscores the degree to which our experience of work has transformed during

00:03:22.160 | the recent past of our species. Our shift from hunting and gathering to agriculture, the Neolithic

00:03:27.420 | revolution only really picked up speed somewhere around 12,000 years ago. By the time of the Roman

00:03:32.960 | empire foraging had almost completely disappeared from the human story, this reorientation towards

00:03:37.920 | agriculture through most of humanity into a state similar to that of the rice farming Agta, grappling

00:03:43.680 | with something new, the continuous monotony of unvarying work all day long, day after day. All right. So I'm

00:03:50.000 | arguing here. I, I, so I, I do a, before this, I went through a treatment of what the foraging and hunting

00:03:56.320 | and gathering life was like in which we spent most of our species existence. Drawing

00:04:01.280 | on some more recent research from the Agta people in the Philippines, where you have the same people

00:04:07.040 | that divided half of the group, went to rice farming and the other half stayed with foraging. There's

00:04:14.400 | interesting research from the 2000s where they compared the lifestyle between these two groups. They were

00:04:18.400 | otherwise very much controlled to be the same people in the same type of geographical region. And they found that,

00:04:23.120 | oh, the, the hunting and gathering foraging group has way more variation in their days.

00:04:27.360 | Busy periods, non-busy periods, busy seasons, non-busy periods. So what I'm arguing here is

00:04:32.080 | that change with agriculture, our days when we were working became uniformly hard days. However,

00:04:38.240 | I'm going back to reading here, the one saving grace in this scenario is that agriculture didn't demand

00:04:43.280 | this homogenized effort the entire year as the busy sowing and gathering of crops is offset by the quiet

00:04:49.680 | of winter. So we went from hunting and gathering and foraging where even within a given day,

00:04:54.880 | there's huge variation in intensity. This hour is really locked in because we're in the middle of a

00:05:01.280 | hunt, but the next two hours we're taking a nap during the midday sun. Then we get agriculture.

00:05:05.360 | I said, no, no, when you work, we have this new idea of like working sun rise to sun set of equal

00:05:11.120 | intensity. Our paleolithic for bearers did not have such uniform intensity, but at a bigger scale,

00:05:18.560 | you had busy seasons and less busy seasons. The fall was very busy. The winter was not.

00:05:24.320 | Moving forward in the human story, and I'm reading again here, the industrial revolution stripped away

00:05:29.520 | those last vestiges of variation in our work efforts. The powered mill followed by the factory

00:05:34.960 | made every day a harvest day. Continuous monotonous labor that never alters. Gone were the seasonal

00:05:41.200 | changes and since making rituals. Marks for all his flaws and overreach hit on something deep with his

00:05:46.240 | theory of entrefundung. I said that perfectly, Jesse, by the way. Estrangement, which argued

00:05:53.440 | that the industrial order alienated us from our business, our basic human nature. The workers

00:05:58.720 | eventually inevitably fought back against this grim situation. They pushed for reform legislation like

00:06:03.120 | the Fair Labor Standards Act passed by the U.S. Congress in 1938, which fixed 40 hours as a standard

00:06:08.160 | work week, limiting the fraction of the day that could be snared a monotonous effort without extra pay.

00:06:12.560 | They also formed labor unions as a counterbalance to the more dehumanizing aspects of industrialization.

00:06:17.600 | All right. So what happens next is the factory in the mill is a brand new thing where we said,

00:06:21.440 | not only are you going to work all day long at equal intensity, there are no seasons anymore.

00:06:25.920 | The textile mill doesn't care if it's January versus June. Now, all year round, there's not going to be

00:06:34.880 | a change. What I'm arguing here in the book is that this was deranging for people. This was like

00:06:41.760 | really difficult. It's not what our species evolved to do, to hard work all day long without change all

00:06:47.600 | year round. Very unnatural. And so we had to, at least to make it bearable, get things like reform

00:06:54.800 | legislation and labor unions to at least try to have like some pushback there. Finally, we get the

00:07:01.520 | knowledge work, which brings us to what us as our audience, you know, to our world. So let me read

00:07:06.960 | here briefly. Then knowledge work entered the scene as a major economic sector. The managerial class

00:07:12.320 | didn't know how to handle the autonomy and variety of jobs of this new sector. Their stopgap response

00:07:16.800 | was pseudo productivity, which used visible activity as a proxy for usefulness. Under this new configuration,

00:07:22.160 | we took another step backward. As in the industrial sector, we continued to work all day, every day,

00:07:26.960 | without seasonal changes, as any such variation would now be received as non-productiveness. But unlike in the

00:07:32.480 | industrial sector, in this invisible factory, we'd constructed for ourselves, we didn't have reformed

00:07:38.880 | legislation or unions to identify the most draining aspects of the setup and fight for limits. Knowledge

00:07:43.840 | work was free to totalize our existence, colonizing as much of our time from evenings to weekends to

00:07:49.200 | vacations as we could bear, and leaving little recourse beyond burnout or demotion or quitting when it became

00:07:55.440 | too much. Our estrangement from the rhythms of work that dominated the first 280,000 years of our species

00:08:01.520 | existence was now complete. So the argument I'm making there is that we've sort of invented, when it comes

00:08:07.840 | to knowledge work, the most unnatural possible way of approaching work. It is as far as we have ever

00:08:14.080 | been as a species away from how we spent most of our time as a species. Kind of had this steady march away

00:08:19.920 | from the way we were wired to exist for the Paleolithic. Agriculture made it a little farther

00:08:25.840 | from that. The Industrial Revolution was even farther from that, and we had to bring in protection so that

00:08:29.840 | at least it didn't overwhelm us, and at least we recognized, "Hey, this is really unnatural, so you

00:08:33.840 | better pay us well and extra pay, and we're going to push back." And then knowledge work got rid of all

00:08:38.800 | that and just left with the really unnatural pace, and everyone now pretended like this was good though,

00:08:43.760 | and this was somehow like what it meant to be productive. So seasonality pushes back against that.

00:08:50.320 | It goes back towards our agricultural roots and even farther back to our forager roots where it says we

00:08:55.120 | have busier periods and less busier periods, a busy time of year and a less busy time of year.

00:08:59.360 | When you better match the natural rhythms for which we're wired, your life becomes more tolerable and

00:09:04.320 | sustainable, and over time you produce better quality work because when you're in this unnatural

00:09:10.400 | edifice of continuous, invisible factory labor, you just burn out. And then it just becomes activity

00:09:16.880 | for the sake of activity. And yeah, you're working all day long, but what does that work? Well, a lot

00:09:21.200 | of it's going to be moving emails back and forth, taking on those unnecessary meetings because at least

00:09:25.360 | it's something that's not too taxing and looks like you have activity. It's not a good way of producing

00:09:30.240 | results, and it's a great way of burning out the human psyche. So I am a big believer in having

00:09:34.400 | variation. I will say as an aside, by the way, Jesse, that whole,

00:09:39.120 | I won't name names, but that whole sequence in this book, which came out of a New Yorker piece

00:09:43.680 | that came out before the book, starting with the hunting and gatherers, moving through agriculture

00:09:48.480 | into the industrial revolution, how we got alienated from our rhythms, quoting Marx and Marx's theory of

00:09:54.480 | estrangement to talk about what happened when we finally separated so much from what we were wired for.

00:10:02.960 | A book that came out not too recently had that exact same arc with the same example, citing some of the

00:10:10.640 | same studies, ending up with Marx and the theory of estrangement or whatever. I don't think the

00:10:16.320 | author like was stealing it. I think his research assistant just like came across my New Yorker piece

00:10:22.720 | at some point, or like, great, here's some ideas. They went into their system. And this is kind of the

00:10:26.240 | problem of the modern way that a lot of these nonfiction writers write, where they just have a research

00:10:29.520 | to have a research assistant gather a bunch of stuff on a bunch of topics. And then the writer

00:10:33.680 | has a bunch of examples to pull from to make their points. I think they just, he just digested my New

00:10:37.920 | Yorker piece, broke it into its pieces. And then the author was like, oh, these are good research parts

00:10:41.840 | and reconstructed my exact arc basically in his book.

00:10:45.680 | Did you read the book?

00:10:46.960 | Yeah.

00:10:47.040 | Have you reviewed it on your monthly books?

00:10:49.600 | You're trying to get me to, you're trying to get me to identify who this is. I will, I will not identify

00:10:56.160 | who this is. But yes, it was the baseball book of why. It was a weird tangent that author went on. I don't

00:11:03.440 | know what Marxist theory of estrangement had to do with baseball trivia, but you know, damn him. I'm on,

00:11:09.280 | I'm on to him. Hey, it's Cal. I wanted to interrupt briefly to say that if you're enjoying this video,

00:11:14.960 | then you need to check out my new book, Slow Productivity: The Lost Art of Accomplishment

00:11:20.960 | Without Burnout. This is like the Bible for most of the ideas we talk about here in these videos.

00:11:28.240 | You can get a free excerpt at calnewport.com/slow.

00:11:33.920 | I know you're going to like it. Check it out. Now let's get back to the video.

00:11:37.600 | All right. Seasonality. So how do we do this? Well, what I do with my summer schedule is one

00:11:41.600 | example, but that's pretty specific to be, I mean, it's like why I became a professor and writer was

00:11:46.000 | exactly to be able to do that. Flexibility and schedule and, and the job, like the writing that

00:11:52.160 | could sort of support me to take time off like that and give me like an activity to do. There's a lot of

00:11:56.800 | ideas you can do that are smaller. That'll give you a taste of seasonality. I have a few to suggest

00:12:01.280 | here as you think about the summer ahead. One, consider having one day each week, maybe a Monday

00:12:07.520 | or Friday, where for the most part, you don't schedule appointments. Don't make a big deal about

00:12:11.840 | this. Don't announce it. Don't tell people when they ask you when you're available that you're not

00:12:15.440 | available on Mondays. Have plenty of options to suggest like, yeah, I have this day, this day. How

00:12:20.640 | about here, here, here? Just don't happen to suggest days on this day that you have secretly

00:12:25.440 | marked for yourself as a no meeting day each week. It won't always work, but many weeks it will. And to

00:12:30.960 | have one day a week or two days a week where especially connected to a weekend where you can

00:12:36.000 | get into your work and get caught up and see what's going on at a more of a leisurely entry pace and not

00:12:41.760 | have to jump into a thousand meetings really makes a big difference, makes the season seem slower.

00:12:46.720 | If possible, try to make one of those days into more of an adventure day. Get out of your normal

00:12:51.280 | workspace if you can, especially if you were able to work remote, go into nature, go to a museum,

00:12:56.240 | go to the mall in DC. It's what I used to do. Make it feel different. Now,

00:13:00.640 | still work, work on your notebook, bring your laptop with you, still get things done,

00:13:05.200 | but in a completely different location. It just makes the pace of work seem different.

00:13:10.240 | I would also suggest, oh, I lost a page here.

00:13:14.560 | Where did I put? Oh, I see what I did here, Jesse. I put my ads in the wrong place.

00:13:20.640 | Okay. So no meeting days. Yes. All right. Another thing you should think about doing is pushing new

00:13:26.080 | project start times. Again, you don't need to make a big deal about this, but you can wind down

00:13:31.040 | your projects as the summer starts. And as you take on new projects, be like, yeah, yeah, yeah. I'm

00:13:36.960 | finishing up some things now, but like late August, I'll get that ramped up. That's close enough in the

00:13:42.400 | future that people are like, yeah, fine. That's reasonable. But what you've done is created a bit

00:13:46.640 | of a lacuna in your schedule here. There's going to be this period where you're not really starting or in

00:13:50.560 | the middle of any big projects. Now you can't do this all the time because eventually the projects

00:13:54.160 | have to get done. So you can't keep pushing them into the future. But once a year, you can start

00:13:59.040 | saying for a little while when new projects are coming to you, like, yeah, I'm finishing up a

00:14:04.800 | bunch of stuff now, but what if like in mid August, I'll take this on? And the reality is you're not

00:14:09.200 | really finishing up that much stuff, but it does buy you some space. Put your hand down more. And what I

00:14:15.040 | mean by that is when it comes to volunteer activities within your work, especially what researchers call

00:14:19.520 | non-promotable activities, like organizing the internal office birthday parties or this or that,

00:14:24.480 | maybe you're that reliable go-to person that likes to help out and people like that about you.

00:14:29.520 | That's great, but put your hand down for a couple months. Just don't volunteer or say, I don't know,

00:14:35.820 | I'm a little busy right now. Again, if you do this all the time, then people might be like, oh man,

00:14:40.700 | he or she looks never useful. It's kind of annoying. You do it for a month or so for six weeks. No one notices

00:14:46.360 | then because like you're mainly volunteering a lot. They won't notice, but it makes these periods feel

00:14:51.160 | different. You just have less of this extra work on your plate. This is a strategy I talked about in my

00:14:57.160 | slow productivity book. The idea of like a highly autonomous blocker project. You take on a project

00:15:02.280 | for the summer that you can keep referring to to try to push away other type of work. Like, well,

00:15:09.220 | yeah, I can do that, but I'm working on the Maguire report, right? And like, that's where my head's down on

00:15:14.280 | this, but after I finished the Maguire report, like I can work on these other types of things.

00:15:18.460 | Choose that proverbial Maguire report to be something that's highly autonomous that you can,

00:15:23.420 | in reality, do like an hour every day and stay on top of it. It is a extremely effective strategy.

00:15:30.540 | People don't know how long work takes. Most people are like pretty inefficient. You take something where

00:15:34.460 | you have a lot of flexibility. It doesn't generate a lot of meetings. It doesn't make you have to be on

00:15:38.140 | other people's schedule. Be efficient and effective and then use that thing to push everything else off

00:15:43.900 | your plate for a little while. You can only get away with that so long, but really we're just looking for

00:15:47.160 | a month or two of respite. Start your day slowly. Apologize only if people notice. Like, start your day in

00:15:54.680 | the coffee shop. Like, get a little breakfast and coffee. Work there for a while. Just change the rhythm

00:15:58.940 | of the day. So, it just feels slower. It's not as frantic as other times of years. Like, this is

00:16:05.060 | psychological, but these type of psychological differences really can add up. Finally, this idea that people keep

00:16:12.880 | getting upset about this. I'm not sure why. Just, you know, every other week, go do something midday

00:16:17.820 | some day that you don't normally do during the weekday. You go see a movie or something.

00:16:22.220 | It's two and a half hours. It's not some big deal. People go to doctor's appointments and stuff all

00:16:25.700 | the time, but it's going to give you a completely different feel for your week. When you're in a

00:16:31.700 | movie theater at two o'clock and you know it's a weekday, it makes you feel like you're not just in

00:16:38.180 | like the normal invisible factory clock and clock out. This is what I do every day. It just gives you

00:16:42.180 | a sense of autonomy and slowness and control. And in the summer, do this more often. Go see some dumb

00:16:46.340 | movies, right? Especially if you're in a situation where you don't have to account for every minute of your

00:16:51.220 | time. Again, if you go to movies every day, if you stop work at two every day all the time, people

00:16:56.040 | will notice. But if once or twice a month, you're gone in the afternoon, you're back a little bit

00:16:59.980 | late. No one knows. No one notices. So you can use these smaller tips if you don't have a huge amount

00:17:05.520 | of autonomy to introduce some seasonality. Summer is a great time for seasonality. So hopefully a lot

00:17:13.600 | of you in the audience are going to join me in slowing down for the months ahead.

00:17:19.820 | Do you ever need more than 30 minutes to do admin work?

00:17:23.600 | Probably. Yeah. Sometimes it extends more. But like I try to just have this mindset of when I get in

00:17:30.100 | there, especially for like Georgetown stuff, which there's not much in the summer, but queries come up

00:17:35.460 | and a dean needs this or that. Just try to be like really focused and work through. The opposite

00:17:41.480 | problem is what I'm trying to solve more often than not, which is there's not a lot of stuff in my

00:17:47.120 | inbox right now. But try to do 30 full minutes of useful admin work, even if there's not just messages

00:17:53.060 | to answer. Like fill in the rest of the time to get ahead of things or to organize things or to

00:17:57.540 | preemptively like step out or prod things along. So it goes both ways. Try not to work beyond that 30

00:18:03.440 | minutes too far, but also make sure you use the full 30 minutes. This is not going to work for

00:18:08.120 | most jobs. It wouldn't work for my job in September, but in July, because it's an academic institution

00:18:14.180 | that's technically shut down and I'm not on salary, it's much more plausible. I also have to deal with

00:18:19.620 | writer and admin as well. But there I'm just really good at all the various teams in my life sort of

00:18:24.920 | know in the summer. Like, yeah, Cal's not, that's not his super accessible time. I guess I thought you

00:18:30.660 | were also handling your personal admin. No, that takes a, yeah, that's different.

00:18:35.080 | Okay. That's different. Yeah. But that's fine. That's summertime. Yeah. You have to like mow your

00:18:39.160 | yard and work on your budgets. I got to do my taxes next week. I'm doing my budget today. Yeah. But that

00:18:44.840 | stuff's fine. That's like you're at home and I don't mind that as much. And that does take a while.

00:18:49.380 | All right. Well, we got a bunch of good lister questions coming up. But first here, a quick word

00:18:55.660 | from some of our sponsors. I want to talk about my friends at Cozy Earth. I love Cozy Earth. This is

00:19:04.340 | one of these products or companies where I use a lot of their products. I absolutely swear by them. In

00:19:09.120 | particular, their sheets are a must have in our family. We have multiple pairs of these sheets so

00:19:14.560 | that when one is being washed, we can switch them into the bed. We have other multiple pairs for the

00:19:18.620 | guest bedroom. We travel with these if we're going to be somewhere for more than a week or two. They

00:19:23.200 | are very comfortable sheets and they're cooling. And we, my wife and I really have a hard time now

00:19:28.260 | sleeping on almost anything else. We have other things as well. We have the duvet cover. We have

00:19:33.560 | the towels. My wife has the PJs. I have one of the sweatshirts, which somehow has like a cooling

00:19:37.720 | effect, just like the sheets, which is really pretty cool. It kind of like cools you down when you're

00:19:41.520 | hot. I am a fan of Cozy Earth. I swear by them and I think you will enjoy them as well. Like

00:19:48.720 | their mission is to transform your five to nine, the time that matters most into the most comfortable

00:19:53.140 | part of your day. All right. And you're going to get that with Cozy Earth. Start with the sheets,

00:19:58.980 | but check out the other products as well. We gave these away at least twice. We've given these away

00:20:05.460 | as gifts. I think I gave some to my mom and then Julie gave some. We, we, we like, so we like these

00:20:14.200 | things. We like, we like Cozy Earth. All right. They have a hundred day sleep, a hundred night sleep

00:20:19.620 | trial rather. Enjoy over three months to experience the unmatched comfort of Cozy Earth. If you're not

00:20:24.080 | completely in love with your bedding, you can return it hassle-free, but guess what? You're not going to do

00:20:28.340 | that. They also have a decade long warranty on their bedding products. This means years of luxurious,

00:20:35.100 | worry-free sleep, knowing that your investment is protected. Luxury shouldn't be out of reach.

00:20:39.540 | Visit CozyEarth.com and use my exclusive code DEEP for up to 40% off Cozy Earth's best-selling

00:20:45.420 | sheets, towels, pajamas, and more. Trust me, you won't regret it. That's CozyEarth.com,

00:20:50.020 | code DEEP. And if you get a post-purchase survey, tell them you heard about Cozy Earth from

00:20:55.200 | Deep Questions with Cal Newport. Sanctuary awaits at Cozy Earth. Also want to talk about

00:21:01.820 | our friends at Shopify. If you run a small business, you know there's nothing small about it. As a

00:21:07.040 | business owner, I get it. My business takes up a lot of my time and attention, and I am always

00:21:13.760 | looking for places to avoid having to waste time dealing with extra products or things that don't

00:21:19.740 | work well. This is where Shopify enters the scene. If you're going to be selling things,

00:21:24.280 | you 100% need to be using Shopify. This is just a consistent theme about all of the entrepreneurs

00:21:31.620 | that Jesse and I know that sell various things, whether it be online or in the real world.

00:21:35.680 | Shopify is a consistent theme how much they like Shopify. Shopify is the commerce platform behind

00:21:41.320 | millions of businesses around the world, and 10% of all e-commerce in the U.S. We're talking about

00:21:48.060 | household names like Mattel and Gymshark as well as to brands that are just getting started. Tackle all

00:21:54.680 | your important tasks in one place. Shopify, you could be doing inventory. You can be handling your

00:21:59.700 | payments, of course, but also analytics. It can help you with your marketing. It can help you with

00:22:04.780 | international sales. They have tools that can work in over 150 countries. You have a real store.

00:22:11.220 | They have award-winning point of sale, and they can connect your online and offline sales all

00:22:15.100 | together. Use Shopify if you're selling things. I don't care where you're doing it or what type of

00:22:19.400 | thing you're selling, online, digital, what country. Think Shopify. So get all the big stuff for your

00:22:25.500 | small business right with Shopify. Sign up for your $1 per month trial and start selling today at

00:22:31.100 | shopify.com slash deep. Go to shopify.com slash deep. Shopify.com slash deep. All right, Jesse,

00:22:39.260 | let's move on to some questions. First questions from Quiz. I'm a teacher in Texas who usually spends

00:22:46.840 | summers on long backtracking trips, which helps me recharge while walking, listening to good content,

00:22:52.020 | praying, and reflecting. Because of a knee injury, I can't do my usually trip this year and don't enjoy

00:22:57.200 | too much unstructured time at home. How would you suggest I structure my eight-week summer break and

00:23:02.560 | daily schedule to get similar benefits without being able to leave home or walk all day?

00:23:06.520 | I'm thinking maybe take up drinking. Good old-fashioned Texas moonshine or whiskey, like an old-fashioned

00:23:16.120 | Texas cowboy. That'll get you through the eight weeks pretty well. Okay, so you can't walk all day,

00:23:21.180 | but I want to push on this a little more. You say you can't leave home. I don't think he means he

00:23:28.260 | can't literally leave his house. He means just like on longer trips, I guess, right? Yeah. Yeah.

00:23:32.340 | Because I was going to say, you know, a few things I was going to say, but my overarching advice here

00:23:37.900 | is have a schedule, have a plan, right? Don't go into your day saying like, ah, whatever. The day is

00:23:46.240 | kind of awful. Just see how it unfolds. There's not much I can do because of my knee. It's that type of

00:23:51.000 | slow, uh, slow through the mud. Like I'm just sort of wasting time. I'm on my phone for three hours this

00:23:58.860 | morning. I don't even know where that when I'm going to watch some TV and see what's going on. Like that

00:24:03.160 | type of, uh, stuck in the mud, mired in time mentality can really be relaxing for a day, but over

00:24:11.580 | eight weeks, I think is really going to be a bummer. So you want to have some notion of a

00:24:15.660 | schedule or plan, whatever those details are. This is what I'm doing this summer. And this should

00:24:19.460 | probably involve some sort of project or projects you're working on. And these might be more intellectual,

00:24:23.780 | cognitive than physical, preferably things you can work on outside of your house. I mean,

00:24:29.400 | you say you can't leave your house, but you can leave your house and go different places. Start

00:24:32.900 | your morning with the paper at the coffee shop. Then you're going to the woodshed because you're

00:24:36.260 | working on mastering some sort of new woodworking tool, you know? Uh, yeah, your knee is hurt. You

00:24:42.020 | can't walk all day, but it's not going to be something where you can't walk on it. I mean, even

00:24:47.020 | people with knee replacements, they want you up and around. So I want you up and around working on some

00:24:51.980 | sort of projects that is not just in your house. It could be cognitive. It could be physical. Some

00:24:56.320 | mix of the two have some initiatives you're going on. I'm going to read all these books. I'm going to

00:25:00.680 | watch every movie that, uh, you know, every Ford movie between this 10 year period and like read

00:25:06.740 | books about or whatever it is, but have things to do that feels like they're rewarding. That's what

00:25:12.140 | the long hikes were. So you want something that's similarly ambitious, but just not something that

00:25:17.000 | makes such a big demand on your knee. And there needs to be a plan here. I think we tell ourselves,

00:25:21.680 | this was kind of my pushback against the sort of Ginny O'Dell hypothesis about doing nothing.

00:25:27.020 | We tell ourselves that some sort of Rousseauian platonic ideal that doing things is a internalization

00:25:34.280 | of capitalist narratives and doing nothing is a brave resistance. And that's where we're going

00:25:37.940 | to rediscover ourselves. But there's only so much of literally doing nothing that we can put up with

00:25:45.160 | before we get antsy. And then what happens is, is we have all these highly polished attention economy

00:25:50.880 | tools that are very much happy to solve our problem for us. Like you're antsy about doing nothing.

00:25:55.140 | look at this screen. Your eyeballs bleed. We can give you something to do. And then time just

00:26:01.020 | disappears and you come out of it feeling depressed. I was just looking at a study that was featured on

00:26:06.080 | NPR clip a couple months ago that was showing these connections between looking at algorithmic content

00:26:12.520 | and your mood. And as they, in an experiment, took people off of their phones, essentially, they found

00:26:18.140 | this steady increase in their actual psychological wellbeing. So it's, you know, doing nothing.

00:26:25.240 | It's not at, we think it is very appealing, but really what we don't like is being overloaded,

00:26:31.220 | burnt out, overworked, but we're not looking for the opposite of that. Or if we are, the opposite of

00:26:37.820 | grueling efforts that feel meaningless and, and pointless is efforts that are meaningful and have a

00:26:51.580 | point, right? So the, the, the, the, the opposite of these things that wear us out is not sitting there

00:26:57.440 | on your phone doing nothing all day. It's doing less things and doing those things better, more meaningful

00:27:01.380 | things. So just keep that in mind quiz and then keep, keep active, get that knee healed. Maybe PT is a big

00:27:07.260 | part of the something you're doing this summer and then get back to those hikes as soon as you

00:27:10.700 | can. All right. Who do we have next?

00:27:12.860 | Next is from Martin. In episode 352, you answered the question, am I working hard enough to get

00:27:18.440 | tenure? I'm a postdoc myself in computer science, but not at an R1 university. A few faculty members here at the,

00:27:25.080 | are at the top of their respective fields, but we do not learn the same toolbox to publish

00:27:28.660 | consistently as a top venues. How do I discover these secrets if the faculty don't even know my name?

00:27:34.920 | I mean, honestly, Martin, like if that's what you want to do, right? You want to publish, uh,

00:27:39.800 | at the top of your field and gun for top academic positions, you need to work for those people. So if

00:27:46.700 | there's a few of those people at your school, like you want to find a way to actually collaborate with

00:27:51.340 | them. Now you might not be able to change your postdoc advisor this far into game, but if you can

00:27:55.300 | find a way to collaborate with them, Hey, I'm making your life useful. I'm writing a paper with you and

00:27:59.280 | doing a lot of the work. That's what you have to do. I mean, it's an apprenticeship system.

00:28:03.780 | High-end academia is an apprenticeship system. You have to find an apprenticeship. It is very hard

00:28:09.380 | otherwise to make your way into this sort of metaphorical guild. So see if there's a way you

00:28:13.940 | can write a paper with these top of the line faculties. I mean, good for you for noticing based

00:28:18.480 | on last week's answer, that's who you need to be allied with. If you have this very narrow goal

00:28:23.520 | of publishing the top places and getting access to these types of jobs. So you got to find a way to

00:28:27.980 | actually work with them. It's very difficult to learn these things from afar. All right, who do we got

00:28:33.460 | next? Next question is from Will. Even if the short form content topics are still relatively deep,

00:28:39.680 | do you still caution against short form content creation and encourage long form writing as a way

00:28:44.360 | to sustainably build an audience? Or do you think today's attention economy is too focused on social

00:28:49.340 | media to the point where the barrier to entry is too high for a new successful blog or sub stack?

00:28:54.480 | So I was questioning a little bit his use of the word short and long. Maybe you have a better sense

00:28:59.900 | here, Jesse. When he's saying short form versus long form, is long form books?

00:29:04.920 | No, long forms is a blog or a sub stack and short form is like a YouTube short or something like that.

00:29:10.700 | Oh, I see. So he's saying, is it, is it too hard to the long form stuff that I talked about before?

00:29:19.220 | Is it too hard now to start with that?

00:29:20.700 | Yeah. Like to start your website and stuff like that.

00:29:22.860 | Um, okay. So it's always been hard. Here's what I think has happened. And I don't know this,

00:29:30.060 | this is maybe a curmudgeonly way of looking at the media landscape. But if you look back 30 years ago,

00:29:38.620 | where what's dominating the, in the printed word media landscape is going to be, you know, books and

00:29:44.700 | the, the big glossy Conde Nast magazines that you, you find on the newsstand, there wasn't a lot of

00:29:52.020 | people who were like, Hey, I kind of want to be in that game, right? Like you didn't have to sort of

00:29:57.200 | like the average person being like, yeah, I want to be one of the authors on the shelf there at Barnes and

00:30:01.340 | Noble or in the van, you know, with my articles and vanity fair, you had like a relatively narrow group of

00:30:06.220 | people who wanted to be professional writers. And that was a vocation. It was kind of competitive,

00:30:09.900 | but sort of cool if you could get there. But the average person was content to be like, yeah,

00:30:15.080 | I read that stuff. I think a lot of this changed with web two. And then particularly after web two

00:30:20.980 | gave way to social media where a big part of the pitch of phase one, social media, and I'll tell you

00:30:27.560 | what phase one means in a second was you can try on, you can cause play producing content for an

00:30:36.700 | audience. I'm convinced that this was the original sales pitch of phase one. Social media was that

00:30:42.860 | Facebook, which was the first big player, uh, they emerged in, in, in Oh four and they're generally

00:30:47.820 | public in Oh six. You have to remember at this point, web two, which had been around since like the

00:30:52.980 | early two thousands, which was blogs basically was pretty brutal. This technology came along is that,

00:30:59.340 | Hey, anyone now can publish something. You don't have to, there's not a gatekeeper at a magazine or

00:31:04.920 | a newspaper or like a web developer. You need this sort of put content. Uh, you can set up an account

00:31:09.760 | and publish something and it's on the web and anyone can read it. And what happened is a lot of people

00:31:14.880 | tried it and no one cared because it's really hard to write things as it always has been that people

00:31:20.000 | care about. And so a lot of people were disillusioned by blogs. It was, it was embarrassing.

00:31:24.480 | It was, um, uh, it was whatever it was demotivating. I, I, I wrote and no one came.

00:31:32.620 | Then Facebook came along and said, okay, here's our deal. You write on here, people will come.

00:31:38.180 | Here's the social contact. You're going to go through the contract. You're going to go through

00:31:41.640 | and you're going to indicate all these other Facebook users who know you, and we're going to

00:31:44.620 | tell them about stuff when you publish it. And yeah, the stuff doesn't matter, but there's going to be

00:31:48.700 | a social contract that they'll come over and they'll click like and poke and write comments

00:31:52.180 | and you'll do the same for them. And we can all sort of, uh, feel what it's like

00:31:56.720 | to be producing, you know, ideas and commentary and have an audience that cares about it.

00:32:03.460 | So we all can kind of cosplay in a lightweight way what it would feel like to actually be

00:32:07.900 | Michael Lewis writing these pieces for Vanity Fair, right? Or Malcolm Gladwell producing these

00:32:12.960 | books that are getting all, you know, these readers. And that was a big appeal of phase one

00:32:16.500 | social media. Now that changed, we got the phase two, TikTok brought us into the phase

00:32:21.800 | two where we said, you know, forget publishing your own stuff. What really matters is just

00:32:25.260 | like put it right into my veins, algorithmic distraction. And we kind of gave up caring

00:32:30.620 | about posting stuff and what people thought about us and was just like, put it right in my veins

00:32:35.440 | until my eyes bleed. That's phase two where we are now. I think that was the original pitch

00:32:39.200 | of phase one social media. And so it really traded on this idea that like, yeah, not only is the ability

00:32:47.260 | to publish words accessible to a lot of people now, but like, it's, it's really like kind of right

00:32:51.940 | there. You get a couple of right breaks and some stuff goes viral and, and, and you just feels like

00:32:56.900 | you're right there. You're posting stuff and people, you know, you have this built-in audience that'll

00:33:00.360 | look at it a little bit and anything could go viral at any time. And you hear stories about it.

00:33:04.540 | And suddenly now this became something that it was like running triathlons or something that like a

00:33:09.920 | lot of people could do. And I think it changed our, our approach to things like professional content

00:33:15.680 | production to like, Hey, maybe I should do this, which again, you don't hear a lot of people in

00:33:20.020 | 1985, just casually be like, walk by and see someone reading, you know, the New Yorker and be like, Hey,

00:33:24.840 | I should write for that. That might be fun. Like I can start doing that because there was just like

00:33:28.240 | less venues and they're really hard. So this has been curmudgeonly, but I'm coming back to this and

00:33:34.120 | saying, well, what content are you producing? Like, do you want to be a professional writer?

00:33:38.560 | Then like, you should go down the path of becoming a professional writer. Don't be seduced into this

00:33:44.780 | idea that there's some sort of shortcut into that that requires a lot less effort and a lot less

00:33:49.540 | hard hoops to make it through where I'm posting short-term content and something takes off. And,

00:33:54.420 | and, you know, next thing you know, I'm a writer on SNL or whatever, like, yeah, that happens,

00:33:58.260 | but really not that often. So, um, is the media attention economy too focused for things like a

00:34:06.200 | successful blog or sub stack or a podcast or books? No, I think it's like, it's always been. It's very

00:34:10.240 | hard. It's, it's, it's no easier, no harder than it's always been. If you want people to read your

00:34:15.700 | articles or books or listen to your podcast, which is kind of like an audio version of like long

00:34:19.820 | articles, it's like, like it was 30 years ago. It's hard. You have to have like an extra expertise and a

00:34:24.380 | voice and, and, and like find, you know, your audience and what you offer and have a couple

00:34:29.300 | swings that finally connect and then build on those as the same thing that professional writers have

00:34:33.280 | been doing all the time. And I don't think it's any harder, any easier than it's been before.

00:34:37.180 | We just have this other thing over here, which is the, the, the cause plane that you get in

00:34:42.460 | attention economy, social media, where everyone feels like they're pretty good. And there are,

00:34:47.420 | there are one lucky break away from being the next Kim Kardashian that I think has warped how we

00:34:52.780 | think about content creation, but I'm a big believer in writing professionally.

00:34:57.860 | I just put these reality checks on it that it's like very hard, right? There's the digital technology

00:35:04.540 | did not make this something that is easier for like a huge number of more people to succeed at.

00:35:09.080 | It changed the forms in which you can do it. It changed the economic for the people who are good

00:35:13.760 | at it. Now you have more control over how you reach and get to your audience, but it doesn't

00:35:18.380 | actually give you a big audience that is still earned. I think the ways that it always was.

00:35:22.760 | So I don't know. I feel like I'm going on a rant here, Jesse, but it's something I've been thinking

00:35:25.980 | about recently is that again, 30, 40 years ago, it would be more rare that someone just be like,

00:35:32.580 | I want to be publishing stuff that people read. It was like the people who really want to be a writer.

00:35:37.100 | And now I think it feels more accessible in a way in part because social media like plays on this idea

00:35:44.000 | of like people are listening to you and like you're, you're like one or two good quips away from like a

00:35:49.200 | retweet storm that's going to make you famous. And so I guess my answer is like the entry, the entry

00:35:55.360 | level is not too high. It's as high as it's always been, which is pretty high. It is hard. I was having

00:36:00.640 | this conversation with someone recently, like, well, how do you succeed with a podcast? It's like,

00:36:03.780 | it's hard. Like typically you have to have something else going that's already big.

00:36:06.820 | All these things. It's like sub stack. It really helps if you're already well-known. You're already

00:36:11.120 | a journalist who's well-known. Then like you could maybe build a successful sub stack.

00:36:14.840 | Do you read a lot of people's sub stacks?

00:36:16.780 | Um, some.

00:36:19.380 | Do you get the email notifications?

00:36:21.400 | I just look at my inbox. Yeah. Yeah. I don't use the app or sometimes it takes me over to the app,

00:36:26.940 | but I just look for my email inbox. I like that form. I mean, email newsletters,

00:36:31.400 | I think is a great form. Podcast is a great form.

00:36:33.360 | Um, I like these forms cause they're independent and you're not beholden. You're not beholden to

00:36:38.900 | like a super large companies that are using your content. I mean, sub stack does a little bit of

00:36:43.420 | this, but like email newsletters and podcasts, I think it's like a golden age of content production,

00:36:46.940 | but the people succeeding in these, it's the same people that were succeeding in like the

00:36:51.480 | magazine industry in the 1990s. It's like kind of the same number of people and some of the actual

00:36:55.820 | very same people, you know, like Gladwell and Lewis podcast. Uh, so the forms are different,

00:37:04.000 | but I don't think this is like the misnomer about media revolutions is they don't make it easier to

00:37:08.380 | be successful. They just sort of change the terms of success. That's kind of my growing theory.

00:37:15.040 | Now that we'll have our new weekly, our newsletter every week, it's going to be good. It's going to

00:37:19.820 | have a logo. It's going to have some good links. Got some good links in it. We're going to have

00:37:22.860 | links every week. I might eventually bring in a, from an archive piece at the end where just sort of

00:37:27.600 | pull out some classic stuff from the decade or so I've been writing about this stuff. It'll be good.

00:37:31.560 | Got to hone in on my topics though. Oh, the video for the podcast. Uh, there is going to be like

00:37:37.400 | featuring the, that week's podcast episode will be right there. So like people who are email

00:37:43.260 | subscribing will always hear like, Oh, here's what's going on in the podcast this week.

00:37:46.160 | Should be good. Yeah. And then on the podcast, what I'm going to do is in a monotone,

00:37:50.740 | read the entire newsletter word to word every week, just all the way through page break. Uh, I'll,

00:37:57.260 | I'll read the punctuation as well because yeah, how else is I going to know? Semicolon space.

00:38:01.400 | Just be like that. And that's really what the podcast is going to become. Uh, instead of me,

00:38:06.840 | I'm going to have a 1980s era Macintosh. Remember how they could talk with that robot voice,

00:38:11.980 | just reading out my newsletter. I think it'll be just as successful with an occasional throwing of

00:38:17.240 | your, uh, French accent with the pipe. Yeah. And then I'll come in and do a bad French accent.

00:38:21.900 | Yeah. I think we've got this all clued in. I think, I think we are about to explode. This world is so

00:38:27.340 | hard by the way. It's so hard to grow podcasts and newsletters speaking to which, I mean, that would

00:38:31.160 | grow it, but what a hard world. All right. What else do we have here? All right. We have Mark.

00:38:35.500 | Mark. I'm an actuary with a job that leads to adversarial interactions with regulators who are

00:38:40.780 | often challenging our assumptions in rejecting proposals. Since filings are infrequent and

00:38:45.960 | mistakes are highly visible, it's tough to design a deliberate practice routine to reduce pushback.

00:38:51.120 | Well, Mark, you're actually in a situation that's really right for deliberate practice because you're

00:38:56.100 | getting super clear feedback. We are really mad because of X. That's great feedback on like, okay,

00:39:02.800 | how do I never do X again? So you, you probably are in a cycle of deliberate practice here that is

00:39:09.000 | making you better. It's not fast in the sense of you're doing 20 reps a day, but like, you know,

00:39:15.160 | each year you're getting probably a dozen or two great pushbacks, right? So great teaching signals

00:39:22.480 | about what's working and what's not working. And the rate at which you are improving at writing these

00:39:27.100 | types of reports that you're filing is probably impressive. If you're looking at the right time

00:39:31.920 | scale, I will throw in, however, if this job is stressing you out and it sounds like it might be,

00:39:37.400 | I mean, it was stress me out. Like you're just constantly being yelled at by people who are mad

00:39:41.780 | with by default are mad about what you're going to send them. You might think about getting another job.

00:39:47.980 | The key, however, is if doing so, take your career capital with you. Don't quit this job to become

00:39:53.640 | a yoga instructor, quit this job to become an actuary somewhere else, but in a non-adversarial

00:39:59.240 | environment, right? So if there's something that's emotionally draining about a position,

00:40:03.000 | see if you can pick up your skills and take them somewhere else that removes that thing that is

00:40:07.460 | emotionally draining. Cause I'm suspecting that might actually be the real problem here.

00:40:10.400 | You know, the sparring, I wouldn't like that either. All right, who do we got next?

00:40:15.800 | Next up is Kayla. As someone who excelled in humanities and liberal arts, I appreciated

00:40:20.700 | your suggestions to read books with opposing views to grapple with different theories and

00:40:25.640 | perspectives. However, I'm curious if you have any recommendations for how to improve

00:40:29.420 | quantitative and technical skills as a way to get smarter in a dumb world.

00:40:33.300 | I would focus on specific projects that require you to master a new technical or quantitative skill

00:40:38.900 | to make progress. The two types of projects that you'll actually return to is either one that has

00:40:44.060 | a professional salient. So you volunteered to do this thing and your organization's like,

00:40:48.860 | yeah, that's good. Do this thing. And now they're waiting for an answer. And to get them an answer,

00:40:53.300 | you have to pick up the new skills. That'll focus your mind or something that's really fun. I think

00:40:58.280 | there'll be really fun to, you know, build this thing that's going to require me to learn circuit

00:41:03.540 | design or I'm going to have to learn computer programming. Maybe I'm doing it with my kids,

00:41:07.160 | but have a project that you either are interested in for professional financial reasons or for fun

00:41:12.660 | sort of like family connection reasons and let that drive you to have to pick up new quantitative

00:41:17.680 | skills. Don't just sit with a textbook and say, all right, guys, it's coming to Torx day and just like

00:41:21.920 | sit there and try to memorize theorems. Like that's not going to be very fun, but having a project is going

00:41:27.060 | to make that easier. I like, by the way, that you brought up my suggestion from before because I always

00:41:32.240 | like to underscore that read two books that have, that are very smart and have the opposite view on

00:41:37.540 | something. Even independent of the content, the dialectical clash of two opposing opinions on something

00:41:44.020 | adds like just a new layer of nuance to your mind. It just makes you smarter. It makes you see the world

00:41:49.300 | in more interesting shades of gray and not black and white. It's a really interesting experience. Even if you

00:41:54.680 | don't care about the topic, seeing that clash just adds like a more sophisticated layer to your brain. And if it's a

00:42:00.720 | topic you really care about, then you extra sure want to read the opposing view as well because that's where you're

00:42:06.460 | really going to figure out why you care about it. It's going to strengthen your conviction. It's going to make you a

00:42:10.760 | better advocate for that. Do not try to avoid stuff that you think is going to be contradicting stuff that you care about

00:42:16.880 | because that's not a belief, right? If it's I'm avoiding contradictory information, that's not a belief acquired in good

00:42:25.420 | faith. It's a, an idol that you're worshiping. It's you, you like the idea of being a part of that cult.

00:42:30.000 | So it's great. The challenge ideas, it's going to make them stronger. It's also great just to make you a smarter

00:42:35.420 | thinker to see what it looks like when you read two very opposing views. All right. Ooh, we got a case study

00:42:41.500 | this week. It's where people send in their accounts of applying the type of things we talked about on this

00:42:45.640 | show in their own life. If you have a case study, you can send it to Jesse at calnewport.com.

00:42:49.420 | do we have a name for this week? I guess not. We'll just call this Peter. All right. Peter said

00:42:57.800 | at the end of 2022, my life reached a turning point. I had just become a father, father for the first

00:43:03.320 | time, released my own music after years of supporting other artists and was deeply involved in serving my

00:43:08.860 | local Baha'i community. On paper, I had achieved, his name is probably not Peter. On paper, I achieved

00:43:14.540 | many of my life goals, but something still felt off. I realized after talking with my wife that I was

00:43:19.880 | constantly distracted thinking about music and work while at home and feeling guilty for not being

00:43:23.960 | fully present with my family or in my community service. Social media, I suspected was a big part

00:43:31.120 | of the problem. It'd be funny here if the case study said, so I got rid of my family and now have many,

00:43:38.920 | much more time for my social media and I've never looked back.

00:43:42.020 | His name is Cyril.

00:43:42.700 | Cyril. Okay. Social media, I suspected, was a big part of the problem. Discovering your TED Talk and

00:43:47.560 | then diving into your books, especially deep work, was a revelation. It helped me see that my habits

00:43:52.080 | needed to change if I wanted to be truly present and effective in all areas of my life. Over the past

00:43:57.340 | two years, I've worked hard to put your ideas into practice. I've become much more intentional about

00:44:01.880 | how I organize my days, treating my music career like a knowledge work job with clear boundaries.

00:44:06.360 | Deep work sessions are now a priority, allowing me to focus on improving my core musical skills,

00:44:11.360 | which are composing, producing, singing, and playing instruments. I'm also more disciplined

00:44:15.460 | with my time. I keep social media off my phone, stick to fixed work hours, and reserve evenings

00:44:20.560 | and weekends for my family. This structure has actually made me more flexible, letting me step

00:44:25.180 | away from work when needed for family or community projects. I've also picked up new habits like boxing

00:44:29.880 | and reading more books, which have brought a steady sense of well-being. All right. What I really like

00:44:35.480 | about this case study is this sometimes paradoxical sounding idea that having some more structured

00:44:40.780 | intentionality and control of your time makes your life slower and more full and more meaningful.

00:44:46.000 | There is sometimes this idea that's out there that once you begin to think more intentionally about

00:44:51.220 | your time, that necessarily means what you're doing is internalizing capitalist narratives and you're

00:44:55.500 | building your life into like a stressful scramble of trying to produce, produce, produce. But that's not

00:45:00.120 | always what's going on. An unstructured life can be busy and exhausting. An unstructured life can be

00:45:04.440 | intentional and more relaxed. And I think that's what we see in Surreal's story here.

00:45:08.400 | I'm going to go back and show you a couple of things that I think are important here.

00:45:11.900 | His job was flexible, right? It's a music production. He's not at an office working for a boss.

00:45:18.380 | Adding clear boundaries made it much better. Here's when I work and here's when I'm done. Without that,

00:45:23.200 | it just bled over into the rest of his life and it was causing real distress and problems. He makes deep

00:45:30.600 | work a priority. There's lots of stuff he could be doing that's probably more fun and maybe I should

00:45:35.120 | be working on the social media pages of the bands that I'm repping and try to make sure that they're

00:45:39.060 | getting attention. But he says, no, on a regular basis, I want a deep work on getting better at music.

00:45:44.300 | Long-term is probably making it much more successful. Short-term is giving him a lot more

00:45:50.140 | psychological satisfaction out of his job because it continually connects him back to the core

00:45:54.980 | artistic element of what he's doing and why he values it. He keeps social media off his phone

00:46:00.520 | and weekends and evenings are just for family. That really helps. I'm not working. I'm not on my phone.

00:46:07.300 | You're going to connect to your family much more. So these seem like simple ideas, but these little

00:46:11.540 | rules and structures made a really big difference in Surreal's life. And so I think it is a fantastic case

00:46:16.500 | study. All right, let's do a call.

00:46:21.340 | Hey, Alex Sorto here. I'm a commercial real estate appraiser. And my question has to do with

00:46:28.360 | the topic of side hustles. So there's a lot of random and conflicting advice on the internet about

00:46:36.640 | side hustles. And maybe some of it is unrealistic and some of it is not. But I would love it if I

00:46:43.240 | could create and develop a side hustle this year to keep me up float during downtimes. And I could be

00:46:49.620 | something that I actually particularly enjoy, like fashion or something along that line where it could

00:46:56.300 | be a passion as well. So I just wanted to see what were some of your ideas or best practices for

00:47:03.820 | creating a side hustle and implementing it into your regular schedule.

00:47:09.720 | seem to be like Alex referred to me as Al. Yeah, but I think he knows your name's Cal.

00:47:14.460 | You think it's because his name, he was queued up to say Alex. Yeah. And then, yeah. He definitely

00:47:21.460 | knows your name's Cal. Yeah. All right. I was going to say Alex. I'll see you in hell, buddy. But I'm

00:47:27.480 | going to, I'm going to assume that was a, that was a runtime error. I was somewhere where I was

00:47:31.900 | recently where I was continually introduced as Kyle. So would you rather be introduced by Al?

00:47:38.320 | Well, the weird thing is my wife who was misintroducing me. I think she just forgot.

00:47:43.360 | No, I made that up. Uh, side hustles. All right. Uh, three things to keep in mind.

00:47:50.580 | One, I typically would say, be wary of a side hustle where ultimately what you'll be doing is

00:47:56.220 | trading your hours for time. So basically wage labor side hustles, they don't scale well. Uh,

00:48:03.920 | you're typically getting less money for your time than you could get just putting those hours into

00:48:08.500 | your primary job. And I, it's not what you're really looking for with a side hustle, which is

00:48:13.860 | like a flexible and stable and potentially scalable source of side income. So it's tempting because

00:48:18.780 | they're easier to get into, it's easier to give, get people right off the bat to pay you for your

00:48:22.740 | time. But if you're effectively just, if I spend five hours versus 10, I'm going to get half as much

00:48:29.220 | money is not where you want to be in a side hustle. That's not a good side hustle. Um, two, I would say

00:48:35.940 | care about your existing career capital. The more that you're leveraging rare and valuable skills that

00:48:41.260 | you've earned, the more value you can get. And the more likely it is you can have like a really

00:48:46.600 | effective side hustle. So if you like something, but you don't have demonstrably rare and valuable

00:48:52.460 | skills in that area, the, the possibility that you are going to get sort of scalable time,

00:48:58.740 | independent income from that direction is very low. That type of low hanging fruit just gets

00:49:02.540 | washed out in efficient markets, right? If there's some like relatively easy way that you can make

00:49:07.460 | like a lot of money on some topic you're interested in and don't know much about anyone who doesn't know

00:49:11.900 | much about it could make money on that topic and they would start trying it. And then that would drive

00:49:15.240 | down the price and that market would go away. So the inefficiencies that produce good side

00:49:18.940 | hustles usually take advantage of skill. The more rare and valuable skill that you're deploying,

00:49:23.620 | the better chance you have of finding a real inefficiency and getting a really actually

00:49:27.780 | attractive side hustle. So don't ignore your career capital. You mentioned fashion,

00:49:31.240 | like you would really need some sort of career capital and fashion before you think that's going

00:49:35.460 | to necessarily be a good side hustle. Finally, there's an idea from Derek Sivers that I featured in my

00:49:40.900 | book so good. They can't ignore you when evaluating your current side hustle and whether you want,

00:49:46.320 | if it's successful or not successful, or whether you want it to become more of a full-time thing,

00:49:49.940 | I'm quoting them here, use money as a neutral indicator of value. All right. Are people paying

00:49:57.860 | me for this? People will tell you your idea is good. People will tell you your idea is interesting.

00:50:02.860 | People will for free, like, yeah, I would love for you to give me a tarot card reading or whatever,

00:50:07.500 | but they don't like to give away their money. And people only give away their money if they feel

00:50:12.280 | like they're actually getting a proportional amount of value for that money. So Derek Sivers had this

00:50:17.320 | idea with the side hustles that he transformed in the full-time roles on multiple occasions throughout

00:50:22.820 | his interesting career is that he would use money as a neutral indicator of value. If this is generating

00:50:27.380 | money, it's valuable. If it's not, it's not. Should I make this my full-time job? Well, here's the

00:50:31.780 | question. Is it generating enough money to be my full-time job? That seems like it's really simple,

00:50:35.980 | but people often ignore that. They want to sort of assign other sources of value to their idea. Like,

00:50:40.640 | well, it's really cool. I know someone else who did really well with this. My friends tell me it's

00:50:45.300 | awesome. It's not really a reliable indicator of value. Get people to give you money. And if they're

00:50:50.880 | not, the idea is not that successful. The hustle is not going to be that successful. All right.

00:50:57.520 | Alex, I hope you enjoyed that advice. Jesse, I sort of reversed it back on him there.

00:51:02.400 | Added a C to his name. All right. Do we have another call? Yeah, we do. All right.

00:51:06.600 | Hi, Cal. I enjoyed your recent podcast, looking at how to become a straight A student as I am a college

00:51:14.120 | student. And that was my initial introduction to you. My question is, you talked for a second that

00:51:21.360 | if you were going to rewrite that, that you would have a section on focus. And I know that you said

00:51:27.700 | that you'd probably never rewrite it, but I was curious what principles you would go over in a

00:51:32.520 | section like that. That's something that I really struggle with and I'm trying to figure out how to

00:51:37.700 | implement that as a college student. So my basic framework about focus is in the book and when it

00:51:46.240 | changed. So there's a core idea in that book, how to become a straight A student from 2006. It's an

00:51:51.080 | equation that says work accomplished equals total hours spent times intensity of focus. And my point

00:52:01.200 | there, like why I made that argument was one of the easiest ways or most effective ways to get

00:52:07.520 | more work done is to not just increase the hours piece of that equation, but to increase the focused

00:52:14.100 | piece of that equation. They're multiplied by each other. Double your focus, you can double your output

00:52:18.740 | in the same number of hours. And back then I was saying what I saw is too many people would study in

00:52:25.300 | a low intensity of focus. So sort of at the library with their friends and reading instead of doing

00:52:30.160 | more difficult, like passive consumption instead of active recall. And they would just try to increase

00:52:34.440 | the hours component to get the amount of work that needed to be accomplished done.

00:52:37.340 | But that's a hard game. Like you got to spend long amounts of time doing that work. And I would say,

00:52:42.460 | no, no, increase that intensity of focus. Come at this thing with laser light focus and you can

00:52:46.200 | drastically reduce the hours needed to get the work done. And that's actually more sustainable in the

00:52:50.200 | long run because you're not spending all your time studying. That central equation still stands.

00:52:54.660 | What's different is the obstacles to intensity of focus today versus 20 years ago.

00:53:01.560 | You have way worse obstacles in particular in your hand, in that phone that's next to you while you

00:53:08.900 | study. That is a much bigger obstacle than we had to deal with in 2006. I was worried about

00:53:15.420 | conversations with people at the same table. I was worried about you just using low intensity

00:53:19.700 | techniques. So using study techniques that weren't very mentally taxing. Therefore, they didn't require

00:53:24.060 | much focus, but required a lot more time. And like walking over to public email terminals to like check your

00:53:30.200 | messages. Like that's the type of stuff I worried about. So you have obstacles to boosting that

00:53:35.100 | intensity of focus now that are way more accessible and appealing and disastrous than what we were facing

00:53:41.600 | back in, in 2026. But I would have one rule, one extra rule that I think translates most of the value of that

00:53:51.280 | equation to today. Never have your phone with you when you study. That would be the rule I would add.

00:53:58.200 | And don't look at this like this is crazy. You're not an ER surgical attending. You're not the Pope.

00:54:05.860 | People will be okay if they can't get in touch with you for 90 minutes. It's okay. You're not that

00:54:10.240 | important. Don't bring your phone with when you study. Go to a library to study. Go to a non-popular

00:54:15.260 | library to study. Don't have your phone with you. Now you just don't have access to those distractions,

00:54:22.220 | right? That's going to make it. Now you can just apply the advice as written in my 2006 book. Now

00:54:27.400 | still you have to use high intensity study tactics. They're going to get you to focus more. Use active

00:54:31.940 | recall instead of passive recall. You know, go to places where you don't have a lot of distractions.

00:54:36.780 | There's like all the stuff I talk about now, now applies equally as well in 2026 as it was in 2006,

00:54:42.640 | as long as you don't have that phone with you. So it would be like my number one advice. You would want to

00:54:48.480 | put a giant sign over every library or public study space that, you know, says, um, abandon all ye phones,

00:54:56.320 | ye who enter here. A little Dante reference right there. And that, that puts you back on a level playing

00:55:03.260 | field. And then you actually have a chance of succeeding with focused high intensity studying.

00:55:08.420 | Let me tell you the effect of this. It will feel like a superpower. You will, and I'm not exaggerating,

00:55:16.120 | study a third of the time of the other people in your class and you'll get good grades.

00:55:20.180 | If you study without your phone and you use smart study techniques, like the type I talked about in

00:55:26.040 | my book that actually work, they're hard, but actually work and not what's easier. And you boost

00:55:30.040 | up that intensity, you focus part of the equation and you do the time management stuff. We talked about

00:55:34.960 | a couple of weeks ago on the podcast that's in that book. You're gonna get really good grades.

00:55:38.900 | You're not going to work late into the night. You're not gonna do all nighters. You're gonna

00:55:42.740 | have a lot of free time. So that's it. That would be the number one thing I would add. You just, just

00:55:47.300 | don't, don't have screen time rules. Don't get cute about it. It is just not allowed to come with me

00:55:52.880 | when I go to study. That'll make all the difference.

00:55:56.040 | For the newer students in the audience, can you, in a couple sentences, explain to them what active

00:55:59.940 | recall is?

00:56:00.540 | The only way to learn new information that actually works with any sort of effectiveness is trying to

00:56:05.840 | produce the information out loud from your own brain as if lecturing a class without looking at notes.

00:56:10.180 | That's how you learn things. The opposite of that is passive recall, which is where you read your

00:56:15.700 | highlighted notes. That is an incredibly inefficient way of learning stuff. I'm going to give you an

00:56:23.060 | analogy. Let's make the analogy trying to get like muscles larger. Passive recall is, uh, I'm going

00:56:30.660 | to hold a shake weight. Have you seen the shake weight, Jesse? They used to sell these in infomercials.

00:56:35.120 | You hold, it's like a dumbbell where the two ends go and they shake back and forth, like powered by

00:56:40.520 | batteries or whatever. And like, it's shaking your muscles and like, it's going to kind of make it

00:56:44.400 | stronger. I guess you're like kind of holding it up or something. Active recall is like, I'm getting,

00:56:48.540 | I'm doing curls with heavy weights and it's kind of sucks because it's like, it's burning.

00:56:52.840 | But I'm doing them, you know, on a really good schedule. Like your muscles actually get easier.

00:56:56.780 | It's harder, but your muscles are going to get bigger. You can stand around with the shake weight as

00:57:00.400 | much as you want, but, uh, maybe get a little bit stronger. So there you go. Active recall is the only

00:57:06.580 | thing that matters. It sucks in the sense that it's really hard, but you have to think of that hardness

00:57:11.620 | as like the, it's not literally this, but like the feeling of your neurons reforming. You're forcing

00:57:16.980 | your brain into a new configuration where that information is accessible. Lecturing out loud as if

00:57:21.680 | lecture in a class. I, I preach this to my students like five or six times a semester. I even wrote, like

00:57:26.860 | I'm teaching discrete math or I was teaching discrete math, uh, to a hundred kids this semester.

00:57:31.240 | Is that what all those exams are?

00:57:32.560 | That's what the, yeah. Do you see those?

00:57:34.040 | I looked at the first question.

00:57:35.260 | Yeah. Not so bad.

00:57:36.300 | Yeah.

00:57:36.880 | I don't know.

00:57:37.180 | Discrete probability.

00:57:38.620 | So I have a, uh, this feels like a FERPA violation, but I was grading this weekend. And so I have all

00:57:43.500 | my exams, uh, in the, in the HQ cause I was grading the HQ, um, the other day, but I always tell them,

00:57:49.640 | I wrote an article about this when I was in grad school because I did well, uh, I got the highest

00:57:55.620 | grade. I've talked about it before, but like, that's when I realized I might have some math

00:57:58.380 | ability when I took this large discrete math classes on the ground and got the highest grade

00:58:01.760 | in the class. And so I wrote an article about, here's how I studied in that discrete math

00:58:05.480 | class. And I'm always telling my students, I wrote an art, do this, do this. Maybe they

00:58:11.800 | do. They do. Uh, what's a pretty good grades. That wasn't a hard exam. I don't think it was

00:58:17.000 | very reasonable exam. Um, so there you go. Active recall. All right. Let's, uh, what we got here?

00:58:23.280 | We got a final segment coming up. I'm going to react to something on the internet you should

00:58:27.340 | know about, but first let's briefly hear from another sponsor. One of the longest sponsors

00:58:33.600 | of this show and for good reason is our friends at Grammarly. That's because from emails to reports

00:58:38.100 | and project proposals, it's more challenging than ever to meet the demands of today's competing

00:58:42.260 | priorities without some help. Grammarly is the essential AI communication assistant that boost

00:58:48.300 | productivity. So you can get more of what you need to get done faster, no matter what it is

00:58:53.980 | you're writing or what tool it is that you're doing the writing in. I have watched Grammarly over the

00:59:00.520 | years embrace AI enabled features that have continued to grow its utility and producing professional and

00:59:07.700 | effective communication in your job. It'll produce drafts. If you want like, Hey, I got to do a quick

00:59:15.760 | draft of this. It can take a bunch of notes. You paste in there and it can put them into a better

00:59:19.340 | summary. It can look at your tone. Hey, what is the tone of this email I sent? Why don't you double

00:59:25.200 | check that for me? Um, and maybe I want to change it to like a more friendly tone or more professional

00:59:29.520 | tone. Maybe this is, you know, you worry about how you're coming across. It's there. It can help you

00:59:33.460 | with that. Maybe you're trying to come up with some ideas or of course the stuff that Grammarly has

00:59:38.880 | always done really well. Just check to make sure that like your, your grammar is good. And this is coming

00:59:42.700 | across as professional. I think of having Grammarly, like having a, uh, editor who looks over your

00:59:47.880 | shoulder with a bunch of ideas and to help you make your writing even better. Like you could hire

00:59:51.480 | someone to be like, let me help you with this email before you send it. Just think about how those

00:59:55.120 | advantages aggregate when you have like help with clarity and effectiveness in your communication

01:00:00.600 | and all the communication you're doing in your job, that's going to add up to you just being seen as

01:00:05.120 | like a much more capable and effective employee. So it's a fantastic, I think focused deployment of AI,

01:00:11.680 | not some grandiose, you know, this is going to take over the world. Like this is a specific

01:00:16.420 | thing that people need to do better in their jobs. And we're going to help you do it.

01:00:20.000 | You can write and edit quickly. 93% of professionals report that Grammarly helps them get more work

01:00:25.860 | done. You deserve to be equipped with the best tools to do your job with the right AI support.

01:00:30.360 | And that is Grammarly. So let Grammarly take the busy work off your plate so you can focus on high

01:00:35.200 | impact work. Download Grammarly for free at grammarly.com slash podcast. That's grammarly.com

01:00:41.280 | slash podcast. You think I would know that word by now? Podcast. I also want to talk about speaking

01:00:48.660 | of technology, our friends at Vanta. Trust isn't just earned, it's demanded. Whether you're a startup

01:00:54.440 | founder navigating your first audit or a seasoned security professional scaling your GRC program,

01:00:59.900 | proving your commitment to security has never been more critical or more complex. That's where Vanta

01:01:06.340 | comes in. Businesses use Vanta to establish trust by automating compliance needs over 35 frameworks like

01:01:12.560 | SOC 2 and ISO 27001. Centralized security workflows complete questionnaires up to five times faster and

01:01:19.100 | proactively manage vendor risk. Vanta can help you start or scale your security program by connecting

01:01:25.040 | you with auditors and experts to conduct your audit and set up your security program quickly. Plus,

01:01:30.340 | automation and AI throughout the platform, Vanta gives you time back so you can focus on building

01:01:35.440 | your company. Join over 9,000 global companies like Alassian, Quora, and Factory who use Vanta to

01:01:42.100 | manage risk, improve security in real time. Here's the good news. My listeners can get $1,000 off Vanta

01:01:49.020 | advanta.com slash deep questions. That's V-A-N-T-A dot com slash deep questions for $1,000 off. All right,

01:01:59.740 | Jesse, let's do our final segment. I want to talk briefly about an article that's important and you

01:02:05.780 | might not realize how important it is. I'll bring this on the screen here for people who are watching

01:02:11.360 | instead of just listening. It's from the Wall Street Journal. It came out on May 15th. Here's the title.

01:02:15.780 | Meta is delaying the rollout of its flagship AI model. The company's struggle to improve the

01:02:20.980 | capabilities of latest AI model mirrors issues at some top AI companies. So what's going on here is

01:02:27.780 | that Llama 4, which they're nicknaming Behemoth, had been the long-awaited next big large language model

01:02:36.900 | from Meta. They were greatly increasing the amount of compute and data, but when they finished,

01:02:44.680 | they realized this thing wasn't that much better. It wasn't better enough versus their last model

01:02:49.300 | to justify yet releasing it. Now, there's an important quote in here. I'm looking through the

01:02:54.400 | article now where other companies have been having similar problems. All right, let me read from the

01:02:59.420 | article. Meta's recent challenges mere stumbles or delays at other top AI companies that are trying

01:03:04.680 | to release their next big state-of-the-art models. Some researchers see the pattern as evidence that future

01:03:09.400 | advances in AI models could come at a far slower pace than in the past and at tremendous cost.

01:03:14.020 | Right now, the progress is quite small across all labs, all the models, said Ravid Schwartz-Ziv,

01:03:19.540 | an assistant professor and faculty fellow at New York University's Center for Data Science.

01:03:24.500 | GPT-5, one of OpenAI's next big technological leaps forwards, was initially expected around mid-2024,

01:03:30.980 | the Wall Street Journal previously reported. In December, the journal reported that development

01:03:35.900 | of the model was running behind schedule. All right, and they give some more examples. All right,

01:03:41.120 | we can take this off the screen now, Jesse. This is pointing towards a very, very important trend in

01:03:46.740 | AI, but the details are a little berky, and I think most people don't know it, and I want to make it as

01:03:51.460 | clear as possible. Here's the best way I can explain this. There was, in the aftermath of OpenAI,

01:03:59.600 | starting to make their moves with their GPT family of models, so in particular, GPT-2 and 3.

01:04:05.520 | There was this idea, they're called the scaling laws. There is this idea that as you continue to

01:04:13.160 | increase the amount of computation and the amount of data that you use to train these language models,

01:04:18.800 | their skills would keep massively jumping forward, and in fact, the chart that showed this relationship

01:04:25.260 | between resources and ability was exponential, and for a while, it was fitting that. GPT-3 was like this huge

01:04:33.260 | improvement over GPT-2. GPT-4 was a huge improvement over GPT-3. It was better at like everything. It was doing

01:04:40.900 | it better. The visions you heard about a year or more ago about like artificial general intelligence,

01:04:49.440 | for example, about all work being essentially automated, was people looking forward on this curve

01:04:58.500 | and saying, this was so much better than that model, this model, so much better than this model. As we

01:05:04.220 | continue to build out the compute and get more and more data, eventually, these models will just

01:05:10.480 | become so capable that we can just build software agents around them to actually execute the stuff

01:05:17.000 | they say. We can just say like, hey, here's a job. Tell me how you're going to do it and have a software

01:05:20.520 | agent actually execute it. And basically, we can automate everything. It will just be able to understand so

01:05:25.820 | much about the world that we can just ask it, describe how you do this or that, and it'll be a

01:05:30.260 | general intelligence that can basically do anything that a human can do. This is going to completely

01:05:34.880 | change the world. This was the scaling vision was the vision for what was going to happen.

01:05:40.620 | The reality, however, is that the benefits of scaling are starting to give massively diminished

01:05:48.620 | returns. It turns out that like this exponential doesn't keep going. Hundreds of billions have been

01:05:54.480 | invested in building out compute and in building out the biggest possible data sets to train these

01:06:01.080 | things. But after roughly that GPT-4 point, the improvements begin to become diminishing. That's

01:06:06.100 | why GPT-5 hasn't come out yet. It's not good enough. It's not that much better. It's why

01:06:11.580 | Llama 4B myth project, they haven't released it yet. It's not that much better. And in the cases where

01:06:17.120 | they have put this in, you look at like Grok, I guess it's three now versus Grok two, massively more

01:06:22.820 | compute. It's a little bit better. So the scaling laws are kind of trailing off. So we're not going

01:06:29.840 | to, this dream of like, if we keep building bigger and bigger data centers, we don't have to do anything

01:06:34.660 | else but get bigger. And we're going to have like AGI. That's not happening. I don't think everyone's

01:06:38.620 | caught up to that yet. There's still so many of these sort of non-technical articles about like,

01:06:43.320 | hey, what are we going to do in a future? We're going to have no jobs. They're still looking up

01:06:47.680 | this curve. Well, look at how much better this model was in that model. If we just do that five

01:06:51.500 | more times, it's the danger of exponentials, the seduction of exponentials. We do that type of

01:06:55.780 | doubling inability five more times and like all work will be done this way. But that's starting to level

01:06:59.900 | out. So where is all the energy shifted? Because they still need to make improvements, right? You've

01:07:04.180 | noticed OpenAI has now released all these models that have weird names. It's GPT-4-0 and GT-4-0-3 and

01:07:12.820 | all these different types of weird names. Well, they're having to put their energy into much more

01:07:17.380 | narrow development. So a lot of what's happening now is they'll take what they call a foundational

01:07:21.340 | model, which will just be an existing model like GPT-4. And then what they're going to do is they

01:07:26.060 | get a particular data set about a particular type of task they care about where they can have a lot

01:07:31.920 | of examples of questions and write answers. So computer programming is a fantastic example of this is

01:07:37.360 | because you can have a question saying write a computer program that does this and then you can

01:07:42.460 | compile the answer and test it and see if it actually does it. Math is another area where they

01:07:47.480 | could do this because you can write math problems and have the right answer, right? And so they take

01:07:51.880 | an already trained foundational model like GPT-4 and they do a tuning step with these type of synthetic

01:07:56.940 | data sets with right answers where they ask it these questions. And if they give the wrong answer,

01:08:02.640 | they use reinforcement learning to kind of move its weights away from that. And if it gives a

01:08:06.580 | right answer, then they use reinforcement learning to make its weights more solidified towards that.

01:08:11.120 | And they can tune the base language model to be better at these specific tasks for which they

01:08:16.960 | have big synthetic data sets with examples of right answers. That's what they've been doing since the

01:08:22.040 | scaling is not working as well. So they're like, look, this model is good at computer programming.

01:08:27.040 | Why? They built a huge synthetic data set of computer programming examples. Over here, we have a program

01:08:32.340 | that's getting really good at math. Why? Well, because we're hiring a bunch of math PhDs and paying them $100 an

01:08:37.420 | hour to write all these problems where here's a problem, here's the answer. We're using that to tune

01:08:41.240 | foundational language model. The reasoning quote unquote advances that came recently was they worked through

01:08:47.080 | built out these big synthetic data sets of questions that require reasoning and then step-by-step

01:08:52.220 | answers. And then they could use that with reinforcement learning to train certain models to be better at like

01:08:58.060 | doing something that looked like that type of step-by-step reasoning. So that's

01:09:02.040 | what's really going on now. We're not just getting this like massive general increases in

01:09:06.980 | overall capability by just scaling compute and data. So instead, the companies are saying,

01:09:11.200 | we're looking for these like specific applications where we can have these really good synthetic data

01:09:15.940 | sets. And then we'll build bespoke models that are like better at that particular thing.

01:09:20.660 | And maybe they get worse at other things. So like the latest reasoning tune engine of open AI better

01:09:27.120 | match this step-by-step reasoning. They trained it with reinforcement learning, but it hallucinates a lot

01:09:32.440 | more. Whoops. Because in trying to actually like match that, it like it helps to make things up more

01:09:38.020 | because you can better like get a logical seeming chain of answer. So the wrong thing to do now is that

01:09:44.940 | this is not now a general extrapolation. You can't say like, now it's good at math and now it's good at

01:09:50.640 | reasoning and now it's good at computer programming. So tomorrow it's going to be good at X. It's like,

01:09:54.060 | no, no, no. These aren't additive or aggregate. GPT-4 is kind of like their state-of-the-art base

01:09:59.880 | model. If you're looking at open AI and they're looking for specific applications where you can get

01:10:04.440 | good data sets for reinforcement learning on particular tasks. Now this is producing and will produce

01:10:09.140 | cool bespoke tools. It's going to be a long product development cycle. You got to figure out places

01:10:14.020 | where you can build these type of tuned models and then build good software agents around them and

01:10:17.860 | integrate them into the workflows where people use them. But it's a far cry from the vision that we

01:10:22.560 | people had two years ago of like, no, no, just GPT-6 will be anything you ask it. It'll do at a human

01:10:28.780 | level. In fact, it can write its own code to make agents around it and we won't even have to like work

01:10:33.820 | anymore. That is not the trajectory we're on at the moment anymore. Without the scaling laws,

01:10:38.300 | we have to sort of hand build bespoke models that can do specific tasks where we happen to have data

01:10:43.440 | for it. There's many things that we do in our lives where there's not going to be good data sets for,

01:10:47.600 | but where there are, I think they're going to eventually have really cool tools. This is a cool

01:10:51.500 | technology, but it is not on one of these, I have to write an article about the end of business

01:10:58.160 | type of trajectories. So anyways, this is like a really important thing that's going on. This is being

01:11:02.760 | discussed a lot among like the high executives in these companies, the death of the scaling laws,

01:11:07.460 | trying to build out these bespoke skills. I mean, so I'll just say real briefly. So why then did we

01:11:14.400 | have like the AI 2027 report come out last month saying, no, by 2027, these AIs are going to be in

01:11:22.960 | charge of everything. And by 2030, humanity might be dead. Well, I look closer at like what was really

01:11:28.420 | going on there because they know the scaling laws aren't working, right? So what's their tech story for

01:11:34.220 | for how AI is going to kill humanity in the next like three years, right? If you look closer,

01:11:39.200 | it all comes down to a single assumption, which I think is specious. The tuning we're doing for

01:11:44.840 | producing computer programs is getting better, which is true like that. We have good data sets for that

01:11:49.660 | and computer programs are text and we can produce a reasonable size, good computer code.

01:11:55.160 | the foundational assumption of AI 2027 is, well, we're better at computer programming with models

01:12:01.400 | now than we were before. So maybe if we keep getting better at that in the next year or so,

01:12:06.380 | the models will just be able to build their own. They'll solve the problem of how to build a more

01:12:10.720 | technically advanced AI. Yeah, we can't just scale language models. We don't know the tech story for

01:12:14.440 | what type of models would be super powerful or super intelligent, but maybe if we just make them

01:12:18.620 | better at programming, they'll figure it out in the year 2026 and then we'll be in trouble.

01:12:22.660 | That's a crazy assumption. Anyone who's actually working with the tuning of language models to

01:12:27.680 | produce computer code to think like, yeah, that's going to lead to giant breakthroughs and architectures

01:12:33.640 | for artificial intelligence. It's a, it's a crazy claim. It's like saying the Wright brothers being

01:12:41.580 | like, we've, we just found the new type of propeller and now we can, we can get powered flight. Now it used

01:12:48.000 | to just be for a little bit on Kitty Hawk beach. Now we can actually go up and circle the field and come

01:12:51.880 | down and do like 10 minute flights. So like, I don't think we're that far away from interstellar

01:12:56.220 | travel. It's like different. I mean, they're both flying, but it was like completely different levels

01:13:01.900 | of complexity and challenges. So anyways, this is an important story. I, you know, I'm, I'm increasingly

01:13:08.260 | believing there's this like weird temperature raising on AI that is becoming more disassociated

01:13:13.280 | from the tech story. And so not to talk about AI too much, but this I think is an important story

01:13:18.620 | understanding that the scaling laws seem to be faltering for language models puts us into a much

01:13:24.100 | more traditional business innovation cycle landscape for, for AI tools right now.

01:13:29.160 | All right, there we go. I'm writing a massive, actually it might be out by the time this, this

01:13:33.900 | comes out on my newsletter. I'm writing a big AI and work article where I'm just dumping a lot of

01:13:38.680 | this stuff I've learned in the one article that I can point people towards. So check that out at

01:13:42.640 | newport.com. It should be live when you hear this podcast episode. Nice. I've just been deep on this

01:13:47.240 | in a while. I was like, I don't want to keep talking about all the time. Let me just put down like,

01:13:50.080 | Hey, here's what I think is going on. And it's an article, at least in the draft that is now,

01:13:53.620 | here's the places where like AI is really affecting the world of work. Here's the places where I think

01:13:58.260 | it will soon. It hasn't been developed yet, but we'll soon here's some places where it's like a

01:14:02.320 | little more hazy, like how exactly is it going to play out, but there could be some cool stuff.

01:14:05.900 | And here's the stuff that's crazy town. I try to cite a lot of things and make the tech story a little bit

01:14:10.880 | clear. So look for that article, calendlyport.com. It should be up. All right. Speaking of which,

01:14:16.120 | I should probably go work some more on that. So let's call this episode to a close. Thank you all

01:14:20.540 | for listening back next week with another episode. And until then, as always, stay deep. Hey, if you

01:14:26.920 | like today's discussion about slowing down in the summer, I think you might also like episode 343,

01:14:31.960 | where I present a minimal protocol for taking control of your life. It's a great compliment to today's

01:14:37.820 | discussion. Check it out. I think you'll like it. So one of the conflicts we've confronted here

01:14:43.120 | is the one between having too little productivity and too much.

The Summer Schedule Reset To Save You From Burnout & Unlock Deep Work

Chapters