How we traded beauty for efficiency

A conversation with David Perell

In case you missed it… we released our annual publication: Whoa, Vol. 2: Conversations on AI and Creativity. We’ll be releasing one conversation here each week.

If you’d rather savor instead of scroll + want early access to all ten conversations, including Oliver Burkeman, Venkatesh Rao, and Anu Atluru, grab the physical + digital zine. And if you already ordered your copy, the next batch will be shipped this weekend.

Only 11 print copies left!

---

This series is made possible by Mercury: business banking that more than 200k entrepreneurs use and hands down my favorite tool for running Sublime.

Running a company is hard. Mercury is one of the rare tools that makes it feel just a little bit easier.

Every other banking platform makes me feel like I’m doing taxes. Mercury makes me feel like I’m building something. The interface is unbelievably intuitive, fast, and lets me handle everything in one place – banking, cards, spend management, invoicing, and more.

If you’re a business owner of any type, visit mercury.com today and find out why so many of us actually love how we bank.

**Mercury is a financial technology company, not a bank. Banking services provided through Choice Financial Group, Column N.A., and Evolve Bank & Trust; Members FDIC.

---

A conversation with David Perell

There are a few people who’ve changed how I think about what it means to create a life and a living on your own terms. David Perell is one of them. He’s carved a way of being and operating that you can’t put a label on — mixing writing, teaching, podcasting, and filmmaking in ways that feel so distinctly his.

If you’re new to his podcast How I Write, this conversation with Ezra Klein is one of my favorites. In addition to releasing weekly episodes with God tier writers, this week he released a beautiful short film called The Modern World exploring the decline of aesthetics and how we traded beauty for efficiency.

I want to be more like David and in many ways, I’m building Sublime to make the kind of work he does possible for me — to go deeper, connect dots, and stay in conversation with the ideas that move me.

This was an incredible conversation, full of insights he hasn’t shared elsewhere including:

• His decision to shut down Write of Passage and why the old funnel of “write, get a follower, and they’ll be loyal” has disappeared.

• The event-based model of publishing he’s betting his career on.

• His brilliant breakdown of how San Francisco, New York, and Los Angeles each see AI differently.

• How he uses AI to laugh, play, and get weird.

• and lots more…

Or listen on Spotify and Apple.

(Best if you want to highlight your fave moments with Podcast Magic).

---

The edited transcript

Alex Dobrenko: This is broadly speaking a conversation about how cool people are using and thinking about AI. What role is it playing in your life?

David Perell: The AI stuff is definitely interesting. It’s the story of our time. But beauty and craftsmanship is really where my heart is. And there’s definitely a way that these weave together that I’ve never spoken about in public and probably haven’t even framed my thoughts on completely.

AD: What I find really interesting is that you bring up beauty & craftsmanship in contrast to the AI stuff. Talk me through that and why you’re obsessed with it.

DP: Well, just look around. The things that we’re making aren’t nearly as beautiful now as they were 150 years ago. When I talk about beauty, I’m not talking about the Paris Opera House. I’m not talking about the exterior of the Sagrada Familia with all the ornamentation. I’m not saying we need to live in some Rococo world, it’s just really simple things. You go to London and the telephone booths have been decommissioned, but they’re still on the street because they’re just so nice to look at. And you look at old radios from the 1920s, people would just collect them because they enhance the room that they’re in. And then you look at the kind of radios that we used to buy at Radio Shack as kids and they’re just boring and mechanical and that story is showing up all over the place and it’s such a tragedy to me. One of the core things that you can measure society on is how beautiful the things that they make are. And I think on that dimension we’ve just completely declined.

AD: But isn’t there a lot of beautiful stuff now?

DP: It’s not like there’s nothing beautiful that’s made anymore but I believe that anyone just by virtue of being a citizen should be able to experience beautiful things as they go about their day to day. I have this podcast called How I Write. And it’s really important to me that we record in beautiful spaces. And basically as I go West in the world, it gets way harder to find beautiful spaces. If I’m in London, I have five rooms where I can record that are stunning. And if I’m in San Francisco or Los Angeles, where literally the biggest wealth boom of the last 20 years has been, they just don’t have those spaces. And that drives me nuts. Like what’s going on there?

AD: What’s going on there?

DP: Well, I think that there’s a few things. I remember my first time in New York. I got out of Penn Station and there’s the Monahan Post Office across the street. I knew nothing about aesthetics, but I saw the yellow cabs and I saw the Corinthian style columns right across the street. And then you walk in, there’s tall ceilings, there’s murals, there’s handcrafted tables, there’s marble all over the place. All basically saying that this isn’t a place we’re just gonna drop off and pick up mail. We’re gonna reach beyond utility and we’re gonna make this space a celebration of the American spirit.

And that’s what a post office was. And even now, if you take a trip to upstate New York, the most beautiful building in the towns will be a post office that was built between 1880 and 1930. It’ll be charming and quaint. But what happens when you walk into the post office? It sucks. You go in, there’s fluorescent lights, no one cares. Going there is the worst part of your day. And this just shows up everywhere.

AD: Yeah. I mean, I think about it in terms of libraries as a place I love going to.

DP: So libraries are slightly different from the post office example, but equally good. The post office example was by the government. The library example was wealthy people doing philanthropy who wanted to give beauty back to the cities. Carnegie built more than 2,000 libraries, and Carnegie really valued the architecture of them. So one of them is government values, and another one is the values of philanthropy.

AD: So why do you think that is in such steep decline?

DP: I think it goes back to the work of Ian McGilchrist, who wrote The Master and His Emissary. His main idea is that we’ve suppressed the right brain, which sees unity and totality, in favor of the left brain, which sees the world in logic and deconstruction. The left brain isn’t capable of seeing the entire picture, and that’s what happened to our culture.

The core shift began with the Enlightenment, which valued reason and logic more than the religious world that came before it. I look at painting and architecture to understand this. We think of aesthetics as superficial, but style is an external expression of a culture’s inner spirit. What we make is who we are. When you compare buildings from the 1970s to those from the 1930s, or even Gothic architecture, you can see how we cast our values in brick and stone.

Some things began to change around 1907 with new technologies like airplanes and taller towers. Painting styles shifted from peaceful Impressionism to harsher Cubism, and the Italian Futurist Movement celebrated speed and technology. Then, World War I came and completely crushed the spirit of Europe. This led to a big fracturing of artistic styles like Russian abstraction and surrealism. We can see on canvas that something fundamental had changed.

World War I was so devastating that we had to make sense of who we were as a people. There was some optimism, especially in America, where the Art Deco movement took hold as a way to get away from the devastation. It was like a football coach at halftime trying to rally the team.

World War II, however, did that for the entire world. After the second great war, styles became more rigid, with right angles and abstraction. You get Warhol. You get Basquiat. Architecture became more militaristic and reason-based. Modernity and ornamentation became antonyms, not synonyms. This is sort of my big-picture history of what happened, and I think we’re still in that era, with engineers who run the world and perpetuate left-brain thinking.

AD: Yeah, I mean the binary of coding is that in the digital realm.

DP: Exactly. Exactly.

AD: Why does that stuff come up in contrast to the AI stuff for you?

DP: Good question. Well, let’s talk about writing. In the 18th and 19th centuries, writing was often maximalist, full of prose and metaphors. Then Hemingway came along. He was a genius who wrote very simply, but with an intentional depth. He knew what came before him, but chose a different approach.

The same thing happened with minimalism and companies like Apple and Blue Bottle. They absolutely nailed it. But then came the copycats. They looked at the tree above the surface, but forgot about the root structure. They copied the surface, and what underpinned the original began to disappear. This created a poor imitation of minimalism, which is just laziness.

My favorite example is the meme of a guy’s bedroom with a mattress on the floor and a gaming console. He says, “it’s minimalist, bro,” but it’s just bad design. There’s so much of this “crap” and “slop” that passes for minimalism.

This same laziness is showing up with AI. If you ask AI to design a room, its default will often be to copy the minimalism of our time. True Japanese minimalism, like in a temple, is beautiful and meticulously thought through—the complete opposite of laziness. But bad minimalism has become synonymous with laziness.

AD: Have you read Filter World by Kyle Chayka? His thesis is that coffee shops are the same everywhere. I think he’s not wrong, but I’d tell him he’s not looking hard enough. There are still cool, weird things out there.

DP: Yes, he wrote a piece called “Welcome to Airspace“ that was one of the first things that turned me on to this.

I know what you mean. Of course there’s cool new stuff, but the new cities and large-scale projects built by financial planners and the government just have this sterile, bland quality. For example, I’ve been living in Austin for five years, and the downtown is so ugly it hurts my soul. The buildings all look the same. And yet, there are individual places in Austin that completely rebel against that aesthetic. So while there are individual places that buck the trend, the overall trend is an undeniable bulldozer that’s consuming everything in its wake.

AD: And AI is gonna accelerate that, you think?

DP: I think it remains to be seen. I use AI every day. I would be so sad if I didn’t have it. I’ve probably prompted it 12 times today and it’s 1:25 in the afternoon. It’s my right hand. I talk to o3 more than I talk to any individual human being. That’s crazy.

And if it went away today, I’d be devastated. And yet I do think that the default thing that people use AI for is outsourcing their thinking to whatever AI is saying. You walk into coffee shops and if you just sort of tilt your eyes to the periphery and you watch people use AI, they basically take the answers, copy and paste them. They don’t really question them. They treat it as dry clay, not wet clay. And that is no way to use AI.

It’s what I call the paradox of abundance. As you get more abundance, what you get is people who get this massive increase in how well they can use it, and then it hurts the average person. An example would be what’s happened with food and obesity. If you walk into the Equinox in Beverly Hills, or on Prince and Broadway in New York, you see these beautifully sculpted bodies, six packs, all that stuff. And yet, if you look at the obesity rates, people are getting fatter, they’re getting unhealthier, there’s an increase in chronic disease. I think that same pattern is going to show up in AI.

AD: I want to talk about your decision to stop Write of Passage. You tweeted that AI played a role in that. Talk me through that decision.

DP: AI is part of a bigger story which is that the internet is in a new era. Write of Passage was based on a simple idea: write consistently, get a following on social media, build an email newsletter, and your life will change. It worked. Our students had amazing outcomes.

But then the internet changed, starting with TikTok. We went from a follower-driven graph to a content-driven algorithm. Instead of distributing content based on who follows you, platforms now look at the content itself and spread it to anyone who might be interested. This has caused the value of a follower to go way down. On Twitter, I have people with a fraction of my followers who get more impressions than me. On YouTube, what matters is your click-through rate and average view duration, not subscriber count.

This meant that what we were teaching at Write of Passage was no longer an honest approach to success. Now, the way I approach writing is to think of a piece as an event. Instead of publishing every week, I might work on a piece that I want to get into the New York Times. The old funnel of “write, get a follower, and they’ll be loyal” has disappeared.

The flip side is that things that really pop can explode in a much bigger way. The tweet with the stick figure that said “tell me a piece of lore about yourself” got billions of views from someone with almost no followers. That’s a fundamental shift, and it’s one of the reasons I wanted to shut down Write of Passage.

AD: But with that stick figure tweet, no one knows who put it up, and no one cares.

DP: Yeah, no one knows. So what’s happened is, strategically, the creators used to benefit more and now the platforms benefit more.

AD: Substack kind of bucks that.

DP: Substack does buck that trend, but you have to be very good to succeed there. There’s people like Ted Gioia who I love. I’m a huge fan of his writing. He’s an amazing thinker, he’s an amazing writer. I’m just much less bullish on the ability for Joe Schmo from down the street or Simple Sarah to just spin up a blog, get an audience, and have their life changed like I was 10 years ago.

AD: Right, you don’t feel confident teaching that anymore because it’s much harder to pull it off.

DP: Yeah. I’m also in a different creative phase. Write of Passage was a volume play—a bet on having thousands of students. I’m not gonna be able to really know most of my students and there’s gonna be a few thousand of them every single year. I’m gonna sit in this very chair. I’m gonna teach and sort of preach the gospel of online writing. And I’m actually less interested in that. And I discovered this over time. I’m less interested in working with a lot of people. I’m much more interested in working closely with a few excellent people who I can just invest in, get to know. And not everyone I worked with felt like that. And that’s ultimately why I didn’t want to run it anymore and frankly why I was kind of a bad leader for the business. What I wanted was very different from what the business needed.

AD: Are you doing that now? Helping, working with people?

DP: Yeah, I’m working on a documentary with my friend Sheehan Quirke from The Cultural Tutor and we’re gonna publish it. It’s about why the world is becoming less beautiful. We’re publishing a pre-pilot soon and then we’re gonna get it on Apple or Netflix. Everything is sort of fewer projects, much bigger if they go well.

AD: And so, maybe I’m not just following, but like the AI piece, how did that fit into the decision?

DP: Writing now has to be really, really good to stand out. People got really upset last December when I tweeted that AI’s writing is already better than the majority of Write of Passage students would be with a day’s worth of work. It made a lot of people upset, but I think it’s true. AI’s writing is great with a good prompt, which is why people outsource their writing to it.

For a piece of content to spread now, it has to be good enough to cut through the noise and be worth the opportunity cost of reading AI. I spend almost as much time reading AI as I do reading articles, and I think that balance will tilt more toward AI in the future, like with deep research reports.

When we think of writing quality, we think of two things. The first is intrinsic quality, like the poetry of Joan Didion. I agree that AI will probably never write poetry like David Whyte. But the second is how tailored a piece of writing is to our exact interests. For example, I’m moving to the West Village in New York and I’m gonna wanna learn all about the neighborhood, so I’m gonna do deep research reports to learn about the West Village, not because the writing is better than the best book written about the West Village but because it can answer my specific questions immediately. The personalization is much better, even if the intrinsic quality is lower.

AD: I like that framing. Obviously AI is better at the information part, but when I read Joan Didion or David Whyte, I don’t read it for information. I read it for that feeling, and the fact that they did it.

DP: 100%. Well, to your previous point, there’s two kinds of reading. There is, I’m sitting down and I want to extract information on something I’m interested in. There, I want to be in the driver’s seat and AI is great for that. And then there’s another kind of reading where I actually want to go in the back seat. You’re the driver.

AD: Yeah, take me on the ride.

DP: Take me for a ride, baby. And when you sit down and you read a beautiful writer, you’re actually signing up to go for a ride that they’ve designed where you’re kind of deliberately not in charge. You’re saying, I want to be surprised. I want to be delighted. I want to fall in love with your brain. That’s how I feel with David Foster Wallace. I’m not really interested in this, but I just know that this is gonna be an absolute trip that you’re gonna take me on. And those are completely separate styles of writing. And a lot of times when people get bogged down in this debate, they’re missing that distinction. And so people just end up talking past each other.

AD: I mean, I definitely am biased toward the “take me for a ride” one because that’s the stuff I write and I don’t want that to go away. Do you think it will?

DP: I don’t know, I think that AI is gonna get really, really good. I don’t wanna speak in absolutes because culture doesn’t work in absolutes. This is a thought experiment, not a prediction, but you could say, okay, AI is gonna get really good, better than any human writer. And then what happens is a bunch of people will say, we miss the humans. And then what will happen is we’ll come up with ways to basically validate that this is actual human beings.

This is a slight change in topic, but the thing that I just don’t have faith in is the game theory of us actually stewarding AI well. I don’t like the game theory of America and China competing in an AI arms race. For a long time, there was a lot of scholarship on how to steward AI well, but in an international arms race, I don’t think people will prioritize a careful, deliberate rollout over winning.

Second, we’ve already shown how we’ve destroyed the internet. Google links used to be so much better 10-15 years ago, and social media platforms like Facebook and Instagram have gotten worse. There are so few places left that have retained that original purity. The internet is already filled with what people now call “slop,” and I have little faith we’ll treat AI any differently.

AD: That’s one of my main arguments. Facebook was full of slop before AI ever touched it. You don’t need AI for slop.

DP: Exactly. But at the same time, I still believe there’s hope. You can look at it in two ways: either we face an intellectual dark age, or we’ll see a blossoming of how smart the smartest people are. Yesterday I gave a talk to some middle schoolers and I asked them how they use AI. One kid named Diego was like it’s my right hand for building and fixing things with my dad. He was using it to level up his skills. But the average kid there is using it to cheat on their homework. And I think that both of these things can and do coexist.

AD: Yes, they always have. The trap we fall into with AI is thinking it’s fundamentally different from everything that came before it.

DP: Yeah. And at the same time though, I do think that there’s been a serious decline in the quality of things that we build and a kind of disposability to the world. This is what really saddens me and it goes back to beauty and the way that that interweaves with writing and AI. We got so good at making stuff in China that if your shirt doesn’t hold up, great. You can just buy a new one and then you can just go to Zara or Uniqlo and you can get it super cheap. And the problem with AI is that those same patterns of disposability are happening to information and writing.

AD: You interview insane writers all the time. How do you use AI in the process of prepping for those interviews? How big of a role does AI play?

DP: So I have an in-person, human researcher who spends a lot of time giving me a prep sheet. So what’s given to me is a 25 page prep sheet. And then I turn that into a document between three and six pages. And then what I do is I set high level goals for the conversation. So I say, this episode will go well if we do this and not that. And that gives me a sense of the big picture. That’s the first page. And then I just have all these conversation topics, which I break down before the interview starts. So that would be like editing, values, childhood, it’s different for every guest. And then I’ll just mess around. So I’ll put a bunch of PDFs of their writing into ChatGPT and ask how their writing changed over the years. Another thing that I’ll do, which always makes me laugh, is I’ll have ChatGPT rewrite all my notes in the style of Theo Von. I do it for almost every interview. And it comes out with the funniest stuff that always changes how I think about approaching the conversation.

AD: Hahaha! What? I was not expecting you to say Theo Von.

DP: Dude, I love Theo Von. He is my single biggest inspiration for interviews. My friend, Brie Wolfson showed me that in the world of cooking, the most high-end chefs at 11 Madison or French Laundry, after the day is done, they’ll like to have Cheetos or something. Like the bell curve meme of high and low. And I think it’s really nice when you have a craft to just pull inspiration from whatever the thing is that is the absolute most different. I don’t listen to any other interview shows consistently that I would consider intellectual. I basically just watch the Theo Von clips where he’s turned into a baby. It’s my favorite AI art form right now – they just get comedians and turn them into babies. I’ve watched these like 500 to a thousand times and I laugh every time. That’s where I get my inspiration.

When people talk about AI, they’re so serious but a lot of the most fun use cases are making yourself laugh and just messing around. And that’s what’s so fun about AI is you just have this kaleidoscope and every time you turn it gives you different information. Laughter kind of turns off your prefrontal cortex and it just gets you in sort of a more intuitive state.

AD: Are there other examples of how you laugh with it or make it make you laugh?

DP: I just tell it to make me laugh. One thing I’ll do all the time is I’ll take three paragraphs and say rewrite this and make it funny. And that’s when I get annoyed. I want it to be funny in the way that someone from like 1830 would think it’s funny, but also very posh like Victorian aristocrats, you know what I mean? So I’ll say give me that style of humor. And then, I don’t know, it’ll just inspire something.

AD: How does the Theo Von interview that you get back differ from yours and how does it make it better?

DP: He’s really good at little one-liners and analogies. Sometimes I’ll ask AI to lean into the Louisiana kind of vibe. Sometimes I’ll say, okay, but now pretend that Theo Von is from India or from China or something. I don’t know, I’m just making stuff up. And the thing is I don’t have a concrete answer for you, because I’m just messing around. Sometimes it’ll open up a little window of something that I didn’t think about or give me new charm for the episode. It’s different every time.

AD: That’s amazing. That’s one of the best answers I’ve gotten in this whole thing.

DP: Thanks. But hold on. Think about that, simple and weird. That’s kind of how to think about AI. It’s just this box where you could do strange stuff and it can just be simple and weird. That’s how people can be using AI. And I think that as a culture right now, we’re so unimaginative in how we prompt it.

I was talking to a greeter at church and he’s like, “I wanna level up my greeting. I really wanna make people happy. I’m gonna get a little iPad and anyone who wants to talk to me, they can give me a little rundown of their morning and then I’ll have the AI write a biblical haiku that we can read together.” That’s so random and ridiculous, but it’s clever, it’s original. Who knows if he pulled it off, but that’s how we should be thinking about this stuff, you know?

AD: How would you categorize the overall vibe regarding AI from these “god-tier” writers that you interview?

DP: Here’s the thing. You can’t take them at their word for it. What they say and what’s really going on are two different things. I’ll speak to both.

First, let’s consider what they say. You have three different cultures to look at. Traditional publishing, which we can call New York. Internet tech, which we can call San Francisco. And entertainment, which we can call Los Angeles.

These are all very different. New York hasn’t really made up its mind; there’s a wide variance of opinions. San Francisco is pollyannaish about the entire thing. They’re all AI this, AI that. I’m not just writing with AI; I’m using Cursor to figure this stuff out. This is where you find the people who are full accelerationists, who’ve just figured out how to “hack writing.” Some of the stuff I’m seeing there is cool and creative. I think a few of these authors are going to make a billion dollars just from exploiting this AI edge they’ve found. A lot of those people are also just going to produce nothing but slop.

Then you go down to Los Angeles, and it’s a completely different story. “What did you say? AI? We do not talk about AI inside the 30-mile zone.” It’s definitely not a topic for conversation. “How dare you even ask me that question?”

So that’s explicitly what’s going on. But then you have the game theory of what’s happening. The people who don’t use AI get a status boost from talking about how they don’t use it. They’re “above it.”

Meanwhile, the people who do use AI generally lose status by talking about it. What you end up with is a situation where the positive examples get squashed and the negative examples get enhanced. So structurally, there are incentives for the people who are using it well to keep it hush-hush.

AD: I would not have expected LA to be that anti-AI.

DP: In LA, the creatives generally work for someone else. In New York, the creatives often work for themselves. The dynamics are very different. Because they work for a big organization in LA, they’re worried about losing their jobs. In New York, people are more worried about not selling books.

AD: Do you fear it destroying or ruining your own personal writing?

DP: No. I mean, I do feel a little bit dejected sometimes when I look at AI, but I’m working on this piece now, and I think in so many ways I can do better writing. It’s so fun writing with AI. You write with AI. You don’t get AI to write for you.

And I think that a lot of people conflate that. They’re like, so AI just does the writing for you, huh? It’s like, no, of course not. Do that and then you’re just gonna get what everybody else gets, right? That’s pure market beta. It’s like Vanguard funds. But when I’m writing something, I want alpha, I’m like a hedge fund, you know what I mean? I want to be de-correlated from what the AI is giving me.

AD: Do you ever feel bad or guilty like you’re cheating?

DP: I don’t. What I do feel is the cries of shame from society, I feel that voice, but I don’t actually feel it internally.

AD: What advice would you give to people, new graduates, people mid-career, who want to write and are afraid or worried about AI?

DP: Yeah, super simple. This is gonna be terse, because I think it’s just clear. Pay for the best models, use them every day, and hold yourself to a standard where using them makes your work better rather than worse. Don’t be lazy and use these tools to do better work.

AD: Don’t be lazy is good. How would you describe your relationship with AI?

DP: It’s a homie that’s surprisingly smart and surprisingly dumb.

AD: What do you wish it could do but fear it never will?

DP: Disagree with us. This ability to push back is so important and it’s just sycophantic.

AD: Yeah, it’s annoying how much it hypes you up. It’ll say: that’s a great idea, Alex. And I’m like, I know it’s not that good, chill…

Everyone’s making predictions about AI. Where do you think humans will be in 10 years?

DP: I’ll point people to something. There’s this great video where Jacob Collier is playing at the O2 in London and he gets up and he starts playing Fix You by Coldplay. And then Chris Martin from Coldplay comes on stage and everyone freaks out. And then what happens is Chris Martin is kind of just playing live and rolling with Jacob Collier’s crazy creative mind. And Jacob’s up there and he gets all the crowd doing live stuff. And you watch this thing on YouTube, man, and it’s hard to not just start crying.

I can’t imagine being there. It’s live. It’s 20,000 people together through the instrument of the voice. And it’s so beautiful. It’s live. It’s human. It’s present. And that kind of art is just going to crush.

That’s what we want. I’m so bullish on anything that lifts the human spirit.

AD: What question about AI and creativity isn’t being asked enough?

DP: How we’re gonna not suffer the game theory problem of what happened to the internet.

AD: Last question, what do you wish people talked about more instead of AI? I think I know the answer.

DP: Beauty and the decline in the things we build. It’s so sad. It’s the biggest problem that I know of that I can distinctively do something about, and that’s why I’m focused.