49. Jasmine Sun - Close Enough to See Clearly

·Writer, Journalist, Operator
Jasmine Sun - Close Enough to See Clearly

49-jas.jpg

Description

Jasmine Sun (Substack, X, LinkedIn) is an independent writer and journalist. She is a contributing writer for The Atlantic and also writes for other major publications, like The New York Times. She previously led core product at Substack.

Jasmine focuses on Silicon Valley and AI, and is something of a participant observer, living among the strange and inspiring people pulling the future forward in San Francisco. In her writing, she plays to both sides: focusing on a more endemic audience with her newsletter while telling the broader world about what she learns in flagship pieces for major publications. Several of these anchor around memes that she thinks may deeply matter: “the permanent underclass,” “chinese peptides,” and “claude code psychosis,” to name a few.

Jasmine has done many interviews about these individual topics, so I wanted to focus on her and her approach: playing to both audiences, her taste in questions and topics, doing both “serious” journalism and more personal writing, how going independent wasn’t so risky, what she admires in great writing, AI and her coming “AlphaGo moment,” China, and more. Please enjoy.

**Dialectic is presented by **Notion. Notion is an AI-powered connected workspace where teams think together and create their best work. Learn more about Notion’s new developer platform and workers here. Check out Brian’s X/Twitter sync worker. You can learn more at notion.com/dialectic.

Timestamps

  • (0:00) Opening Highlights
  • (1:30) Intro to Jasmine
  • (2:13) Thanks to Notion
  • (3:24) Start: Being a "Historian of Vibe" and Learning to Look
  • (15:00) Taste for Questions & The Depth Behind Memes
  • (24:28) Translating Between Silicon Valley and The World
  • (40:27) Substack vs. "Serious" Journalism and Integrity as a Writer
  • (47:35) Integrity when Using AI and the AlphaGo Question
  • (58:42) Strategy Across Publications & Maximizing an Idea's Reach
  • (1:06:45) Going Independent, Risk, and Commercial Tradeoffs
  • (1:24:35) Great Writing: Style, Voice, and Resisting Summary
  • (1:35:35) Literary Inspirations, Favorite Essays, Writing vs. Thinking, and Getting Better
  • (1:51:09) Writing to Publish, Authenticity, and Art
  • (2:00:38) Grab Bag: China, Silicon Valley's Virtues and Problems, AI Transition, The Relational Economy, Parties, Debates, Self Belief, and More
  • (2:35:16) Thanks Again to Notion

Links & References

Select Articles From Jasmine

Transcript

(0:00) Opening Highlights

(1:30) Intro to Jasmine

(2:13) Thanks to Notion

(3:24) Start: Being a "Historian of Vibe" and Learning to Look

Jackson: (3:24) All right, Jasmine Sun, thank you for joining me.

Jasmine: (3:27) Thanks for having me.

Jackson: (3:28) I'm very excited for this one. I feel like you're also kind of having a little bit of a moment. You've been all over the place.

Jasmine: (3:33) Oh, I don't like that.

Jackson: (3:34) You don't like that? Why?

Jasmine: (3:37) I think it's best not to think about your external perception too much.

Jackson: (3:42) That's fair. But you could still be — you were at Substack, you were interviewing Benjamín Labatut, you're all around town.

Jasmine: (3:50) It's been fun, but I'm mostly very excited to stay put for June. This has been the promise I made to myself — I canceled all of the tentative trips I had for June and I was like, I need to go back in the writing hole and just have big blocks of time on my calendar and be antisocial and just not get on a plane for a month.

Jackson: (4:08) And is that pretty typical of your life? Yo-yoing between what you just described and then a slightly more external way of doing things?

Jasmine: (4:18) I think so. I mean, a lot of writers tend to be more introverted types, and you have these ideas of most writers as just holing up by themselves for extremely long periods of time, never talking to the public, maybe drinking themselves away or going into fits of madness or something like that. I'm not quite like that.

Jackson: (4:36) You have time. You have time to get there.

Jasmine: (4:38) Yeah, give me another year or so. But I think for me, writing is about the process, and I also really do care about the output. I write in order to get particular ideas out into the world. I write because there is some idea or feeling in my head that I want to communicate to many, many more people than I could otherwise. And so the process of doing the distribution after a big story — going to workshops, doing podcasts, just meeting a bunch of people who are interested in the same things I am — does feel like part of the project to me. A whole writing cycle is, let's call it, one or two months of actually doing the research, reporting, and writing — ranging, frankly, between four hours and four months — and then another month afterwards, if the piece goes well, of just talking to lots and lots of people in the field about what I'm seeing, talking about the piece, trying to get the ideas out in as many mediums as I can. And then I start to drive myself crazy. I get sick of the sound of my own voice. That's the phase I'm in right now, and I'm like, okay, it's time to go in the hole.

Jackson: (5:38) And we're podcasting still. We're sick of it. Well, thank you for doing it with me. You described yourself on your Substack as an anthropologist of disruption, which is wonderful. There's an old post — it wasn't your very first post after going full time, but it was the beginning of 2025, so maybe the second or third — and you wrote that writers serve the public as historians of vibe. They tell us what it feels like to be here now. Which I think is a great way to describe, at least from my perspective, what it is that you actually do. And it also fits into the reason that the approach you just described makes sense. So, zooming out — what makes a great historian of vibe?

Jasmine: (6:24) I think I'm really interested in trying to capture the emotional texture or the zeitgeist of a moment. That's a better way of saying vibe, maybe — the emotional texture, the zeitgeist. Though I do like vibe.

Jackson: (6:43) Vibe is a funny word because on one hand it's interesting in that it doesn't mean anything. But on the other hand, it's more like a je ne sais quoi, or like a quality without a name — one of these things that's pointing at something actual or something real.

Jasmine: (6:55) I think it is. A big part of it for me — and the reason I use "anthropologist," the reason I use "historian," the reason I use these particular sets of words — is that I think I'm actually quite naturally a relatively judgmental person. And I don't think the best writers are very judgmental people. One of the things I love most about reading critics — I went through a period where I read a lot of literary, cultural critics — is that they kind of separate writing into two phases. There's a phase of close observation, just purely trying to arrive at the perfect descriptive explanation of a phenomenon. And then there is a separate phase that they can choose to enter if they want, where they make some sort of value or moral judgment on how that thing is. I was actually trying to pull myself away from moralism in a bunch of ways, trying to make myself a less judgmental person. And I find that writing — maybe because it's such a long and deep and reflective process — it forces me into stepping away from my instinctual reactions to phenomena and actually challenges me to just say, how can I be the best historian or archivist or anthropologist of this person, of this moment, of this phenomenon? To just focus on: can I capture it almost as one would capture a photograph, but in text? And to do that and try to master that separately from any attempt to assert some kind of judgment on the thing.

Jackson: (8:27) Do you think all writing is like that? Because that's not intuitive. What you just described totally makes sense, but in my experience, when I go to write, it almost feels like I'm just trying to come up with a take. So maybe my question would be: is that a mode? Is there a specific type of observation that gets you into that kind of capturing?

Jasmine: (8:54) I think it's like a normative belief about how I want to write and how I want to be as a person. There are certainly people who are excellent takes writers — I've definitely written takes before — and there are people whose goals are much more to investigate or to uncover, or even not really to describe but to express something deep within them. For me, it was that I wanted to get away from takes writing. I wrote a lot more takes as a college opinion columnist or something like that, but it felt shallow, and I didn't like the kind of person I became when I was always in takes-generation mode. So it was more that I wanted to be better at observation and noticing and empathy. I noticed that in general, writing caused me to do this, because it's such a long process that gave me a lot of time to reflect on my initial reactions and really interrogate them. The version of myself in long-form writing is a much more measured, much more good-faith, much more even-handed and nuanced version of myself than the version of me in conversation or on Twitter or anything else. So I think I choose long-form writing because it makes me better.

Jackson: (10:01) That's pretty beautiful. It's a little bit of a rehash, but I'd love to read a quote that I think describes part of what you were just saying. You had quoted Susan Sontag saying, "The aim of all commentary on art now should be to make works of art — and by analogy, our own experience — more rather than less real to us. The critic's first duty is to explain, not to evaluate." And then you wrote: "Too much writing on tech still rushes to gawk or condemn. It's why I focused this year on training my journalistic and ethnographic eye instead." You also wrote somewhere else: "The writer's lens follows you away from your desk. Nothing is too small to put under the microscope. Put on your anthropologist hat and you'll instantly see in 4K." Reading that last line, it made me think of that story in Zen and the Art of Motorcycle Maintenance where the teacher has the girl — I don't know if you know this.

Jasmine: (10:57) I've never read the book.

Jackson: (10:57) Okay, so — I actually haven't read it either.

Jasmine: (10:59) I know the story.

Jackson: (11:00) Right, so she's supposed to write about Bozeman, Montana, and she comes back and she's like, "I don't know where to start." He's like, "Okay, write about the downtown." And she's like, "I don't know where to start." "Okay, write about the opera house." And she's like, "I don't know where to start." He's like, "Okay, I want you to go sit across from the opera house and start at the upper left brick." And she comes back with all these words.

Jasmine: (11:19) Oh, I love that.

Jackson: (11:19) So I'd love to hear how you think about starting to observe. You do a ton of anthropology and observation of San Francisco and Silicon Valley and these sometimes narrow and sometimes very vast ideas. What is that process of first starting to look, maybe without judgment? How does that actually happen?

Jasmine: (11:48) I think it starts with my Notes app, first of all. I'm a big Notes app person. I don't believe in having these complicated systems or whatever — as soon as I have a theme or an idea or the seed of an idea, I just start a note. When I travel, I'll start a note that just has the place and the year: "China, 2026," or "New York, D.C., November 2025," something like that. And I literally just start writing bullet points in that note everywhere I go. Every weird little conversation I have, everything I notice just goes into the Notes app. In that moment, I'm not trying to make sense of it. It's very scattered — miscellaneous things like, "Wow, there's a lot of construction everywhere," or "Why are there so many flowers in Shanghai? There didn't used to be this many flowers." Or "Why don't Chinese people take the stairs? They never take the stairs — even if the escalator requires you to wait in a very long line and the stairs are empty and it's going down, everybody takes the escalator." It makes no sense.

Jackson: (12:56) That's like California, or at least LA.

Jasmine: (12:57) Really frivolous things, just into a giant bullet point list in my Notes app. That's all the observation phase — collecting mode, collecting mode, collecting mode. And then once I either feel like I have enough material, or if it's travel, once I've returned from my trip, that's when I really revisit the notes, read them over and over again, and start to notice what themes emerge. So again, even in the actual work, the process of observation and the process of synthesis and judgment feel like two separate steps to me. And it's important to me personally to keep them separate. For reporting it's a little bit different. When I'm doing capital-J journalism and actually reporting out a piece rather than just traveloguing, I almost always start with a research question that I do not know the answer to. One thing that bugs me in some journalism is when I'll read an article and it's really clear that the reporter already had a thesis and they're just trying to prove out that thesis with a set of quotes. I like to pitch stories where I have a question — like "Why is everyone taking Chinese peptides?" or "Does AI actually have any impact on the labor market?" — that I genuinely do not know the answer to, that I am distinctly not an expert in, that I know there's discourse around but I really want to go broad during the interview phase. I talk to people intentionally on every single side of the issue, not trying to extract judgment. And often it's only after doing 10 or 20 or even 30 interviews that I even start to have an opinion on what I think. This has actually made it sometimes hard to pitch stories to editors, because some editors really want you to have the angle before you write — otherwise it's going to be boring. And I'm like, well, if I had the angle, I wouldn't be writing it. I want to write this story because I want to embark on this process of exploration. I feel pretty sure that if I talk to 25 people about one of these things, there's going to be something in there. But I can't promise you in advance, before I go and talk to those —

(15:00) Taste for Questions & The Depth Behind Memes

Jackson: (15:00) 25 people — that presumes you are asking good questions. Your confidence, I mean, it's a mix of a few things. It's confidence in the rigor you're going to approach something with, but perhaps also that you don't have the answer. The fact that you don't have the answer now, or very quickly, is probably indicative of the fact that it's an actually interesting thing. It's not a yes or no question. And I suspect that you have some kind of tacit taste for the types of questions to ask that are going to lead to a lot of string to pull on.

Jasmine: (15:36) I hope so.

Jackson: (15:38) Maybe to ask it a different way — do you ever start down this process and realize, whoa, there's actually just not that much here and it's much simpler than I thought?

Jasmine: (15:48) No, but I feel like my internal narrative for the same thing is that everything is interesting. I'm thinking of that Annie Dillard quote, which I think I cited in one of those pieces — it's like, "admire the world for never ending on you." I really believe if you just look closer and closer and closer, it will become interesting to you. I wrote this New York Times story on the peptide phenomenon, and before I wrote it, I was not interested in peptides. It kind of happened accidentally. I report on AI mostly; I don't really follow health or bio or pharma or any of that. But after like a month of reporting on peptides, I was fascinated by the whole thing. So one thing is, I do think that everything is interesting if you look closely enough. And also if you see things as instantiations of trends bigger than themselves. The peptides thing isn't just about peptides — it's about this weird community epistemology around biohacking and people who trust ChatGPT and Reddit forums more than they trust their doctors. It touches on all of these medical trust issues. So it's about something bigger than peptides, but peptides is a very fun and quirky entry point into the broader discourse. When I started investigating it, I had no idea. I just thought it was funny that people were buying vials of gray market drugs from China and injecting themselves. I just think that the world is much more interesting than we all think it is.

Jackson: (17:11) Yes. And we were talking about this once and you said you're trying to write things that are relevant on a two-to-three month time horizon — not two weeks or two days, but also not much longer. So to take the peptide example, the permanent underclass is very much in the middle of your strike zone in all of these different ways. What caused the peptide story? Is it just like, I feel the temperature rising around me, I've heard this enough times — what made you do the month of work that got you interested in it, as someone who wasn't interested initially?

Jasmine: (17:53) I mean, the real answer is not very interesting, which is that I was trying to pitch Tracy Alloway from Odd Lots on doing a Chinese peptides episode — not me doing it, I was trying to pitch her on doing one with somebody.

Jackson: (18:05) Somebody's got to. Will the devs do something?

Jasmine: (18:07) Yes, will the devs do something? I was like, Tracy, this is hilarious. It's about supply chains, it's about US-China stuff, it's about biohacking, it's about tech bros. It is a perfect story. I was trying to convince her to do it, and Tracy was like, you're so right — who should I interview? And I was like, I don't know, but you gotta find someone. And then Tracy just kept being like, do you want to do it? And I was like, no, I don't know anything. But eventually I was like, you know what — I sort of convinced myself in the process that it actually was an interesting story.

Jackson: (18:35) Because you were telling her, yeah.

Jasmine: (18:37) Yeah. Okay, maybe I guess I will do it, if I really want somebody to cover this topic, because I knew it had to be covered. And then I ended up chatting with my Times editor and it all worked out. I think, insofar as there is some taste for asking questions, one is that I do gravitate towards things that feel like they sit at the intersection of a lot of issues. I was excited that it touched on supply chains, US-China, pharma — I could see that from the beginning. I like reporting on memes, but not in the stupid way — in the deep way. I spend a lot of time thinking about the difference between trend reporting versus culture writing. Trend reporting is like, I saw three TikToks, let me just tell you what it is. Culture writing is interrogating: okay, the fact that this meme has taken off signifies something much deeper going on in the culture, and I actually want to investigate what deeper cultural currents are causing some concepts or trends to stick above others. So with Chinese peptides or the permanent underclass or whatever — the fact that a meme is rising in the zeitgeist tells me there's something simmering below, and I want to find out what that is. It also conveniently means, if I'm being a bit more tactical about it, that when you write the story, it's going to do quite well. Which I have now also learned.

Jackson: (20:02) You had a great conversation about that earlier point — the trend thing — you talked to Celine Wynne about this.

Jasmine: (20:08) Oh, yes.

Jackson: (20:09) And I think there are a lot of people saying, I've noticed three TikToks, I have a take. It goes back to what you were saying at the very beginning — you're almost saying the fact that something is bubbling up so much means there's probably a more profound there there, whether it be on cloud psychosis or any of these things. It's fascinating because if you really squinted, you could describe what you do as sort of being like a San Francisco meme reporter.

Jasmine: (20:36) Sure. I mean, I would prefer not to, but sure, whatever.

Jackson: (20:39) To your point, I think memes are actually these black holes of interest.

Jasmine: (20:46) Yeah.

Jackson: (20:47) Or at least they can be. Obviously they can be really flat, but when they aren't, there's something there — and for better or for worse, something like the permanent underclass gives us something to hold on to.

Jasmine: (20:57) Yeah. And I think sometimes the memes and the jokes are obviously hyperbolic, and I try to be real about the fact that they're generally hyperboles — but they are also release valves for emotions, for the things you can't say. One thing a lot of China watchers do is spend a lot of time on the Chinese internet. There are all these great Substacks like Chinese Doom Scroll that collate Chinese internet memes and translate viral phrases. And right now, if you look at the Chinese internet, so many of the phrases are about the grind of work. There's a term that's like "work smell" — basically, if you've been working too much, you have work odor. You have 996, of course, which was originally a Chinese phenomenon. You have all of these terms that refer to — "code farmers" is another one — just the grind of working. And in a culture where there's a lot of speech censorship and not a lot of public action or policy advocacy, one of the ways you can understand how a lot of Chinese young people are really feeling is by looking at the internet memes. It's kind of an acceptable and non-threatening funnel for very real concerns about unemployment and affordability and so on. And similarly, yeah, sure, the permanent underclass is a meme. Will we actually be in this permanent useless class of people stuck on welfare checks forever with our robot overlords? I don't know — probably not. At the same time, part of what made me report on it was that I have friends in San Francisco who are literally making their career decisions based on these memes. They feel like if they do not get a job at a frontier lab, they will be poor forever. That's obviously, in my opinion, not actually true. But the fact that it's actually guiding action, and the fact that it expresses anxieties even among the people working on AI — who are not that techno-optimistic — was one of the things I realized. Like, you guys don't think that AI is going to create equal abundance for everyone or make everyone live like kings? Actually, there's an assumption baked into the permanent underclass meme that AI will widen inequality, that there will be winners and losers, automators and the automated. And if you look closely enough at the meme, you uncover the ideology behind it. And I actually think that's really important.

Jackson: (23:16) It also ties to your — you have some comment in, I think the main piece, where you say when you talk to these people off the mic, they all stop being optimists. No, they don't all stop, but there's a lot of that.

Jasmine: (23:25) Yeah, it was very frustrating. I literally talked to somebody as a friend and then I'd get them on the mic for an on-the-record interview and they would just totally change their tune. And I get it. I don't want anyone to self-incriminate. I'm not doing journalism in order to screw people's reputations over. And so I take what they say on the mic. But I added that parenthetical to the New York Times piece — I think the day or two before it went out. It was one of the very last edits that I made. I made it on my phone. I remember where I was: lying awake at night in my bed, feeling unsettled about the piece. And I was like, you know what's really frustrating is I've done this work to try to take these conversations happening in whisper networks and private conversations and make them public. But even the things that I'm quoting publicly are in some ways less extreme than what I hear privately. How do I communicate that? And then I was like, you know what, I'll just go for the meta. I'll literally add a parenthetical that's like, just so you guys know, people are more pessimistic when they are in private. And I could not get the same people to say the things on the record.

(24:28) Translating Between Silicon Valley and The World

Jackson: (24:28) Your main beat, to be overly generalized, would be San Francisco and Silicon Valley culture. There's this fascinating thing where on one hand you've called San Francisco the most happening place in the —

Jasmine: (24:44) World — notably not a word, for the record.

Jackson: (24:47) One of, if not the most important places in the world. You've compared places in China in a similar way. On one hand you have traditional media, which doesn't really know how — they're both mad at San Francisco and don't totally know how to actually talk about it. And then on the flip side, San Francisco doesn't really know how to talk about itself. And so in some sense you've become this sort of translator across these two worlds. On one hand, you're doing a little bit of the Didion thing, reporting from the inside. In another way, I think a lot of people from the outside looking in might be like, oh, she's one of those people. You're friends with the people who work at all these labs, and yet you're not among them. I'm curious what you've noticed about that divide, especially as you try to bridge it. When you're writing on Substack, you're probably writing more for San Francisco talking to itself. When you're writing for the New York Times or the Atlantic, you're reporting from within to the outside. What does that tension feel like? How does it come up when you talk to other journalists who are maybe less on the inside? It feels like you're in a pretty load-bearing position.

Jasmine: (25:55) Yeah, it's interesting. I think a lot of my role in general has been doing various forms of translation between the tech industry and the external world. I mean, I used to work at Substack as a product manager, but I also helped out a lot with comms and product marketing. I spent a lot of time writing blog posts that explained why we would attach a social network to the precious newsletter ecosystem, because most writers are extremely suspicious of tech companies and social media and algorithms and AI. Part of my role was to make the case for why it actually might be values-aligned and good to do that. I feel like, because I personally feel very half in, half out with the tech industry — I worked in it, I do have a lot of friends in it, there are a lot of things I really love about Silicon Valley — I have this simultaneous suspicion that comes from the parts of myself that are maybe more engaged with journalism and politics and whatever. It creates a tension within myself, and then it also creates this desire for, like, why won't my friends get along? I want them to like each other more.

Jackson: (27:16) Are both sides skeptical of you, or have you found — from the outside looking in, it seems like you've done a pretty good job of having both sides feel like you're not too swung to the other.

Jasmine: (27:31) I mean, I try my best. I think part of it is how do you maintain genuine independence? How do you make sure that you don't seem like a shill for whether it's effective altruism or the tech industry or the progressive left? There are all of these tribes and all of these camps. And one thing that frustrates me about a lot of the Silicon Valley new media world, or even just Substack stuff, is that they're not really doing journalism or truth-seeking. They're doing advocacy for their camp. This is a progress studies Substack. This is a rationalist blogger. This is a new media thing celebrating the spirit of enterprise, or Andreessen Horowitz, or whatever.

Jackson: (28:11) A lot of that, at least, is Silicon Valley talking to itself.

Jasmine: (28:15) Yes.

Jackson: (28:16) Is it worth delineating? There's Silicon Valley talking to itself, and then there's Silicon Valley trying to do propaganda, maybe.

Jasmine: (28:23) Sure, they don't do a lot of it.

Jackson: (28:25) Okay. So you're more talking about when it's Silicon Valley talking to itself.

Jasmine: (28:30) Yeah. But what was the original question?

Jackson: (28:36) The original question was pretty open-ended, which is: what does it feel like to be this load-bearing translator between these two worlds? And maybe inside of that question, to be honest, is — there aren't a lot of you.

Jasmine: (28:49) Yeah.

Jackson: (28:50) One hypothesis I had — I think I was even talking to Celine about this in terms of what I should ask you — is that maybe you have to kind of appreciate something to be able to really see it clearly.

Jasmine: (29:03) Absolutely, yeah.

Jackson: (29:04) And so maybe that's one of the reasons you're able to occupy this space well. If you loved it more, you'd probably be working at Anthropic. And if you loved it less, as most journalists do — which is a super overdue generalization — they just stay away. So it's somewhere in between.

Jasmine: (29:21) Yeah, I think that's a really good point. Celine should really do the talking for me. I think when you love something, you want to understand it and you want it to be understood. Part of this is what I said about trying to observe without judgment a lot of the time. When I think about my reporting and I interview someone, even if internally I might have very different beliefs than that person, to me the mission of journalism — or the kind of journalism I do, at least — is to capture their belief. The thing I am reporting on is the essence of what they believe. And even if I can only get one or two sentences into my final piece because you just have to cut a lot, I never want them to look at that and feel like she was cherry-picking. I want them to look at it and think, that was the essence of the thing I said. So even in some pieces I've written that have been really controversial — there's one from about a year ago on the defense tech revival at Stanford, basically asking why all these students want to work at Anduril and Palantir now. I interviewed a lot of people at defense tech companies about why they did it, and I wrote them up in a piece. The reaction was really interesting, because the people I interviewed, when they read the piece before any public reaction, they liked it. They were like, this is an accurate representation of my views. You asked me why I want to work at these places and I said it's because I care about Ukraine, or I think more women should work in national security, or whatever. One of the lines that went viral was "all of my most moral and effective friends work at Palantir" — because this was a genuine belief that this person held. So when she read my story, I wasn't cherry-picking. My goal was in fact to represent exactly what the person said in its fullest essence. The public reaction is a different question. The public reaction to that piece was: these kids are murderers, and thus you tricked me.

Jackson: (31:10) The feeling of the person who maybe gave the quote might end up being "you tricked me," even though it wasn't decided.

Jasmine: (31:15) It wasn't. I mean, it was a psychologically challenging thing for me because I wasn't trying to trick them, I wasn't trying to throw them to the wolves. And I also quoted quite a wide range of things. But it was really important to me that at least — public reaction aside — they would not feel tricked.

Jackson: (31:30) Yes.

Jasmine: (31:31) If they only feel tricked after the public reaction, that's very different. But I think that both when I talk to people in Silicon Valley or I talk to people in politics or whatever, I want them to feel like I have captured them, I have created a portrait, I've taken a photograph that feels correct and true, regardless of what the reaction is after. And if I hold to that, I think people will give me a little more credit. And then of course, there are a lot of things I really like about Silicon Valley, and I think people can see that. I also have a lot of skepticism — I think people can see that too. And I'm pretty open about my personal beliefs. I write opinions on a blog. So before people chat with me, I feel like they can get a pretty good sense of who I am and how fair I am.

Jackson: (32:26) Why do you think Silicon Valley struggles so much to talk to the rest of the world about itself?

Jasmine: (32:32) Two things, I think. One is they don't try, and two is there's a genuine self-selection for a different set of values here. They try less now than they used to. They used to try harder in the "tech for good," Google-whatever era. And I think if you look at, for example, "going direct" — Lulu's thing — I respect Lulu a lot, Lulu's very sharp. Going direct worked for particular goals, which was that Lulu was saying you should identify your audience. Oftentimes your audience is recruits — if you're an AI company, you want to recruit the best talent, or you want to land very specific enterprise deals. So rather than cultivating your messaging to the broad public, direct your marketing toward a very specific group of people, and that'll be much more effective. I think she was right about that. But sometimes the things you say to convince a bunch of very odd AI researchers to join your lab — when the rest of the world hears that, they're like, "dude, what the fuck?" So there's an extent to which the insularity, and the fact that Silicon Valley spends most of its time messaging to itself, to investors, to potential recruits, to enterprise tech companies that might make a purchase — it means they don't know how they sound to everyone else, because they're not trying to message to the public. They feel like their business is not contingent on whether the public likes them. It's not like Coca-Cola. Coca-Cola needs public goodwill, but a lot of these companies don't, especially at the beginning. Before OpenAI had a consumer product — for the first five years or whatever — they didn't need to talk to anyone else. Sam Altman could write these blog posts saying AI might make the world worse off and shift all the power from labor to capital, and that honesty might be really respected by these very AGI-pilled researchers. But by the point where you're trying to IPO and you need investors and the public to like you, it becomes a lot more troubling. And I think that's one of the tensions they're coming up against. The second thing is that the values of Silicon Valley are genuinely very different from the values of the broad American public. Of course, everyone wants welfare and happiness and whatever, but San Francisco is an industry town, and what that means is that people come here oftentimes because they want to be part of a specific culture and subgroup. You have these nerds who felt really misunderstood in the places they grew up, in their other contexts — maybe they didn't fit into their university or institution — and they thought, "here's a group of people who will appreciate me, who will appreciate how good I am at coding, how high-agency I am, my idealism." Where other places might have a tall poppy syndrome and try to cut them down. So you really concentrate a group of people who have quite similar personality characteristics and drives, and it's not actually a very good representation of the overall population. I'll notice this in things like when people describe their post-AGI utopias — they'll talk about wanting to live forever, or wanting UBI so they can engage in leisure all day. There are public polls of Americans on living forever and UBI, and none of those things are very popular. Or it's like, "oh, you can have a superhuman tutor in your pocket, the equivalent of Terence Tao tutoring you in math." And I'm like, how many Americans really want Terence Tao in their pocket if you ask them what utopia looks like? Some of them, but actually not that many. A lot of people want jobs. People like feeling like they have a purpose. So there's also a genuine way in which even when Silicon Valley attempts to message about these positive narratives around AI, they don't realize what really matters to people — that maybe living forever and not having a job is not everyone's idea of the good life.

Jackson: (36:17) You could paint a similar version of that for coastal elites generally. Why do you think this is so much more true in San Francisco than New York or LA or London or other hubs where ambitious people from all over the world flock to?

Jasmine: (36:37) I mean, New York and LA and DC all suck in their own ways, right? The out-of-touch politicians, the Brooklyn media class, all Hollywood movies are secretly about Hollywood. I think it's definitely true for every industry. The other thing maybe is that storytelling and messaging is generally undervalued in Silicon Valley. If you just think of some of the traits that are valued in people — being a nerd is often associated with being kind of prickly and bad at communication and hard to work with, but you're a genius so it's okay. Whereas in New York, LA, and DC, I think being broadly likable is seen as a more appealing quality. I actually think sometimes you lose points for it in Silicon Valley.

Jackson: (37:27) This is changing, by the way.

Jasmine: (37:28) It is changing, especially —

Jackson: (37:29) The funny thing is you were making the point about how these companies never had to talk to the rest of the world, but if anything, it feels like Silicon Valley is doubling down on brand more than ever, on charisma more than ever — all this new media.

Jasmine: (37:43) Oh totally, yeah. It's great for us, right? But it's so funny — when my interest in Silicon Valley started, roughly 2017, I went to Stanford, and at the time being interested in writing and sociology and stuff like this was very low status. No one gave a fuck about this. Everyone was like, "You should just code. It's really stupid that you can't code."

Jackson: (38:03) Well, now, losers.

Jasmine: (38:05) I know — who's getting automated now? But these things change. And because of this, the media ecosystem here is really new. It's really young. It's still figuring itself out. I think new media can figure itself out, but it's going to take some years. They're coming up against these hundred-year-old institutions on the East Coast. And another thing I think about all the time: there are very few journalists who live in San Francisco. Just by numbers. There are so few journalists who live in San Francisco relative to how important it is as a center of power and wealth. So many people report on San Francisco from New York and D.C. And to me, for example, the idea of reporting on the White House from San Francisco would be insane. I wouldn't trust a White House reporter who never went to the happy hours with all of the staffers. In D.C. you have this much more integrated — and in a way mutually beneficial — ecosystem between sources and reporters, where they're going to the same happy hours, hanging out in the same social scenes. So one, it means reporters have better story taste, because they start to hear about things early. They're actually in the scenes where they care about things.

Jackson: (39:16) It's intuition almost. It's like the temperature got hotter, or whatever.

Jasmine: (39:20) Yeah, exactly. I feel like that's what I do here — I can feel when the temperature goes up on something in a way that maybe a reporter in New York can't. But it also means you have more of a pact, more trust. Politicos understand that there are ways to work with reporters that are advantageous to them. Like, "I can feed you my side of a story on background, and that might actually be better for the way you portray this in the public eye." And also because there's some shared social tissue, it means you're not going to burn me in a deeply unfair way most of the time, because it's going to screw everything else up. Whereas here, most people in technology know zero journalists personally, and there is just a default distrust. And I think that distrust is even partly warranted, because the journalists aren't in those rooms. They don't really know how people are thinking. There isn't really an advantage to telling somebody who has no context your side of the story, only to be taken out of that context and painted as a villain. And I think that's what happens with a lot of tech journalism.

(40:27) Substack vs. "Serious" Journalism and Integrity as a Writer

Jackson: (40:28) Speaking of different audiences — you have these amazing flagship pieces in big publications that are certainly speaking more to the rest of the world about what's going on in Silicon Valley from the inside. And then you have your Substack, which is obviously much more inside baseball, not necessarily speaking directly to Silicon Valley, although I would guess it leans that way. Is all of your writing journalism?

Jasmine: (40:55) I don't think so. I think a lot of the Substack is not journalism.

Jackson: (40:58) Well, what is it? How do you think about what it is?

Jasmine: (41:07) Writing. It's all writing. I like the word "writer" because I like how expansive and big-tent it is. And I like that it's also kind of understated. "Journalist" comes with baggage. "Very serious blogger" is almost too diminutive — it's not serious enough. "Writer," to me — I love how big-tent it is and how it contains so many things within it. Writing has research, writing has craft. There are so many things. With Substack, I don't really think about what it is. I don't think about myself as writing in a form. It's really loose. Some of them are travelogues, some of them are basically columns or op-eds, some of them are interviews that I transcribe. I'm really not thinking very hard about the form.

Jackson: (41:53) It's very you, though. It's not always parasocial or that kind of thing, but it's very much your lens on the world, which some of your other writing is not. With those there's a glass pane.

Jasmine: (42:04) Oh, totally. The Substack is just like — I want to take my brain and transport it onto the page. I feel like there's a bunch of stuff going on in my brain and your brain and everyone's brain. And the thing I really love about writing on the internet publicly is that I can go through the exercise of seeing how much I can turn that weird blob of ideas into a form where it's shareable and scalable. I can take all the fuzz of ideas and concretize them.

Jackson: (42:32) Well, it's almost a version of your Notes app thing. It's like one long funnel of the little things you've noticed in the world as you're driving around — which most of us do, but we don't write down.

Jasmine: (42:44) Yeah.

Jackson: (42:44) And so you've built that, like, it —

Jasmine: (42:47) It goes through a filter for digestibility before it makes it onto the Substack.

Jackson: (42:49) Right.

Jasmine: (42:50) The Notes app is a real disaster. I would never share that. But I do sort of try to make it a bit digestible and then share it. Because even before I had very many subscribers, writing long-form on the internet is one of the best ways to meet like-minded people. I've just made a lot of friends by doing this. So I want to share how I think, because I'll find people that way. With reporting and journalism, it's a much more specific thing. I actually do believe in a lot of the ideals of journalism — these attempts at objectivity, talking to sources from all sides, going through this very careful reporting process: on the record, off the record, on background. You're trying to engage in a very specific practice with a set of professional norms around it. And I take those norms pretty seriously, because I do think that for all the flaws that capital-J journalism sometimes has, when something is printed in the New York Times, you assign it a level of credibility that me posting a meme on Substack does not have. So I'm willing to play along with the process. And I also think it's good for me to go through it in order to produce these bigger pieces, because it gives the ideas more weight and more credibility when people know you've done all of those steps to get there. It's also largely split based on audience. When I write in a bigger publication, it's oftentimes something I want to share to a broader group — primarily people who are not in Silicon Valley. And I also like that it gets me to do more reporting. I don't report as much for my Substack, whereas I'll do so many more interviews and go so much deeper and more rigorous when I know it's going to be printed in an outlet where I can't just make a little edit the day after because I changed my mind.

Jackson: (44:35) Some of my favorite pieces of yours are the companion pieces to the big publication on your Substack. Like the permanent underclass one — the bonus episode. That was juicy.

Jasmine: (44:49) Yeah.

Jackson: (44:52) Maybe on a similar note — and I'd actually be curious across both of those modes — what does it mean to write with integrity? There's journalistic integrity and all those things you were describing, but what does it mean to write with integrity even in the more casual mode? My sense is that it's something you think a lot about.

Jasmine: (45:22) Yeah, I mean, I try to write things that are true. I try not to let my prior ideological affiliations get in the way of truth-seeking and just be really honest with myself about that. I try to talk to people, especially when I think I'm going to disagree with them, and force myself to listen to them — which is healthy, because oftentimes you do come around. I think maybe another thing is there are probably debates in journalism about how much you should feel responsible for the outcomes of your piece and what happens once it's in the world. Some writers and journalists are very much like, "I owe nothing to the world. I'm here to create this thing, but I don't want to be audience-captured. I don't want to be swayed by the court of public opinion. So I write for an audience of one, which is myself, and the world can do what it will." For me, part of integrity is actually caring a bit more about what happens afterwards. It just really doesn't feel good to release something that will throw your sources to the wolves, or that is too casual and too loose with other people's lives. Journalists command an incredible amount of power. Writers command an incredible amount of power. It's not wealth in the way that working in an AI lab gives you wealth, but being able to have a microphone to shout to the tens of millions of subscribers of the New York Times — "hey, this is what I think about this issue, and this is what all these people are saying" — that is a huge amount of power. There are people whose entire jobs it is to get a single mention in one of these papers. So to me, holding that power with integrity means both the truth side of it, but also being at least aware and responsible about some of the impacts you expect. People debate about this, though.

(47:35) Integrity when Using AI and the AlphaGo Question

Jackson: (47:33) Well, thanks for pontificating on it. You also talked about integrity with AI recently on Substack. You said, "Being real with yourself is the most important cognitive skill for the AI age. AI makes it easier to lie to yourself. You've got to be able to honestly answer: am I actually thinking with AI, or am I letting it do the hard part for me? Is this essay, product, or business a good idea, or did AI convince me it was?" Somewhere else you said — I think you were quote-tweeting something — that AI writing is bad for the writer, much more so even than for the reader. I read this as almost a different kind of integrity. It's like an internal integrity around whether or not you're going to cheat yourself.

Jasmine: (48:17) Yeah, yeah.

Jackson: (48:18) How do you keep yourself tethered to the right balance there? Even something as simple as — you were talking about using Claude as an editor, and you're like, "Hey Claude, I don't want you to change the writing at all." That's a great aspiration, but it seems really hard when you're in the weeds and you're thinking, "Wow, that idea from Claude is pretty good. Maybe I should just use that sentence." What are you holding onto?

Jasmine: (48:47) I really don't want to make myself dumber. This is a big fear I have in general — Claude or no Claude, AI or no AI. I write because I think it makes me smarter, I think it makes me kinder. It makes me a bunch of things I want to be. I write full-time because it helps me live the life I want to live. All of the reporting, for instance — I became a much better listener, I became much more empathetic. Most people don't spend hours of their lives listening to people who have very different beliefs and focusing solely on the understanding aspect. So for me, writing is almost a selfish thing — it's an exercise in personal development. And I don't want to do anything that will hinder those goals. I wrote that Substack note about AI because I was trying to think through all of these arguments about whether you can use AI in writing. In a way, I'm a moderate. I use AI all the time as part of my research process. I'll ask Claude for feedback on my essay structures. I use AI pretty regularly. But I really don't like slop — I really don't want AI to ever generate sentences or paragraphs for me, and I'm careful about that. And oftentimes I'll get flack from both sides: people who are like, "Why are you using AI at all? Do you hate thinking? Do you hate your human editors?" And also people who are like, "Why are you discriminating against people who can't express themselves in sentences?"

Jackson: (50:21) Or maybe even something like — is there a chance that using it more would make your writing better? I mean, it's sort of a rhetorical question.

Jasmine: (50:31) I mean, with regards to the people who really don't like any AI use at all, I realized why I don't fear being de-skilled: it's because the way I use AI is so effortful. And I just know what it feels like to think — I know what it feels like to be getting better and getting worse. It's not like I never use AI to summarize a book for me. Of course I do that sometimes. But I don't lie to myself about the difference between reading a book and reading a summary of a book. That was true before AI too. I know the difference between SparkNotes-ing a book and actually reading the book. It's kind of like if you plagiarize something — it's not that the output is bad, it's just that you should know that you plagiarized. You should know you were cheating yourself. If you're token-maxing, right — there are all these coders and companies spending millions of dollars on tokens, and some of those tokens are probably valuable, some of them are making people more productive. But a lot of it is the illusion of productivity. You might be lying to yourself, thinking that the lines of code you write or the tokens you use are representative of an increased level of productivity. For some people, it is. Other people are literally just lying to themselves. And this is the thing I kept thinking about: AI makes it so easy to generate the illusion of thinking, the illusion of productivity, the illusion of everything being a good idea — because it's so perfect at justifications. If it's that powerful, the only thing you have is knowing, in some deep internal sense, whether you are doing real productive work, whether you are learning, whether you are thinking. In the same way that if you are lifting heavy, you know you're pushing yourself at the gym — it's the same feeling. You can lie to yourself if you want, but you're not going to be stronger. I really don't want to be dumber. So I try not to lie to myself.

Jackson: (52:24) The fear is that the more we use these tools, the more it will become harder to know when you're lying to yourself. Unless you're really vigilant about it. Clearly you have a very firm boundary on that. My sense is, personally, if I'm not careful about it, the boundary is getting blurrier.

Jasmine: (52:49) I mean, maybe I'm not careful enough. Like, maybe you're right, and I actually am just getting dumber because of it.

Jackson: (52:55) No, that's not what I'm saying. What I'm saying is I think you're actually very attuned to this and you're thinking about it a lot. At least as you've just stated, writing is a huge part of making yourself smarter.

Jasmine: (53:11) Yeah.

Jackson: (53:12) And maybe that's completely true. Maybe that's a story you're telling yourself. Whatever. But I think for so many of us, as more and more of what we're doing — this is probably too much of a sidebar, but like, an example would be the prep I do for an interview. I don't use very much AI currently, and I think I could probably use more AI and do better interviews. But there's also a version of that where I take it too far. It's currently important that I do a good amount of it manually. And so I think that's what I'm trying to point out: as these tools get better and better, part of it is I get to rely on the fact that if I do it manually, the interview's better. But what if at some point Claude's research is better than mine?

Jasmine: (54:09) Yeah. I mean, I use the tools because I want to know. Every one of my personal evals of the AI systems is roughly once a quarter, or as often as I do a freelance piece. I take my pitch and I take all of my interview transcripts and I feed them into Claude, the latest version of Gemini, and the latest version of GPT, and I say, "Write this 3,000-word New York Times article for me. Go do it." And to be clear, I never use any of the content — it's just an eval. So, one, I do want to know if it gets to that point. I don't think it's there yet. I feel like I'm still better at writing than the AI. But I don't fear finding out, and I think that's kind of important.

Jackson: (54:48) You don't fear finding out?

Jasmine: (54:49) No, I really don't.

Jackson: (54:51) Which is like — you could know that it's going to write a better one than you...

Jasmine: (54:54) Yeah.

Jackson: (54:54) And you still feel confident that you're going to write it yourself.

Jasmine: (54:57) Yeah, because I want to be smarter. It's very — I just can't be fully instrumental about it in that sense. It just can't be about the output alone. I actually believe that writing for me is a process of personal self-development. And so that's why, like with cheating in school — you can cheat on your homework, maybe you guarantee that your teacher's never going to find out, but you're not going to be able to do stuff later. I just really don't want to be that person. I use AI plenty, but for me, I know that it actually makes me less lazy. When I didn't have AI tools and somebody said a word I didn't really know, or I was interviewing someone and they were explaining some complicated economic concept or referring to some paper — a lot of the time I probably would have just let it bounce off me and moved on, because I didn't have the time to go find out. I didn't have the time to read a really complex paper or learn a totally new discipline. But now, by lowering the bar to learning a little bit about something, I'm way more likely to look stuff up. Like, "Can you summarize this paper for me?" And I know that the net amount of learning I'm doing is more than it was before. I'm still very aware that "summarize this paper" is not the same as actually reading the paper. But it feels pretty clear to me personally — and this is what I was struggling so hard to articulate in that note — I don't know how to give any advice besides: be real with yourself. You just have to know if you're doing something that works. You just have to know.

Jackson: (56:44) I think the challenge is that everyone is going to have to calibrate themselves on what that actually means. And part of it is that most people are mostly doing things instrumentally and haven't actually stopped to say, "What is the thing I want to do, even if the computer can do it better than me?"

Jasmine: (57:03) Yeah. I think that's why I always think back to AlphaGo. That story has so much resonance for so many people over so many years. We're at like the tenth anniversary of the Lee Sedol match, and people still talk about that match more than they talk about a lot of other milestones in AI that have happened since. Because I think this very specific experience — we are going to watch the machines excel at the thing that we do, we are going to watch them beat us, and we are still going to have to ask ourselves: do we want to play anyway? One of the things I find really beautiful about games — and I know you interviewed C. Thi Nguyen, and I love his Games: Agency as Art — is that you are playing them to exercise your own agency, to develop a set of skills. You are not playing a board game because you're going to win money from it or because it's going to accelerate your career. And everything is going to become a little bit more like playing board games soon, and you are just going to have to choose: do I want to do this for its own sake?

Jackson: (58:06) He makes this great point where he's talking about playing a board game with his wife and he's like, on one hand we both have to try to win, we can't not care. But if I go spend all this time separately studying strategy guides and get so much better than her, it's not going to be fun anymore either.

Jasmine: (58:23) I love his work so much. It's more than board games, especially right now. There's so much to it. I've thought a bit about what are the books I would recommend to people for the age of AI that are not actually about AI — and Games: Agency as Art is one of them, because I think that's the disposition we're going to need.

(58:42) Strategy Across Publications & Maximizing an Idea's Reach

Jackson: (58:42) I think you're right. A couple final things on the Substack versus maybe the serious writing stuff. You wrote in your end-of-year recap, kind of at the start of this year, that you were going to try to dial up — I think it was like 40% Substack last year and like 80% this year.

Jasmine: (58:59) Yeah, but I didn't do it.

Jackson: (59:00) Didn't do it. Why? And what are you thinking about for the second half of the year?

Jasmine: (59:04) I mean, I used to —

Jackson: (59:05) Oh.

Jasmine: (59:06) Oh God. Yeah. It's really interesting, because I worked at Substack for a long time, and I was more brought up as a blogger on the internet than anything else. I've never worked in a newsroom. I've never been really trained in journalism in that way. My default loyalty is to independent media and the independent ecosystem. At the same time, I think partly with AI in particular, I've come to believe that legacy media and telling good stories about AI in legacy media is a lot more important. And so I think it actually comes from a sense of duty more than anything else — it's less about me and my preference. It's more a sense of public impact, where I think there are not that many AI journalists, not that many AI journalists in San Francisco. I do feel like there are certain stories I can tell that other journalists aren't equipped to, and I should exercise that by trying to portray certain ideas in the biggest medium possible, where they can reach the most number of people. And that's not my Substack. And so I've become a lot more open to doing more freelance work. There's just nothing that compares to, say, the reach that the New York Times has. It's incomparable — just the amount of attention, discourse, and action that comes out of writing a big New York Times piece. And it feels really important right now, given how polarized Silicon Valley is from the rest of the world, given the AI backlash, given all of these things, to spend time trying to bridge that gap, to do the translation work.

Jackson: (1:00:49) If that's the case, could you make the argument that there are other mediums that would be even higher impact? Like Ezra Klein is spending a lot of time podcasting.

Jasmine: (1:01:00) Oh, yeah, absolutely.

Jackson: (1:01:01) Or even something like the Abundance book, and like Derek —

Jasmine: (1:01:06) You and I, the book to go on the podcast.

Jackson: (1:01:08) Right, right. You drop the boulder in the lake and it's about the ripples. If you were thinking about it purely instrumentally, and the goal is that the world needs to better understand what's happening —

Jasmine: (1:01:17) Yeah.

Jackson: (1:01:17) — why is a great big, highly reported piece every quarter the right shape for that now?

Jasmine: (1:01:29) So one thing is, I write the piece and then I spend the next month going on podcasts, going to conferences, giving talks, and talking to people about it. So I'm very aware that writing is not the thing that reaches the most number of people — though in mediums like that, we're still probably talking like 10 million or something. It's still a pretty big number.

Jackson: (1:01:50) The permanent underclass was the headline for the Sunday issue, right?

Jasmine: (1:01:55) Yeah, it was the Sunday opinion cover — or actually the Sunday business cover. My sense, and they didn't tell me the numbers, but my sense is 10 million-ish would not be surprising to me. So it definitely reached people even though it was writing. And then beyond that, the reason I write less than other writers — most full-time writers write more often than I do — is that I devote more time to going on podcasts, because I agree that most people are not going to read the writing. You do have to get the message out in as many mediums as possible. I'm very comfortable doing that and I want to do it. I see it as part of the work: if I do all this research, I want as many people to hear about it as possible. The reason I start with writing is basically because I think it makes me the smartest. That's basically it. I like the process of writing.

Jackson: (1:02:37) It makes the idea the most solid too, to build on.

Jasmine: (1:02:39) Yeah, you just synthesize. Podcasts are very exploratory, writing is more like synthesizing, and they're both really interesting. For me, when I'm reporting and interviewing, I'm in exploratory mode. Then I bring it back in, compress it down into a single 3,000-word piece or whatever. And then I go broad again when I go have all these conversations about it. That feels like the best thing for my learning as well as for the impact. And also, finally, a lot of important people still read. Most people don't read, but I'd argue that among people with influence in business, in politics, in media and tech, a lot of the most influential people are still big readers.

Jackson: (1:03:17) You gave this Substack talk at the Substack event recently about AI and writing and so on. One of your points was that the value of polish is going to go down and the value of charisma and style and weirdness is going to go up. In some sense that would describe your two forms of writing — New York Times and Atlantic stuff is very polished, frankly a lot less personality, and your Substack writing for me is much more fun to read. Granted, some of that is that I'm more inside baseball already, and the two go well together. But I was thinking about that in terms of maximizing impact. Obviously the New York Times is going to be fine, the Atlantic is probably going to be fine. The direction of travel, though, along the lines of your point would be that if anything, we want to read more weird, stylish, charismatic writing. Do you think there's a world — maybe this is a weird way to ask the question — where in two or three years the writing someone like you is doing in the New York Times or the Atlantic is even more stylish?

Jasmine: (1:04:21) Yeah, I've thought about this one a lot. The thing about polish — perfect grammar, house style, et cetera — part of that is about the publication as a consistent brand. But it's also because polish used to be a very good proxy for how much work you put into something. It just meant that some human spent a lot of personal attention being meticulous about every line, every word, every fact-check, every citation. And that is becoming less and less true, because very soon — I'm pretty sure now, actually — if you fine-tuned some open-source model, you could build the New York Times house style bot. You could put someone's really crappy notes into it and out would come a piece in perfect New York Times house style. No personality, but perfect New York Times house style, because it's very rigid, especially on the news side rather than the opinion side. I felt it was basically like code syntax — every "she said," every word, really means something. So in a world where polish is no longer a very good proxy for how much work went into something, you're going to need other proxies for the fact that there was a human who spent a lot of effort on this.

Jackson: (1:05:22) Yeah. Katie Weaver talking about bread was not very formulaic.

Jasmine: (1:05:25) It was amazing. So good.

Jackson: (1:05:27) Granted, subject matter matters. But is it so insane to imagine a reform of real reporting that reads more like that?

Jasmine: (1:05:35) I hope so. There's a lot of distrust in media right now from the public, and I think part of that is that they don't like that reporters are pretending there's this single objective view from nowhere.

Jackson: (1:05:43) Yes.

Jasmine: (1:05:44) And it would be better, I think, if reporters were a little bit more open and honest about the fact that they are people with subjectivities and very particular emotions. When you share that on the page, I think people trust it more, not less.

Jackson: (1:05:56) It's a fine line, but I think you're right.

Jasmine: (1:05:58) Yeah, it is a fine line. I think it's more interesting, more engaging — in a way, perfecting your own voice and style is actually more evidence of work than having the polished thing that ChatGPT can turn out now. So I think that's really important. And I've talked to editors at these legacy publications who have explicitly told me they're trying to be more stylish. I believe the New York Times newsletters now allows their writers to write in the first person. Before, it was unthinkable to allow a New York Times reporter to use the word "I." Only recently have they started to think, maybe in our newsletters they can use it. So I think they're coming around.

(1:06:45) Going Independent, Risk, and Commercial Tradeoffs

Jackson: (1:06:45) All right. I want to talk a little bit about writing and Substack broadly, but also maybe the original decision to go independent. Fun fact: you and I both started what we're doing, I think, the same month — November 2024. Which is a fun synchronicity. I think my first question on this is — and you've written about this — you left a successful career. You were the head of, I think, like, core product.

Jasmine: (1:07:13) I led the core product team.

Jackson: (1:07:14) Led the core product team at Substack. You'd been there for a while, you were doing very well, ascendant. And you did this theoretically very risky thing to go become an independent writer. Why was that not so risky?

Jasmine: (1:07:29) I think a lot of people overestimate how risky career decisions are. My thought process was fairly simple, which is that I'd spent long enough at Substack, and Substack was a company people knew, such that I felt I could get another product job within a year if the writing thing didn't work out. That was basically my insurance — if I hate writing or I'm really bad at it, one year from now I'll say I went on sabbatical, and I'll just get another job as a product manager at a different startup. I felt pretty confident I could do that. All I needed was about a year or two of runway. And then I asked myself: okay, now let's imagine a world where I make $0 for one or two years — is that okay with my personal financial situation? It was fine. It's fine for, frankly, most people who work in tech, at least. So I was like, okay, seems a little bit risky, I'll try it for a year. If I hate it, I'll get another job, and I won't have gone into the red.

Jackson: (1:08:23) When you put it like that, it sounds so simple. So what is catching most people?

Jasmine: (1:08:29) It's very baffling to me. What do you think is catching most people?

Jackson: (1:08:34) Well, I think it's worth making a few points. One is, it's definitely not the case that most people could quit their job for a year and be fine.

Jasmine: (1:08:40) Yes, sorry — most people working in tech who had spent the same number of years at a tech company as I did.

Jackson: (1:08:45) Because you also — I mean, you were three years out of school, four years out of school. You weren't super established.

Jasmine: (1:08:51) Right. But I had like $100k in my bank account. That's all you need.

Jackson: (1:08:56) Why didn't you do it sooner, to flip the question back on you?

Jasmine: (1:09:00) I really had a sense of what I wanted to do at Substack. Before I joined, I was a sociology major in undergrad and spent a lot of time thinking specifically about social media governance — how the way you design a social platform impacts the kind of information ecosystems that occur, the social dynamics, et cetera. It was a longtime dream of mine from undergrad to design a social media platform from scratch. So I specifically joined Substack when it was still just an email tool, because Chris said they were going to do social media later, and I really wanted to be a part of designing it. I got to do it, I got to scale it, people started using it, and we had product-market fit. And then there was nothing else at Substack I wanted to do. My team included podcasts and video, large publications, the iOS, web, and Android apps, DMs and chat — all of that stuff. There wasn't really anything else I wanted to take on. I felt like the core mission I came in with — to impact the media ecosystem in a positive way, and to learn to design a social media platform based on different incentives and have people actually use and like that thing — I'd kind of accomplished both goals. So I was like, okay, I'll move on to my other goals now.

Jackson: (1:10:07) I think the other part of this is — and maybe I should ask you directly — to what extent do you think more people would correctly identify how risky independent writing actually is if they had more confidence in what it was they actually wanted to do?

Jasmine: (1:10:24) Yes, I think so.

Jackson: (1:10:26) Like, what level of fidelity did you have over what this life could look like and what it would mean? You had, I think, some list of goals which were like, have fun and make money and a couple other things. Did you have a sense of what success would look like, with a lot of clarity?

Jasmine: (1:10:44) Yeah. So my first year, I treated as a trial year — 2025 for me was not "I am a full-time writer now." It was "I am trying this out and I'm totally fine with going back to tech after." I was truly fine with going back into industry. I didn't know if I was going to stay with it. So I told myself: am I having fun? Am I writing stuff I'm proud of? Am I on track to financial sustainability even if I'm not making that much money now? Those were my exit criteria for the first year, to decide whether to go into the second year. That's what made me comfortable with it. Now I have broader goals — I'm like, okay, I'm actually a writer now, so I've got to get good at writing.

Jackson: (1:11:19) So how far into the trial year did you know?

Jasmine: (1:11:23) Maybe halfway through, basically.

Jackson: (1:11:24) And was that mainly a product of the financial side? Mainly a product of the meaning and the fun? Mainly a product of, like, people liking this?

Jasmine: (1:11:33) I was having a lot of fun, and I was pleasantly surprised by how much demand there was for the type of writing I wanted to do. It is clear now that there's an incredible amount of demand for Silicon Valley culture stories. I guess I wasn't paying that much attention at the time — it just so happened that those are the things I'm personally interested in. But there is tons of demand for journalism from people in San Francisco writing about AI right now. So I got really lucky where I actually ended up going to basically the fastest growing part of the media market, and stuff — just growth and pitching and whatever — was easier than I expected.

Jackson: (1:12:15) Yeah. In some sense you were catering to an area where there was a ton of demand. I'm not saying it was totally deliberate. You weren't like, "I'm gonna go write creative poetry."

Jasmine: (1:12:29) Oh, absolutely.

Jackson: (1:12:31) Which is probably more crowded and has less demand.

Jasmine: (1:12:33) Yeah. I think the risks would be super different if you wanted to go into poetry. A lot of people could still afford to take a year off to do poetry, and that might be really valuable for them even if they end up going back into a corporate job afterwards.

Jackson: (1:12:48) And you just don't know. I mean, I didn't start a podcast for so long because I thought it was just too late. But why is patronage — and maybe this has evolved to the point we were just talking about, writing more for big publications — why is patronage the right business model, and such a good business model, for someone like you?

Jasmine: (1:13:10) Interesting. I don't think of my business model as patronage.

Jackson: (1:13:13) You think about it as gated content?

Jasmine: (1:13:16) Well, frankly, my income comes from a real mix of things.

Jackson: (1:13:20) How much — roughly — is Substack subscriptions? Presumably a meaningful chunk.

Jasmine: (1:13:25) There's Substack subscriptions, there's philanthropic grants, and there's freelance writing and speaking.

Jackson: (1:13:33) So one third each. I mean, arguably two of those buckets are patronage.

Jasmine: (1:13:40) That's true. I guess it is. I've never thought of it that way. And especially —

Jackson: (1:13:44) — because some Substackers, it isn't really patronage. It's like a subscription.

Jasmine: (1:13:48) Yeah, but I don't —

Jackson: (1:13:49) You don't paywall much.

Jasmine: (1:13:50) I paywall very rarely.

Jackson: (1:13:52) Right. So you're basically relying on a lot of goodwill, and then occasionally the Atlantic or the New York Times.

Jasmine: (1:13:57) Yeah. Which pay not enough.

Jackson: (1:14:00) Do you think your business model will primarily be patronage in two years?

Jasmine: (1:14:04) I think it would be nice to have things like speaking also be a bigger part of it, and to do some more paywalling, mostly for diversification reasons. The patronage situation I want to avoid — which is what I associate with the word "patronage" — is having one primary patron, because then you are highly likely to get audience captured and you're not really independent. Depending on their ideological goals, personal proclivities, whatever, it just feels like you're on a leash. Whereas the thing that is good about diversifying income sources — even within philanthropic grants — is that I take grants from very different organizations with very different political and ideological points of view, because I really don't want to be pigeonholed and I fear what it will do to my writing. So I would still be excited about continuing to get, say, roughly half my income from patronage or more than half. But what I think about more is: is it sufficiently diversified that I both am perceived as independent and actually feel that I am independent, and won't make my funders mad or something?

Jackson: (1:15:09) You and I already chatted about this a little bit, but you are someone who — probably unlike most journalists — could not only work in tech, but could probably go get one of the most coveted jobs in the world, whether that be working directly at an AI lab or working in partnership with one of them. It seems like you've found a way to support yourself. Maybe this even goes back to working on the Substack side. But one of the big stories of Substack is not just that writers can support themselves, but that writers have built massive businesses. To what extent do you think about the ability to build a really great business? Presumably someone like you could turn down the big money for a little while in the name of being independent, but at some point something's got to give. My sense is it's actually quite important that people on the far end of the power law of any discipline be able to do really, really well for themselves for that to become a long-term path. Like Emily Sundberg — I don't know how much she makes, but she's doing very well.

Jasmine: (1:16:33) She's doing very well.

Jackson: (1:16:34) That is very good for the world. And one of my senses is that the reason a lot of these big traditional media publications are struggling to retain talent is because the best people can make way more elsewhere.

Jasmine: (1:16:57) Yes.

Jackson: (1:16:58) To put it another way, if you can make way more money doing podcasting like Ezra Klein has, you probably are going to do that — or have a greater impact there.

Jasmine: (1:17:11) I agree with about half of what you're saying. A big part of the reason Substack was founded was that Chris thought it was really messed up that all of these incredible bloggers who had changed people's lives and who had huge readerships — like Scott Alexander — were doing it for free rather than getting rich as they deserved to be.

Jackson: (1:17:29) And doing a day job on the side.

Jasmine: (1:17:30) And doing a day job on the side. He was like a psychiatrist. Chris thought it was crazy that if you work at the New York Times or any legacy newsroom, the person bringing in 100 times more views is getting paid maybe 2x the salary because of salary bands. He felt the market didn't really value writers and creators accurately, and that you should be able to get rich by being a top one-percentile writer or journalist or anything. And I completely agree with that. I think it's really important that Emily Sundberg and Scott Alexander and whoever can get rich doing what they do. It's amazing. At the same time, I really don't think I'm doing this to get rich. If I wanted to get rich, I would go work for an AI lab.

Jackson: (1:18:10) Right. That's my point.

Jasmine: (1:18:11) I took a very large pay cut. I think I'm going to make up that pay cut on accident — like, I think I will eventually make up my old tech salary — but I was perfectly happy to take a very large pay cut in order to do the thing I really wanted to do. And I have turned down a lot of opportunities to make more money. Yeah. Maybe this wouldn't hold if I had kids. For sure.

Jackson: (1:18:34) My question is more about the durability of it.

Jasmine: (1:18:37) Right.

Jackson: (1:18:38) Like, the most important thing for any creative endeavor — or really any endeavor at all, you think about this in the context of investing or whatever — is being able to stay in the game. And so I guess that's my root question, which is: a thousand true fans is a really cute idea, but actually $100k isn't enough for most Americans to live in a big city.

Jasmine: (1:19:00) Yeah.

Jackson: (1:19:01) And so maybe I'm too much of a capitalist, but it seems to me that it's really, really important that we find ways to create great financial upside, because otherwise everyone is going to go work at Anthropic.

Jasmine: (1:19:16) No, I mean, I completely agree with this. I worked at Substack because I thought it was really important that the best people get paid. And so my opinions about the ecosystem are pretty separate from my opinions about myself, because I understand that as a 26-year-old with no children, I just have different needs. I'm totally fine being kind of broke for a little bit so that I can just have a lot of fun.

Jackson: (1:19:37) But I want you doing great reporting in 10 years.

Jasmine: (1:19:39) I do too. I think that would be good. But I don't plan that far in advance. One of the keys to happiness for me is I literally never plan more than a year out. I do not know what I will be doing in 10 years — it's very unclear to me. I would make a sub-50% bet that I'm still a writer in 10 years, by the way.

Jackson: (1:19:56) I think you will probably be very successful financially and otherwise.

Jasmine: (1:19:59) I hope so. But I think there's also a way in which people overrate financial capital. There are a lot of forms of capital in the world — attentional capital is super valuable, social capital is super valuable — and I think a lot of people underrate those other forms by focusing too much on money. And I am building capital right now.

Jackson: (1:20:20) Yeah.

Jasmine: (1:20:21) I'm just building other forms of capital —

Jackson: (1:20:23) — that you can launch a VC fund one day.

Jasmine: (1:20:25) I know I can cash these things in when I want to or if I need to. I know that if I wanted to put enterprise sponsorships on my Substack tomorrow, I could do it and make a lot of money. I know I could do comms consulting for tech companies and really help them out. I know how I would make money. But in a way, I think it's harder to build attentional capital and cultural and social capital than it is to build financial capital. And more importantly to me, I have more fun and I get to do the things I want. I feel this immense luxury of having the freedom to not do a 9-to-5, to travel and write and talk to all these people. And if I need to start cashing some of that in, I know how I'll do it. The fact that I know how to turn it into financial security is enough financial security for now.

Jackson: (1:21:24) I think the fear for people listening — especially on the more journalist side of things — is that the only way to cash out is to fully compromise. So maybe it actually just goes back to your original point, which is that you think —

Jasmine: (1:21:46) — like an Emily Sandberg is compromised? You think that an Acquired podcast is fully compromised? I feel like the great thing about independent media, whether it's Substack or podcasts or whatever, is that you're not. At least I don't feel that you're that compromised being one of these types of businesses.

Jackson: (1:22:02) I do not think that Emily nor Acquired are compromised. I wonder about the different types of media and reporting, though. It's actually quite a rare thing to have someone who could work at a big AI lab be reporting on AI labs. Maybe one of the reasons there aren't more journalists covering AI is because they all work there.

Jasmine: (1:22:30) I mean, I think about this a lot. I know a lot of people who worked in editorial at publications, who worked in policy writing, who have now gone to work at one of —

Jackson: (1:22:38) The cynics' view is that this is just what capitalism does over time. But I think it goes back to your original point, which is that being diversified is the resilience against this.

Jasmine: (1:22:49) Yeah. When I talk to podcasters about how they think about their ads, a lot of them really focus on diversification so that if they accidentally offend one tech company that might be a sponsor, they obviously don't want to lose the sponsorship, but they've diversified so that it is possible to do so. Or you have news and business splits in newsrooms, and even some large Substacks I know will do the same thing — the person selling the ads is different from the person doing the writing, and when the writer is writing, they have no idea what ad is going to appear above or below it. These are small things, and you can make arguments that they're imperfect in a bunch of ways. But I think it's possible to be ethical and conscious about how your sponsors, your advertisers, your patrons are going to compromise your independence. It's obviously harder for someone doing hardcore corporate accountability journalism. I think that kind of work is probably going to have to be a combination of philanthropically and subscriber supported, because corporate sponsors are not going to go for a lot of that — which is both rational and sucks, because I do think accountability journalism really matters. And maybe it's possible that I piss somebody off so badly with my reporting this year that, right now I still think I could go work in tech again, but maybe they're already so mad at me that they don't even want me doing comms for some tech company. And I think I just have to take that risk. It's okay.

Jackson: (1:24:16) I think also, ultimately, the negotiation between you and your readers is that you are going to be a good steward of this.

Jasmine: (1:24:22) Yeah.

Jackson: (1:24:23) Like, I think that is really the trust thing.

Jasmine: (1:24:26) Yeah. I mean, again, I might be making terrible financial decisions. I just don't plan that far into the future.

(1:24:35) Great Writing: Style, Voice, and Resisting Summary

Jackson: (1:24:34) I want to talk about good writing. You say the point of long-form writing is to say something not reducible to a tweet. "The TL;DR of a Susan Sontag essay could only be every word of it," wrote A.O. Scott. How do you get closer to writing where that statement could be true?

Jasmine: (1:25:04) Part of it is style and craft. If you have a really distinct voice and a really unique style — where you're not using clichés, you're using crazy metaphors and retelling anecdotes and whatever — then the style becomes part of the content and part of the substance. You would never read a ChatGPT summary of a Sam Kriss essay. You just lose everything about it when you lose the style. And I think that caring about voice and style, using fun words and writing fun sentences, is part of that process. I also see writing as a form of entertainment — it's not just "eat your vegetables." Tweets are not entertainment. TL;DRs are not entertainment. When I write about AI, I try to make it somewhat interesting to people. Most AI writing is pretty dry and technical, and a lot of it's for business reasons so I get it, but I see part of my job as entertaining readers and keeping their attention captivated even while I'm trying to share serious information. And that's another way you avoid people wanting to TL;DR you or have their AI agent summarize you in a news roundup. If the process of spending five or ten or twenty minutes with your essay is intrinsically enjoyable, people are not going to want to TL;DR it.

Jackson: (1:26:33) Yes, yeah. Maybe there's some payoff or some education at the end, but that's almost the bonus versus the inversion, which is like, what's the point of the article? Why are you clickbaiting me?

Jasmine: (1:26:43) Or whatever.

Jackson: (1:26:44) What about voice? To what extent do you think you have consciously evolved or improved your voice? Or is it almost one of those things where you actually have to try less in order to have your voice come out?

Jasmine: (1:26:59) Yeah, definitely both. Sometimes when I feel really stuck or I feel like my writing's really dry, I'll take out my paper notebook and just start writing in it. Because somehow that feels like a less self-conscious form of writing — it's a more natural brain-to-page. And then my voice comes out more. I never thought about my voice until maybe earlier this year, late last year. This is actually one of the things that Claude helped with, where sometimes people would tell me, "I really like your writing voice," and that's great to hear, but I actually had no idea what my writing voice is. I don't try. I just write how I want to write. So I had Claude read all of my archive of Substack posts and describe my voice back to me, and I actually found that super helpful.

Jackson: (1:27:43) What did it say?

Jasmine: (1:27:44) One thing is a blending of registers. I talked to Celine about this as well, but I think people's voice is an amalgamation of the experiences they've had and the social context they're in. As somebody who spends a lot of time in the AI industry, a lot of time on the internet and in communities with a lot of weird online slang, but who also likes to read literature and critics and serious journalism — the metaphors I choose, the words I use, the references I make are going to be an amalgamation of all of those different influences. And I think that's true of every single writer. So one characteristic of my voice is that I tend to mix the rationalist jargon with the internet slang with the literary critic's point of view. Another one is that my voice is very first person. I just naturally write as an "I" — I say "I feel," "I see," "I'm not sure about things." Even when I'm tackling politics and AI and whatever, I just find it natural to write in first person. I like scene-setting and I like anecdotes, relative to writers who operate at a more abstract level. The way I think is so often tied to very specific conversations I've had, or quotes I remember, or things I see in the world. So when I'm writing, I'm just trying to take the thing in my head and give it to somebody else and hope that it also hits. Maybe it's not as rigorous as reading a paper or looking at a poll to tell someone an anecdote, but I tell people the—

Jackson: (1:29:28) —anecdote. But it might be as true. Yeah, "true" might not be the right word, but — is it important? Do you care at all about writing how you speak?

Jasmine: (1:29:46) I don't try to. I try to write how I think more than how I speak.

Jackson: (1:29:50) Yeah.

Jasmine: (1:29:50) Do you think I speak like how I write, or is it different?

Jackson: (1:29:58) The vibe is the same.

Jasmine: (1:29:59) Okay.

Jackson: (1:29:59) But again, I don't know.

Jasmine: (1:30:00) I'm happy to hear that because it makes me feel like I'm not—

Jackson: (1:30:02) Like, it's authentic. Yeah.

Jasmine: (1:30:05) Yeah, I have no idea.

Jackson: (1:30:07) I mean, I guess the real litmus test is when someone who's read you for a long time meets you in person. Does it feel—

Jasmine: (1:30:16) I've actually asked a couple people this question. They tend to think I am roughly the same. I don't feel like I switch a lot — I talk to a lot of people from very different communities, but I think I'm roughly the same in all contexts. So I don't think people are very surprised. I like to context collapse. I like to introduce my friends to each other. I think one thing that some people don't like about having a blog or being very open online is that they feel uncomfortable with the idea that people from different parts of their lives will see the same self they present. If you write a personal essay, your mom and your boss and some stranger are all going to see it. I have a relatively consistent self, so I've just accepted that they will.

Jackson: (1:31:00) Another version of this is just: are you as fun to hang out with as you are to read? Which I'm not sure is the case for every writer.

Jasmine: (1:31:06) Oh God. I've definitely been to some conferences where most writers are just not very compelling speakers. And that's okay because they have other talents.

Jackson: (1:31:16) Yeah, yeah.

Jasmine: (1:31:17) But many writers can't really speak.

Jackson: (1:31:23) You wrote, "Blogging is not only a search query" — referencing Henrik's excellent post — "but a kind of prompt for human wisdom. The world will respond with the seriousness that you put into it." Can you say more about that?

Jasmine: (1:31:42) I think when I put a lot of effort into writing something, it's proof of work, and it means that people — strangers — will put a lot of effort into the response. One thing I think about a lot is that I feel really rich in information. I love information, I love knowledge, I love facts, I love knowing stuff. I'm not very capital-rich right now, but I get to know a lot of stuff and I get to talk to a lot of people.

Jackson: (1:32:14) This is what Tyler is — he wants to be like an information trillionaire.

Jasmine: (1:32:17) Yes, totally. I completely understand that. And one of my favorite things about writing is, if I pour my heart and soul into a piece and it's really evident — I remember I wrote this 7,000-word China essay last year. It was about my family, it was about China, it was about technology, all of these things. I spent a lot of time on it.

Jackson: (1:32:36) You got a Kevin Kelly comment too, which is so—

Jasmine: (1:32:38) Oh yeah, I was really happy about that. But my favorite thing about it was getting all of these emails from people who had either lived in China or spent time in China at various points throughout the last 50 years, who just wrote me essay-length emails about their own stories and experiences — like growing up in Shenzhen and seeing the phrase "skyscrapers sprouting up like mushrooms," which I think is an incredibly beautiful, evocative phrase. And I'm like, someone just wrote this in a freaking email to me, a stranger. That's the thing I feel really fortunate to get. If I spend a lot of time writing a piece, I get rewarded many times over with other people's memories, disclosures, whatever. And I just think that's really special.

Jackson: (1:33:27) One of the ways the Internet can be the most special is that phenomenon specifically — almost inside of it is like, you're allowed to care this much. Not quite exactly the same, but it's something to that. The Internet is this place where you can find the weirdos who care as much as you do.

Jasmine: (1:33:46) Yeah. And they really can be anywhere. The original dream of the Internet is still alive to me in that sense, where I think that through writing I'm able to maintain a lot of friendships with people who live in other countries or other cities, especially if both of us are writers. And I've noticed that if I like someone's tweets, for example, it doesn't actually mean we're going to get along in person. But when someone and I like each other's long-form writing, we almost always hit it off. There's something that is deeper — the connection is stronger. That, to me, is just really exciting. Do you think it happens as much for podcasting?

Jackson: (1:34:24) I probably get less wordy responses, which is probably telling, but it's certainly like a bat signal.

Jasmine: (1:34:34) Yeah.

Jackson: (1:34:35) I mean, I certainly have friends who I've met through interviewing people, and I have friends who I've met because they like the podcast. Yeah.

Jasmine: (1:34:45) Do you think that your podcasting self is roughly who you are?

Jackson: (1:34:49) Oh man. I mean, you and I talked about this a little bit. There is an element of this that's performance. I'm conducting the orchestra with one hand and I'm playing in the orchestra with the other hand.

Jasmine: (1:35:01) Yeah.

Jackson: (1:35:03) And there's a version of me that could not have any notes in front of me, do no prep, and would be maximally present — less performance, more like me in real life. And then there's a version of me that was just perfectly scripted. I'm trying to find the right balance between those two. I'd have to ask people who know me, but I think it's probably fairly similar.

Jasmine: (1:35:28) You seem relatively the same as you did when we got coffee.

(1:35:35) Literary Inspirations, Favorite Essays, Writing vs. Thinking, and Getting Better

Jackson: (1:35:32) I'll take it. You wrote somewhere that you carry around printouts of your favorite essays and put them in your laptop bag. What are some of your favorites?

Jasmine: (1:35:47) The one I always talk about and probably look at the most is Gideon Lewis-Kraus's 15,000-word essay about Google Brain in 2016.

Jackson: (1:35:56) I don't know it.

Jasmine: (1:35:58) It's called "The Great AI Awakening." It was in the New York Times Magazine, and I think it is the best piece of AI journalism I've ever read. It's everything I aspire towards in writing, which is that writing has so many components — there's reporting, there's prose style, there's the ideas and whether the ideas themselves are good. And I think this magazine story is all of those things. It's really funny. He writes about the different characters — these AI researchers and engineers — in ways that turn them into these delightful, hilarious characters, as if you were writing a novel. The prose style is really wonderful. It's very deeply reported. He literally just hung out in the Google Brain offices over and over and talked to Jeff Dean and talked to Quoc Le and spent a lot of time with these people, really learning about both who they are and the technical work they were doing. He can summarize very complex AI papers for a layperson in ways that are beautiful and yet also accurate. And then it was just so prescient. The thing that I think is maybe most amazing is not just that it's a great work of nonfiction — which it is, he's a very good prose stylist — but that this was 2016. In 2016, DeepMind was doing AlphaGo and OpenAI was doing robot hands or something. No one had really cracked LLMs yet. But in this piece, he's writing about Google Translate upgrading to a more LLM-based model and he's like, yeah, maybe language is the pathway to artificial general intelligence. He says all of these things that we now take for granted as being true about AI progress — as a New York journalist in 2016, writing about Google Translate upgrading their model. I think it was so prescient, so brilliant, and it just really stands the test of time. There are very few pieces about AI that stand the test of time like that. That's a ten-year-old piece and I still reread it.

Jackson: (1:37:57) That's pretty amazing. Maybe zooming out but on the same thread — you wrote somewhere that you want to be able to observe the world like Maggie Nelson. Are there other essayists or writers or journalists who have been influential on you? We talked about Susan Sontag a little bit.

Jasmine: (1:38:16) Yeah. I love Susan Sontag. I love Maggie Nelson in the culture critic camp. I think Gideon's excellent. I think Dan Wang's work is very good. A lot of the people I admire have both a cultural lens to their work — a deeply humanistic lens, very interested in people specifically — but they're also sharp, and I like sharpness. I really like Anna Wiener's work. Meghan O'Giebel, who wrote God, Human, Animal, Machine — she's an essayist. Her first essay collection was about the Midwest and Christianity, and her second one was about almost like the theology of how we think about machines and artificial beings.

Jackson: (1:39:13) Wow.

Jasmine: (1:39:14) Because she has this theology background.

Jackson: (1:39:14) So cool.

Jasmine: (1:39:15) It's super cool. I like the Christianity essay collection as much as I like the AI-ish collection. Meghan O'Giebel.

Jackson: (1:39:23) I have to check this out.

Jasmine: (1:39:24) She's really cool. I love her work. It's just these non-linear ways of thinking, and again, it's a mix of disciplines. You don't think that these disciplines should illuminate each other, but they do. I love Peter Hessler's China books and essays. He's one of the classic guerrilla ethnographers — he's not trained in ethnography, I think he was trained more as a journalist. But his books are super funny, very character-based. They're not just "let me cite a bunch of stats and charts." It's like, I'm going to hang out with a Uyghur tradesman in Beijing and eat lamb skewers together, and I'm going to go road-tripping around the Great Wall and then get caught by the police —

Jackson: (1:40:02) Almost Bourdain-ish.

Jasmine: (1:40:04) — that style of living in the world as the credibility for your writing. And again, it's such a human focus. It's all about these very particular characters he meets and the real relationships he builds, like when he taught in Chengdu. That kind of writing is really influential to me.

Jackson: (1:40:22) What does sharpness mean?

Jasmine: (1:40:25) Ooh. That's a good question, because sharpness and precision are two things I really value, but I've never attempted to define them.

Jackson: (1:40:40) Are they the same?

Jasmine: (1:40:44) No, but they're related. I think it has something to do with clarity of thought. It's like knowing what your point of view on a piece or a story or an issue is. You're not just doing summary. You're not making 10 tangential arguments at the same time. There's an angle to things. There's a thing you're trying to accomplish, whether it's a frame or a question or a conclusion. I think being sharp has to do with your clarity of thought and how much you know what you're trying to do with a piece.

Jackson: (1:41:21) Yeah, it cuts very cleanly. There's something really elegant about when somebody can make a point like that.

Jasmine: (1:41:27) Yeah.

Jackson: (1:41:30) You told me — or maybe we even talked about it earlier — that figuring out what you think is very different from when it's converted. Like when it's in your head and when it converts into writing, there's this blob that needs to convert. Is that just about clarity and precision? What I'm asking is: will you think about something and come up with your conclusion and then convert it into words? Or are you actually thinking on the page? Does that delineation make sense?

Jasmine: (1:42:06) Yeah. A lot of people say writing is thinking. I believe this is true for many people, and writing definitely helps me think. But the primary work of thinking is, to me, separate from writing. Maybe I'm doing some writing — taking notes or scribbling or something — but they feel like separate parts of the process. The act of synthesis, of pouring over a bunch of notes and interviews and papers and essays side by side and trying to figure out what I actually think, is a separate process from: now that I know what I think, now that I know what I'm trying to accomplish and the frame I'm trying to introduce, how do I make that legible and entertaining to the reader?

Jackson: (1:42:47) But when you're in the thinking process, are you just sitting there arranging puzzle pieces in your head? Or is the thinking actually this mix of reading and listening and going on a walk?

Jasmine: (1:42:58) Yes, all of these things. When I come back from China, I have like 150 bullet points in my notes app — what the fuck is going on in this notes app? I'm rearranging them into little blocks, trying to find clusters and themes that emerge. If they were sticky notes, I'd be doing it on a wall, but I just copy and paste around my notes app.

Jackson: (1:43:16) It's kind of writing, but it's not prose development.

Jasmine: (1:43:18) No, it's not prose necessarily. I also call people. I'll talk to Claude sometimes. With the AI and labor piece, I spent a lot of time talking to friends who I thought were really smart thinkers about AI and the economy, where I was just like, okay, this is a really complicated issue. It's very controversial. Economists haven't decided what they think is going to go on in the labor market. And I need to figure out what I actually think is going to happen, and then what I want to say in my piece. I would just debate on the phone, think out loud, get pushback from friends who I thought were smarter than me. And through the process of conversation we would arrive at an outline, and then I would stare at it again and push back here and there. But it's really messy. It's not a clean process.

Jackson: (1:44:01) You also wrote — maybe it's a similar note — that at some point in every project, it stops living in the Google Doc and starts living in your mind. Everything else you see and hear gets filtered through the frame of the essay. Is that just a byproduct, or is getting to that point actually a part of your writing process? Does it change once you're seeing the world through that lens?

Jasmine: (1:44:27) It's a good question. I don't think it's intentional, but I actually think it is part of the writing process now that you bring it up. I think the piece gets a lot better at that point.

Jackson: (1:44:34) Yeah, there's additional stuff that can happen only then, right?

Jasmine: (1:44:38) Because that's when I start to make non-obvious connections. When you're just doing research, it's very scoped — here's my reading list syllabus, here's my sandbox, I'm going to do these 10 interviews, and you're thinking within the sandbox. But once I'm immersed enough in a piece, every conversation — it's probably kind of annoying to my friends — I'm always bringing up what I'm thinking about. And I don't care if the person's an expert, I just want to know how they react to it. Or everything I see in the world kind of connects to the piece. Like, the companion piece to my permanent underclass thing that I did on my Substack — I was in China finishing the permanent underclass piece, and I was there to learn about open source AI or something. The point was not for me to think about the labor market. But I couldn't stop looking around and noticing how many fake jobs people were working in China. You have street sweepers in every park who are literally sweeping clean streets. You have people checking bags in the subways, and they're not actually checking — we all know they're not actually checking. There are security guards monitoring extremely safe areas every 200 meters in broad daylight. And it only became clear to me because I was in this mode, writing this piece, unable to stop thinking about the labor market: oh, these are jobs programs. I'd been to China a lot of times before, and it never clicked that they were government jobs programs — that because China has high levels of unemployment, the government is very motivated by social stability and is willing to have some inefficiencies in its bureaucratic spending to give people something to do. Because the Chinese government believes that giving people work is just an inherent good in and of itself. But I literally would not have noticed that had I not been in the mode of my New York Times piece, with everything getting filtered through it. A park I'd run through dozens of times — I'm suddenly like, why is that guy sweeping a clean street? I never thought about it before. That's when you start to make all of these crazy connections, and that's one of the most satisfying parts of writing for me.

Jackson: (1:46:32) Yeah, I have the same thing. After interviews, I just start — like, the rest of the next week I'm just filtering back, almost processing it after the fact. Fully digesting it. Henry Carlson talks about this when he's working on a piece. He's like, "It's like a room that I have a key to, and when I publish the piece, I'm throwing away the key." So sometimes he won't publish it because he wants to keep that headspace a little while.

Jasmine: (1:46:59) That's cool. I like that. I do feel a little bit like that too, because when I publish, I don't return as much to it anymore.

Jackson: (1:47:06) Or at the very least, now you're ready to go talk about it and spread the ideas. It's formed now.

Jasmine: (1:47:11) And so I kind of know what my thesis is. I'm not in the exploratory mode anymore. One of the things they do in reporting, in interviews — and this is some of the capital-G journalism stuff again — but a lot of journalists, when they're asking people for interviews, they're just really trying to get one quote out of them. They're going to a person because they think that, based on the organization they work for, this person is going to say this one sentence, and they need to plug that sentence into their piece. They just want you to be like, "Oh yeah, Elizabeth Ward's policy is good or bad," and you can predict what they're going to say even before you do the interview. For me, the thing that I find most fun about reporting is the process of discovery. I'll go into especially the first set of interviews — like, the first 15 calls or whatever — and I'll tell the person, "I don't know what I think about AI in the labor market yet. I'm still trying to figure out this issue. So just so you know, I'm not just trying to get a single quote out of you. I'm trying to work through the issue. Can we just talk?" Because I'm in this learning mode, this discovery mode, and I really like being in discovery mode. It's much more fun. And I think it also makes them trust me more. I will never say I'll never reach a conclusion — they know I'm going to have an opinion eventually — but it is malleable for quite a long time, and I like feeling like the ideas are malleable.

Jackson: (1:48:23) Yeah. I think you wrote somewhere that you hate all your old writing, or you think it's all bad. But I'm curious how you think specifically you've gotten better.

Jasmine: (1:48:44) One thing I noticed between when I wrote at the beginning of last year and when I write now on Substack is that I hedge much less.

Jackson: (1:48:53) You're sharper, probably.

Jasmine: (1:48:54) Yeah, I think I'm sharper. When I read some of the pieces that I thought were pretty good at the time — I did this one on what AGI means, trying to disambiguate the term — I sound so nervous and insecure and self-conscious. Every other sentence, I'm apologizing for not knowing anything about AI. And it's because at the time, I really felt that way. It was an accurate representation of how I was thinking. I don't know if I regret it, because it was warranted humility — I hadn't spent that much time in those debates, in those worlds yet. So I did feel like I was in the process of learning. But I think it was a less effective piece of writing because I was hedging so much. And I also had this complex when I was writing last year where I evaluated all my own work by what my most expert friend would think about it. When I wrote about AI, I thought about the people I knew who had been working in AI the longest. When I wrote about China, I thought about the China scholars and professors I knew. And when you're in that mode, you think all your writing is really bad. It's not good for the psyche.

Jackson: (1:50:00) Well, on one hand, that probably does make your writing better.

Jasmine: (1:50:03) It does.

Jackson: (1:50:04) You aim for the stars or whatever. On the other hand, it can't be the litmus through which you judge the final product, because it's not for them.

Jasmine: (1:50:11) Yeah, it's not for them. That's the thing — I think it took me until late last year to realize that I'm writing for an audience that is smart, but I don't write for the world's expert on this issue. I'm just trying to move their understanding forward. And who knows, you might surprise yourself —

Jackson: (1:50:28) You actually might end up influencing the experts too, in an unexpected way.

Jasmine: (1:50:35) Sometimes I do. Like, I actually do feel like — Gideon —

Jackson: (1:50:39) Gideon Lewis-Kraus, maybe, to that point. He actually maybe called it the language model point?

Jasmine: (1:50:45) Yeah. I mean, I think sometimes the outsiders do see things more clearly than the known experts in the field. But yeah, I read a lot of self-consciousness and hedging and insecurity in my past writing, and I have a little bit less of that now. Which is not to say I think I'm an expert — I just realize that the best end product is not constantly apologizing for its own existence.

Jackson: (1:51:08) You said you wouldn't write if you couldn't publish.

(1:51:09) Writing to Publish, Authenticity, and Art

Jasmine: (1:51:11) Oh, yeah.

Jackson: (1:51:12) And yet you also have this amazing piece somewhere where you're talking about Hanya Yanagihara — excuse me, I'm sure I'm butchering that — and this idea of the audience of one, this more autotelic, non-instrumental thing. How do you square those two things? And are you sure you wouldn't write if you couldn't publish? I'm not sure I believe you.

Jasmine: (1:51:38) I believe me, because I'm not a journaler. I don't journal.

Jackson: (1:51:43) No?

Jasmine: (1:51:44) Maybe I journal once every two to three months. I don't write when I'm not publishing.

Jackson: (1:51:50) So hold on — when you were at Substack, you were sometimes publishing on Substack. And we talked about how you've been writing since you were a kid.

Jasmine: (1:51:56) Yeah.

Jackson: (1:51:57) You've basically only ever been writing if you could share it.

Jasmine: (1:52:00) Yeah, basically. I mean, sometimes I try to write something because I think I'm going to share it, but then it's crappy and I don't. But everything has that intention of having an audience. And to be clear, sometimes I write emails. I like writing emails to my friends and my mentors, and to me, that counts as publishing.

Jackson: (1:52:15) Totally, totally.

Jasmine: (1:52:16) Because someone's going to read it. When I worked at Substack, we had an internal Substack, and a lot of the way that — I joined as an intern with no college degree, very junior in the Substack hierarchy — a lot of the way that I ended up leading the core product team and moving up in the organization was that I wrote a lot on our internal Substack. I'd write my product takes and my strategy takes, and enough people liked them and read them that they eventually involved me in better conversations, gave me more responsibility, things like that. But for me, writing is a form of politics. It's a way to change the world, a way to meet people, to shift discourse. It's communication. Because again, my thinking process is separate — I don't think by writing. I think in some other weird, blobby way that involves conversations and notes and readings.

Jackson: (1:52:58) This is interesting.

Jasmine: (1:52:59) Because writing for me is completely a communicative act. I'm trying to entertain people, I'm trying to share an idea, I'm trying to make a dent in the discourse. And because of that — like, if I was just trying to think, I wouldn't be writing.

Jackson: (1:53:15) Well, maybe it's not an audience one, but there was this — I think it was when you left Substack, maybe you were still there. You said, "I think working a lot has made me a worse writer. My prose is tight and the excess trimmed. Deformation efficiency is paramount. I write like the $12 dust salad, the bar that packs 20 grams of protein and plastic into a 1,200 calorie brick. But good writing, like a good meal, needs fat. It should indulge readers. It's meant to be chewed and enjoyed, affording a generous escape from the prosaic and the mundane." And then you go on to say you're talking about increasingly feeling like so much of your life is instrumental. You say, "As I've gotten older, I've increasingly viewed art as enjoyable but auxiliary, secondary to the real work of politics. In high school I traded creative writing for debate," so on and so on. "But I have begun to come around on this. Art, or at least good art, is defined by its non-instrumentality. Art is not useless, but it is use-agnostic, and that's what makes it so useful." After all, most of your writing is pretty instrumental. That's not to say it's not artistic or creative — I think it is so many of those things. Do you have an ambition to do more artistic things?

Jasmine: (1:54:37) I do. I don't know that I'd be good at it.

Jackson: (1:54:39) That's okay.

Jasmine: (1:54:40) I really admire Hanya Yanagihara so much, because when you read that New Yorker profile of her — which I think that piece references — she doesn't think about the audience. She doesn't care if people like her. When she drops the boulder into the lake, she doesn't look to see if it makes a splash. And it does, because her work is very good and she cares about the craft purely internally. But I just don't know that I'm built like that. I think it's good, because audience capture does pose a lot of risks. And if you don't care about the craft and the rigor for its own sake, you're not going to produce work that matters anyway. But I think I see myself more as a consumer than a producer of art, for now. Art is something that I consume and appreciate. I don't actually see my writing as art — my writing is mostly communication.

Jackson: (1:55:43) Yeah, it's very, very creative.

Jasmine: (1:55:46) I try to be. I try to have fun with it. Not every piece is — I'm not writing policy memos. But it's always for an audience. I think that's the thing: even when I'm trying to be creative, it's because I'm trying to entertain.

Jackson: (1:56:02) Maybe on the last note, another question that Celine brought up that I loved — what does it mean to do beautiful or expressive nonfiction writing? You said somewhere else that you don't write fiction because there's no grander game than what is actually going on. Which is awesome. Sure, many would disagree. But yeah, on that last note of maybe just being entertaining — it isn't lifeless. I don't know if I would use the word beautiful across all contexts, but some of it, like that China post, was amazing and so lively. And certainly expressive.

Jasmine: (1:56:43) Yeah. I'm reminded of our conversation about AI writing and AI art. At some point the AIs will get good enough that they'll make movies and write essays that are better and more enjoyable for the reader than a lot of our work. I'm ready for my AlphaGo moment on the writing. It's going to happen. But to me, the point of writing is to communicate the life behind the work — the fact that the person who has generated the text is a real living being who has desires and fears and motives, who is a subjective person with weird biases and family baggage and trauma and whatever else.

Jackson: (1:57:24) Benjamín Labatut talked with you about this, with Lee Sedol — he's like, Lee Sedol is super into K-pop.

Jasmine: (1:57:28) Yes. I thought that was amazing.

Jackson: (1:57:30) That's what makes him a human.

Jasmine: (1:57:31) Exactly. And I always think about when people listen to a pop song — sometimes they're just listening because they want to listen. But when someone's a real loyalist to a particular pop star, they usually have this grand narrative about their life and the dramas in it. Like, "she wrote this album because she's going through this breakup," whatever. And again, when you marvel at Michael Phelps swimming, or Alysa Liu and her ice skating performance, it's not just the fact that she's a good skater. Of course she's a good skater, and we definitely appreciate that. But it's the life behind the work. We're compelled by Alysa Liu because we hear about her dad, we hear about her quitting, we hear about her coming back and learning to love it for herself.

Jackson: (1:58:10) And you see it on her face — you see it on her face looking down at her.

Jasmine: (1:58:12) It's the expressiveness. It's not just a technically perfect performance — it is a technically perfect performance, and the liveliness shines through. A unitary robot Alysa Liu would just never be as compelling. So I think nonfiction writing is quite similar, where I am trying to transmit ideas, but I'm also trying to evoke that there is a life behind the work. There's a very particular point of view with a set of footprints, flesh-and-blood experiences that I have, that don't have to be representative, don't have to be objective, but that I'm hoping people can connect with. And for me, it feels more authentic, it feels more real, it's more enjoyable.

Jackson: (1:58:49) Perhaps you kind of just answered the question, but which of your work — either what specifically, or what aspects of it — are you most proud of?

Jasmine: (1:58:59) I am really proud of the China piece. I think it's probably my favorite thing I've written. It felt structurally ambitious to try to weave together some of the family history stuff with a travelogue to China, with stuff about the Chinese tech ecosystem and Wang Huning and more abstract political stuff. I remember the moment where I figured out the structure, and that was super satisfying, because initially when I wrote it, it was just a day-by-day trip report of observations, which would have been fine. But it wasn't until I unlocked both the Wang Huning America Against America frame, plus this realization that different migrations from different members of my family — and so much of the AI story is a talent story, it's about why individual people decide to seek out opportunity for themselves — that the piece felt like it clicked. And it was also just important because it was emotional. My grandmother had just passed, and I was in China in part because I wanted to learn more about her life. I was going through hundreds of photographs with my mom and my grandfather, from her early life in Indonesia to when she moved to Fujian by herself, leaving behind her parents and eight siblings, to her college days. I think that personal motive — I wanted to write something that would honor the story — was another big part of it.

(2:00:38) Grab Bag: China, Silicon Valley's Virtues and Problems, AI Transition, The Relational Economy, Parties, Debates, Self Belief, and More

Jackson: (2:00:36) I think you did. I have a speed round of miscellaneous — we can race through a bunch of random things. A few things on China, actually, quickly. First of all, it's the same reason I love San Francisco: for all its thorns, China is a place where things actually happen. Chinese engineers also seem more practical than their American counterparts. They're here to build tech and make money. Risk management is for bureaucrats. Policy is only relevant insofar as it helps or hurts your work. This is something I think Westerners often get wrong. If you live in a single-party state, you are, on average, less ideological yourself. The politics have already been decided. No point wasting extra cycles coming up with something new. You're pointing at something I think is really underrated there. It made me think about people always wondering in the U.S., like, what must they be thinking about over there? It's like you're brought down one level of the decision tree. Can you say more about that — is there almost a sort of freedom that the average Chinese person gets from not having to think about this stuff?

Jasmine: (2:01:45) Yeah, it's really interesting because it's such a contrast to my own experience. I spend a lot of time thinking about political philosophy and how I want the world to be organized. And recently I've been thinking a lot about the AI backlash. So during my last trip to China, I was asking a lot of people — both AI engineers and my family members and random people — is there an AI backlash in China? Do people think about it? And it's similar: people would say the idea of an AI backlash is kind of silly, because AI and its diffusion is taken as a given. Especially because the Chinese government has decided that AI is part of their strategy for national advancement. And so there's not even a culture of resistance, or even the idea that one might oppose a new technology, because it is taken as a given. In the same way that you don't really spend a lot of time thinking about what alternative political systems might emerge if you've grown up in a society where there is only one.

Jackson: (2:02:44) So those two things are almost mirrored.

Jasmine: (2:02:46) Yeah. And of course, I'm generalizing. Of course there are dissidents in China. There are people who do critique AI. There are people who are doing underground, more active — like Americans.

Jackson: (2:02:59) We're gonna debate everything.

Jasmine: (2:03:00) Yeah.

Jackson: (2:03:01) We have an opinion on everything.

Jasmine: (2:03:02) Yes. And the majority of Chinese people just don't spend that much time thinking about it because there is a precedent where if you try to organize a protest, you're gonna get shut down immediately, you're gonna get thrown in jail. Why even try? What is the point of resisting? And there's a way you can look at this as being very bleak. And in many ways, I do think that sucks. I love pluralism. I think free speech is really important. I could never live in China because I value free speech and pluralism so highly. But I also don't think it's — there are still a lot of degrees of freedom that you have at a lower level, outside of the big questions of how is our society organized, and like, can you stop AI? Questions like: how will I use AI? I think a lot of Chinese people spend more time thinking about how AI can benefit them because they've accepted that it's a given in their lives. And then it's a question of, okay, do I want an elder care robot? Am I going to lose this job? What is the actual role that this thing is going to play in my life?

Jackson: (2:04:03) There's a world where, on some of these things, if they really are deterministic, they might be better prepared to think about the details of them.

Jasmine: (2:04:08) Yeah. I think the median urban Chinese person is probably doing a better job thinking about how AI can help them than the median urban American.

Jackson: (2:04:17) Right. Another one — there's a saying that goes something like: after one week in China, you feel like you could write a book; after one year, you think you could write an article; after ten years, you realize you know nothing. I think you express in that piece that you're sort of in the middle of that spectrum. But what's inside of that? Is it the history? Is it the fact that China is in many ways an ancient culture? Is it the vastness? The size?

Jasmine: (2:04:44) Yeah, I think all of those things. I mean, it's like the United States, right? I grew up in America — do I even understand America? I don't think so. I grew up in Seattle, I live in San Francisco, I've spent time in New York and D.C. These are extremely self-selecting pieces of what America is. And so I don't feel very capable of speaking for Americans. And I've never lived outside of these coastal elite hubs — to attempt to speak for China, when I go to Shanghai like once a year and I've been to a few other cities once or twice, just feels impossible. Plus there are language barriers. There are all of these things. I think it's that I want to afford China, as a country and as a society, the same level of respect for complexity that I afford the United States — which is something I think a lot of observers don't do.

Jackson: (2:05:41) To be fair, I don't know if most Americans afford the United States that complexity either. You've compared China and San Francisco — what, at its most virtuous, do you admire about San Francisco or Silicon Valley?

Jasmine: (2:05:53) I think it is the agency. People think it's cringe now, but I am such an agency believer. I studied sociology, I was very interested in social justice stuff as a teenager and a young adult. I worked at nonprofits, did college activism, ran in very lefty circles. And there's just this pervasive mood in a lot of those environments that you're screwed, everybody's screwed. No matter how hard you fight, capitalism and white supremacy and all these crazy superstructures are going to win in the end and they're going to ruin your life. Plus everyone's operating in a scarcity mindset and trying to cancel each other all the time. I remember even — media is a bit like this too — when I talked to journalists in undergrad and I was like, should I become a journalist? So many of them were like, "Nope, it's over. Journalism's over. We're all dying, we're all getting laid off, there's no future." And the thing I really loved about the tech industry and Silicon Valley is that there was always a sense, an insistence on there being a future. There's a delusional insistence on your ability as a random individual with no qualifications to shape that future. And there's a celebration of the people who do that. I thought that was amazing. It actually reminded me in many ways of reading books as a kid about civil rights leaders — reading about MLK Jr. or Yuri Kochiyama or whoever — where you do have to stand in the face of a superstructure that is greater and more powerful and wealthier than yourself, and rather than retreating into fatalism, you say, "Screw it, I'm an idealist, we're gonna make some shit happen anyway." I think Silicon Valley at its best embodies that kind of delusional idealism. And sometimes it is delusional, but you don't even get any of the good unless you actually believe it's possible. That's what saved me, in a way, from a life of being an extremely depressed and cynical leftist — meeting a bunch of people in Silicon Valley. In 2020, one of the formative experiences in my life was living in a group house for two months in Cancun with a bunch of Thiel Fellows. And they were the ones who convinced me that I should do writing and start a magazine, that I could be a journalist again. I had called all these journalists who were real journalists and they were like, "Nope, it's doomed, this industry's over." And it was these Thiel Fellow startup dropout kids who were like, "Wait, why can't you run the New York —"

Jackson: (2:08:17) Times, by the way? Probably don't care about journalism.

Jasmine: (2:08:19) They don't even care. They just believed in me believing in myself. And they would argue with me — I'd be like, nobody wants to fund journalism. And they would literally just sit on the couch and argue with me for hours. And then they'd introduce me to my first funders for my old print magazine. That sense of belief in different futures being possible is a thing that I love more than anything else. And so I don't know if I'm an optimist or pessimist about the tech industry. It's done a lot of good. There are things I don't love about it, but I really wish that other parts of other industries, other sectors, other people, other communities had the same sense of belief. It's almost this deterministic sense: if I set out the future, I know I can get there. And I think that's guided me quite a bit too. Just being able to set a goal — I want to design a social media platform from scratch and make everyone use it. You can't guarantee it, but you're sure as hell not going to do it if you don't think that you can.

Jackson: (2:09:21) Beautiful. What do you think this place is most undercalibrated on, or could use a dose of?

Jasmine: (2:09:32) I think it's insularity. Not being interested in, not respecting, not being curious about the rest of the world. There's such a strong belief that we can reinvent everything from scratch, that we can first-principles it, that we are smarter and we can figure out how to do government better, how to do —

Jackson: (2:09:53) It's the exact opposite side of the good coin. It's the exact same thing.

Jasmine: (2:09:55) I mean, agency is important, but I don't think agency and insularity have to go hand in hand. Agency doesn't mean that you're okay if everyone else hates you, or that you think you don't need to collaborate. And the attitudes that bother me in Silicon Valley often come from this anti-democratic, anti-pluralistic instinct where you don't think you have stuff to learn, you don't have to do the same thing as everyone else. But are you not even curious? The vice-signaling stuff — I think it is a virtue in and of itself to try to make people like you. It is a virtue to be pleasant. It is a virtue to try to communicate what you're doing to others and to hold out hope that they will understand and support what you're doing. At the point where Silicon Valley companies are no longer even interested in making that case to the public, they're basically saying, screw you, I don't care what you think, it's not important to have democratic buy-in for my project.

Jackson: (2:10:52) Yeah.

Jasmine: (2:10:53) And that's what really bothers me.

Jackson: (2:10:55) We talked around it a little bit, but one of the things you and I spoke about in the past — you made this point in your writing too — is that maybe AI can be a tool that empowers us. And you rightly make the point that the same tool that empowers the senior designer is going to automate the junior designer. And one thing you said to me really stuck, which was just: the point is to one-shot everything. How do you hold that view and feel that inevitability? You've made the point about getting AlphaGo'd. And that view is sort of driving all these people to lose their minds over the rat race — talking about the Chinese competition, all this stuff of just, I have two years before nothing matters.

Jasmine: (2:11:43) Right, right.

Jackson: (2:11:44) How are you sitting with that?

Jasmine: (2:11:46) I think it's really good that the world is this complex. I think we should all feel really fortunate that it is super hard to one-shot all the jobs and that we are really far away from it. There are all these lessons in management and organizational sociology — the org chart is not the org chart, the world is extremely illegible, the real world is all edge cases all the time.

Jackson: (2:12:11) Not that many things are software problems.

Jasmine: (2:12:13) Not that many things are software problems. Yes, there's standard operating procedure. But how often is that standard operating procedure actually followed? Even if it's 80% of the time, that 20% is a lot of edge cases. How many times do you do your podcast setup and some tiny thing goes wrong and you have to adjust because the lighting's different or something broke? The real world is all edge cases all the time. And so it's really hard to automate, actually. It's really hard to make everything legible. You can collect a lot of data, and despite all the internet data being sucked up into these systems, there's still so much tacit knowledge. There's so much that's never been written down. There's so much that is literally in people's hands, that's just a vibe that you cannot understand unless you're swimming in it and doing it. So on one hand, I think it is explicitly the mission of these companies to one-shot everything, to be able to do your job and my job in a single go. But I think humans are just a lot more general and flexible and complex — and the organizations we build even more so — than we give them credit for. Which is going to slow down the automation by quite a lot and give us time to adapt and to do new things. I am somebody who, if I got a UBI check someday and didn't have a job, I feel like I'd have something to do. But I think if you do that to people without them being ready for it, people are not going to spend their time well. I don't know what's going to happen then. But really what matters with automation is the pace at which it occurs and how much time we have to both adapt our jobs, and also adapt mentally and psychologically to the world we're living in.

Jackson: (2:13:58) Yeah. You have this line in the permanent underclass piece — I'll keep quoting it — quoting Carl Benedikt Frey: "Most economists will acknowledge that technological progress can cause some adjustment problems in the short run. What is rarely noted is that the short run can be a lifetime." You're getting feedback from the Times piece. If anyone should have a sense of what's happening across D.C. and San Francisco — do we have a chance of smoothing this curve?

Jasmine: (2:14:27) Yeah, I think we have a chance.

Jackson: (2:14:29) Or what's it going to take? I need all the solutions, all the answers right now.

Jasmine: (2:14:37) A lot of people working on this one. Part of the reason I wrote the piece is that even while economists disagree very strongly about the long term — some think there will be more jobs than ever, others think there's going to be a permanent underclass — there's actually quite a bit of agreement across the labs that this transition period is going to be pretty tough. Even in Sam Altman's more optimistic tweeting, he says we're going to augment people, not replace them, someday people will be more fulfilled than ever. The a16z post says the same thing. They all have a line that says there's going to be a painful transition. Every single one of these optimist posts says the transition will be hard, the change will be hard. And the reason I keep quoting that Frey line is I really want us to spend more time thinking collectively about how we will make that transition better. Think about the China shock and deindustrialization, where maybe a few million factory workers lost their jobs in the early 21st century in the U.S. Way more jobs were created somewhere else. But the fact that those factory workers lost their jobs and never got a good job again created the populist backlash that very arguably led to the deaths of despair and led to Donald Trump. And for those people, that is their entire lives. They might have spent decades developing a set of skills, going into debt to get educated, moving to a location to do one thing. If we end up one-shotting that job, it sucks. It's not their fault. And what are you going to do with it? I'm really interested in what the policy solutions are going to be. People talk about expanding unemployment insurance to give people more time to figure out the next thing — UI is six months in California, but maybe in a world where you have to reskill to something entirely new, you need a couple of years. People talk about subsidies for junior employees. Right now there's not a lot of incentive to even train a junior person. So through subsidies and adjustments to the tax system, you can give tax benefits to companies that hire young people, and that might make it easier for them to train that person into something and buy time. And I think in the long run, if we do start to see more automation, maybe we want to do things like give everyone a shorter work week instead of having some people employed and some people not. Maybe it's better if everyone just works a little less. I don't know.

Jackson: (2:16:54) You're not going to stop the people around here working less. The people with mythos are not going to be working less.

Jasmine: (2:16:59) Oh, no. I mean, that's one of the funniest things — they're all like, "Oh, I can't wait for the post-AGI future, we're all just doing leisure." I'm like, bro, you work a hundred-hour week.

Jackson: (2:17:05) Yes. I don't believe you.

Jasmine: (2:17:07) I don't believe you. Anyway, I'm not a policy wonk, but I think there could be a lot more attention paid to this. It's going to be really hard over the next ten to twenty years. That's a lot of people's lives, and we really need to focus on that.

Jackson: (2:17:21) Yeah. One of my favorite ideas you have is enjoying the sweet, sweet relational goods. Can you talk a little bit more about that world you imagine — one-on-one tutors for students, like, what does that look like?

Jasmine: (2:17:35) Yeah. So when people force me to do the thing where they're like, "Outline a good world" —

Jackson: (2:17:41) Right, yeah.

Jasmine: (2:17:42) I've talked about this with Jack Clark, I've talked about it with some other folks. The idea of the relational sector, relational goods, which has existed as a concept for a while. But Alex Imas, the economist, has been writing a lot about it recently. He says there's some class of goods and services that we prefer a human do, even if a machine could do it equally well — teachers, nannies, let's say. A robot teacher with an LLM in its brain could equally provide curriculum, grade homework, tell you good job, whatever. We might still, for various human reasons, prefer that a human is on the other side providing that service. People debate about how large the relational economy is and how many jobs there really are where we actually care if a human did it. How many jobs are there for artisan bowl makers versus Ikea machine-generated bowl makers? I don't know. But I do think there's a set of jobs in the care economy — one-on-one relationships, coaching, facilitation, teaching — where the majority of people will want another human in that role. And what Jack said was, a lot more people would be teachers if teaching paid more. I think this is true. I know lots of people who would love to be teachers if it paid —

Jackson: (2:19:03) — more. Or if you didn't have to. One of my best friends teaches ninth graders and he has a class of forty-five kids.

Jasmine: (2:19:07) Literally. It's crazy.

Jackson: (2:19:09) It's insanity.

Jasmine: (2:19:10) Insane, right? So one thing I think about is, okay, let's assume there are way fewer management consultants and investment bankers and software engineers, because we don't care if those jobs are done by people — but we really care that teaching is done by people. Maybe we can decrease classroom sizes and give everyone an Oxford-style tutor, where you have a one-on-one relationship with someone who is smarter than you in that field. They actually get to know you as a human. They can tune the curriculum. You build a relationship, human to human, separate from the purely academic one. And that probably makes you do the reading and learn more. A society where we use the windfalls from AI and find a way to tax and redistribute them — maybe in part towards funding ten times as many public school teachers as we have today and paying them all twice as much — that seems like a pretty amazing world to me. I think there are a lot of people who would be happy to live in that world. There are probably a lot of things I'm missing on the policy side, but it makes sense to me. We know nurses, teachers, therapists are so overworked, and if we could pay them all, train them all, do one-to-one matches — it seems like it would be a better world.

Jackson: (2:20:15) You've said everyone deserves to be bet on. Who has bet on you?

Jasmine: (2:20:22) I think one of the best things about going into an independent writing career is realizing how much mentorship I'd receive, which I could not have predicted, but which ended up being true. Maybe it's because writing is fairly precarious and there aren't a lot of people who take this kind of risk — so older and more senior writers are really eager to help young people figure it out. I felt really fortunate with a number of journalists and writers who have offered advice, introductions, mentorship. My apprenticeship for real, capital-J journalism was working with Kevin Roose on his big AI book for about eight months last year, two days a week. He just brought me in. We interviewed Demis Hassabis together. We would go to the OpenAI office and do the tag-team interviews. It was amazing because you get all of the tacit knowledge of someone who is at the top of their game, who's done this a million times. I got to participate, to see the conversation happen, to ask questions, sometimes to be in the room. There's no better way to learn the craft of journalism than spending many, many hours with someone who's really good at it. Gideon Lewis-Kraus has given me a lot of amazing advice, especially on approaching long-form work, bigger pieces. He's read drafts of things where I'm like, "I'm really scared this company is going to be mad at me — do you think they're going to yell at me?" And he'll talk me through that. Dan Wong has been amazing at talking me through the ropes too. Dan is the number one person in my life telling me to work harder and grind harder, which I actually really appreciate, because most people in my life are like, "Jasmine, take a break," and Dan's like, "No, never let up." I just felt really fortunate to have a lot of other writers decide — at basically little to no benefit to themselves — to take me under their wing and share what they know.

Jackson: (2:22:35) Why is the Peter and Valentine, Locke and Demosthenes subplot of Ender's Game meaningful to you?

Jasmine: (2:22:45) Oh, man. It's quite obvious, really. I read Ender's Game basically once a year from ages ten to twenty. The main plot is about this video game genius, Ender. And he has these two siblings, Peter and Valentine, who, while he's off accidentally committing alien genocide, are posing as bloggers on an internet forum — in the eighties, where there isn't really an internet — writing essays so good that they solve world peace. When I was a preteen, I too was blogging anonymously on the internet because I didn't want everyone to know I was twelve. And so the idea — I believed in this idea of meritocracy, this marketplace of ideas, the original vision of the internet where nobody knows you're a twelve-year-old, and if your ideas are good enough, which mine were not at age twelve, you could really change the world for the better. I just loved that arc. I thought about it all the time. Yeah, it just comes back to that.

Jackson: (2:23:52) What makes a good party?

Jasmine: (2:23:55) I think that in a good party, most people should know at least three other people, but not the majority of people in the room. And you want the social clusters to overlap but not be isolated. I have a network graph in my mind — I can't really paint a picture, but I feel like the sociology part of my brain has a network map. You want enough overlaps between all the different social clusters, because that gets different groups of people who don't know each other to talk to each other.

Jackson: (2:24:28) Classic.

Jasmine: (2:24:28) Yeah. But you don't want people who are so isolated that they never leave their current friend group. So one thing is the invite list and making sure you have something like that. I also think it's fun to have a very lightweight theme or activity that is optional — that people can partake in, but that doesn't dominate the party. Last year I hosted maybe five or so happy hours for writers in the Bay Area: magazine editors, journalists, bloggers, all sorts of people. They would often have a thing like, "bring a copy of one of your favorite books, put it on a table, write a note on it, and do a book swap." What that meant was it's not a very imposing activity — you can totally ignore the table and just drink with people and chat if you want. But if you meet someone and you have no idea who they are, you just say, "what book did you bring?" And I think it automatically builds in a way to form a relationship. Another one I did for my birthday last year: since it was my first year of being a journalist, my party was secrets-themed. I had a wall of sticky notes that said "secrets" on it, and everyone had to write a secret — a piece of celebrity gossip, personal gossip, anything — and stick it on the wall. So lots of my friends who do know each other can just chat, but if you don't know somebody, you say, "what was your secret?" It was great. And I still have the wall up.

Jackson: (2:25:43) What makes a good debate?

Jasmine: (2:25:48) A good debate is like a good game — both people have to want to be there and have to want to play the game for its own sake. Part of a debate is about clash. Debates are not about consensus, being really polite, and pretending that you actually agree with each other on a lot. Debate is about actually seeking the parts where you disagree and focusing on them, which is naturally uncomfortable for many people.

Jackson: (2:26:11) Yeah, yeah.

Jasmine: (2:26:13) And when you do a debate well, both people have decided that for the purpose of this game — whether it's putting on a performance or understanding the issue better — you're both willing to really drill in on that tension, on that disagreement, and to push hard for your own side. You actually don't want someone to say "nevermind, you're right" in the middle of a debate. The way it works best is when you have two fairly balanced sides of a controversial issue and both people are making the best case for their side. Then as an audience, you really get to weigh the best cases against each other. After the debate, of course, you can say, "actually, you totally convinced me" — we're good friends, whatever.

Jackson: (2:26:51) Keep the show alive.

Jasmine: (2:26:52) You've got to keep the show alive. I hate it when something's framed as a debate and people just end up agreeing the entire time. It's very annoying.

Jackson: (2:27:01) Do you think we're going to see many more of the sort of Free Press, Pirate Wires, Substack new-institution type thing? Obviously you're a little biased, you work there. But it seems to me more and more obvious — why would you ever be a writer, if you have any leverage, as part of anything bigger? Why wouldn't you just be independent?

Jasmine: (2:27:26) I think because a lot of people just don't like to be independent. This has been interesting for me to think about, because I probably — and you too, since you're independent — I love figuring out my own writing career as part of the game. I like trying to figure out how to pay my bills, trying to figure out, should I podcast, should I write long essays, should I write short things? I like being full stack. When you are a staff writer at a publication, oftentimes you write pieces because your editors tell you to — your editor hands you a piece. I like coming up with the pieces I like, and I don't want somebody else to hand me a piece. But some people like to be handed the idea. Or maybe the part they feel really, really good at is style, prose styling. There are other people who are amazing investigative reporters but hate the writing part, and their editor will basically rewrite the actual prose for them. I personally just happen to like full stack. But a lot of people don't. Or people don't want to go on podcasts. People don't want to think about distribution.

Jackson: (2:28:28) The world is definitely rewarding the person like you, though. It is the modern world.

Jasmine: (2:28:32) Yes, I think so.

Jackson: (2:28:33) Right?

Jasmine: (2:28:33) Yeah.

Jackson: (2:28:34) You wrote a post — I don't think this was the title, maybe the title was in poor taste — but the subtitle was "10 Alternatives to Having Good Taste." It reminded me of Nadia Asparouhova's piece called "Being Basic as a Virtue." Anything come to mind as additional points you'd make to the virtues of not having good taste, or being — I don't know. There was this excerpt I loved: "I don't know how to pick my preferences from a catalog of cool when they are so invariably not. I don't want to wipe these snapshot memories because they've since lost their shine. I am, as my taste, an unapologetic hoarder of every place I've ever been, every person I've ever met, everything I've ever loved."

Jasmine: (2:29:19) Yeah. I mean, it's kind of like the reason I write — I think everyone has a specific life and a voice. One thing I think about when people are writing: sometimes I read a writer and I feel like they are LARPing a writing style. They are trying to write like Didion, or they are trying to write like David Foster Wallace. And you can tell it's a poor fit for them. You can tell who they're trying to emulate, but they didn't have the experiences that Didion had or that Wallace had. It's like wearing your mom's high heels when you're a kid — you're trying something on that doesn't quite fit, rather than trying to develop your own voice. One of the things that annoys me about some of the taste discourse is that I think it drives a lot of people to acquire good taste as its own end, by kind of mimicking other people with good taste, or picking from a catalog: buy this coat, not that coat; eat at this restaurant, not that restaurant; watch this film, not that one. To the point where you are not attempting to develop your own point of view. That kind of taste is more a form of "I know the references" — cultural capital. And that kind of bugs me. I'd rather other people just figure out for themselves what they like.

Jackson: (2:30:37) You wrote somewhere that you wanted to be a professor when you were younger. Do you think any of that has persisted? Is any of that inside of what you do now?

Jasmine: (2:30:49) I really like research and I like teaching. I don't like the part where you spend a year on a paper that 10 people will read, and then you have to live in a tiny college town in some suburb.

Jackson: (2:31:03) Fair enough.

Jasmine: (2:31:04) A lot of professors want to be Substackers. This is one thing I've learned since being at Substack full time, because I still meet professors — I like my academic friends and I talk to them — and I meet so many professors who I can tell desperately want my job. But my job didn't exist before, so they had to get tenure and publish real articles.

Jackson: (2:31:24) That's got to be hard, to be like, "I was born at the wrong time." You're describing — this is very early on, I think, right when you left Substack. As for the project itself, the tongue-in-cheek tagline I've been using is Dwarkesh meets Ezra Klein, but a girl. How do you think that's going?

Jasmine: (2:31:48) So embarrassing that I wrote that. Pretty well, I'd say — except for the part where I don't podcast.

Jackson: (2:31:56) Yeah, okay. You said a solo newsletter and podcast on technology, politics, and culture — fewer podcasts these days.

Jasmine: (2:32:02) Yeah, I mean, I guessed more. I don't host, and both of them I think are very good interviewers, very good podcast hosts. But subject matter wise, roughly, yes. The thing that I really respect about Ezra is that he is really deep and really smart and really wonky, but he views his role as shaping a broad public discourse. He cares about accessibility. He writes for the New York Times because it does have a particular impact, and his theory of change has to do with choosing that over making millions of dollars doing a Substack. Then maybe subject matter wise, I cover AI, I cover Silicon Valley — I am more of the world that Dwarkesh embodies. He's very deep in the AI culture, he's known for reading all the papers and trying to understand the technical work. Pushing that hard at depth while also doing the public translation is something that I aspire towards.

Jackson: (2:33:01) What do you think keeps you earnest? It's not quite the same as optimistic.

Jasmine: (2:33:10) Yeah.

Jackson: (2:33:11) But I think it's something that's maybe even more respectable than pure optimism. I read you as very, very earnest.

Jasmine: (2:33:21) I think not being earnest is bad for the soul.

Jackson: (2:33:25) That's a great answer. I guess this is a little related — my last question. There's an old post, I think one of the first ones I could find, from 2020. The title is "Things that Don't Scale." And I think obviously the background is it's Covid and you're in school. You say: "When I think about the future, it feels vast and overwhelming. I don't have a job lined up post-graduation in 2021. Hell, I don't even have a job lined up for this fall. I refuse to attend Zoom classes either way. I want to figure things out, but there are too many unknowns and not enough processing power in my brain. Nothing about long-term planning makes sense anymore." What would you tell that Jasmine?

Jasmine: (2:34:19) I think she's thinking about it the right way. I'm really glad I didn't go to Zoom school. I'm glad I basically dropped out and decided to take a bet on the unknowns and figure it out, rather than forcing myself to do Zoom school just because I didn't know what else I was going to do. I did get a job shortly after — I just needed like a month more to figure it out. I think, going back to the earlier conversation about risk, most people have this idea that you have to know. When I quit my job at Substack, I didn't have any grants, I had literally no financial plan other than, "I guess I could lose money for a year and it wouldn't be the end of the world." Most people just need to take the leap first and trust that it's going to work out. There are more and less stupid ways to do that. But I'm really glad that my 2020 self, facing all of those unknowns, told herself it was just going to work out.

Jackson: (2:35:13) That's all I got. Thank you.

Jasmine: (2:35:14) Amazing. Thanks so much.

(2:35:16) Thanks Again to Notion

← Back to all episodes