How Every Builds a Writing Team in the Age of AI

[00:00] When I was the only editor working with our writers, I really used AI as a way to try to enforce our editorial standards. [00:09] What I was finding is the work just came back at a different level from each of them. We basically created a project where every draft before it came to me [00:19] If it was an internal writer, they would need to just like run it through our editor. We have a style guide. When I came here, it was like 400 rules. Hopefully by the time it gets to me, the sort of floor has been lifted. It's not about accepting what AI says blindly at all, but it's not generic and it's not random. It's trained on our stuff and it's trained on what's worked. And so your job as a writer or editor is to consider it. [00:44] *music* [00:58] Kate, welcome to the show. Thank you. Thanks for having me. I'm so excited to have you. One, because we've worked together for so long. You've been part of Every for so long. We've been working together on Every for like three and a half years. For people who don't know, Kate is our editor-in-chief. She does everything involving the newsletter and has been doing that for like a really, really long time and the business surrounding it to like sponsorships and events and courses and stuff like that. [01:28] Thank you. [01:28] Oh, man. You joined us at quite an interesting time because I think your first day was the day that Lex launched.

[01:36] And so it was like, like everything was like, uh, it was, uh, Lex is a, uh, an AI app that we incubated, like maybe, I guess, three and a half years ago, um, or three years ago, something like that. And, um. [01:48] It went super viral. Yeah. [01:50] And at the time we were really just known for being a newsletter and so we have this viral software app and then we spun that out. And so we went back to being a newsletter and we had to kind of like look at each other and be like, [02:03] what are we doing here and what do we want to be? Um, and so I want to talk about, I want to talk about you. I think, I think your career is incredibly interesting. I want to talk about, uh, how things have progressed, uh, for you and, and, and how you've seen every and what that journey has been. Cause we've honestly never talked about that. So just curious about it. And then I think the, the last like interesting thing on my mind is, um, um, [02:27] uh, [02:28] I feel like you have had this really interesting transition, which you just said, I think you framed it really well. Like, I'm sort of like an early adopter at child stuff. And you're sort of like a knowledge worker who's excited about this, but is not just going to use it because it's like cool. And I feel like you've had this like major transition in your workflow over the last like month or two. That even though we've been working on this for like three and a half years together, like, [02:54] and you have obviously adopted AI in your workflow, something has shifted for you. So I really want to talk about that. For sure. We've had a really interesting career that I think, [03:04] I think the kind of career path you have is the kind of career path that a lot of people at Every Have. And it's the best. It's like you've done so many things. Thank you.

[03:13] So my career is in media and technology. [03:18] And I think there's the version of like the how immediately did I get to Every, which, you know, you can kind of there's one version. And then like how did my career maybe set me up for that? But my background is actually in book publishing. [03:34] agent way back when. Kate was like a star agent. There's literally a New Yorker article from the front of the New Yorker being like, [03:42] Kate Lee is a star. Very kind. I was the subject of a talk of the town piece. [03:50] when I was just starting out in publishing and it really jump-started my career. It was extremely overwhelming at the time and I will just never forget it. Albreed? [04:00] Twenty... [04:05] 27? That's right. 27? Something like that. [04:09] I still have not been to talk of this. [04:13] Yeah, I mean, I guess I would recommend it. I don't know. But really wanted to ultimately, after 10 years as a literary agent working among [04:23] My colleagues were the best of the best of agents, representing the best of journalists, novelists, literary fiction writers. [04:33] big commercial nonfiction politicians, businessmen, et cetera, Nobel Prize winners, Pulitzer Prize winners. [04:40] Felt like ultimately having done it for 10 years, like I wanted to see what else was out there. It wasn't something that I necessarily wanted to do for the next 30 and really wanted to move into tech. And my, um,

[04:51] network at the time was just [04:54] you know, people who were founding things in New York, which was a much sort of smaller ecosystem than it is now. And I just knew I wanted to be a part of that. I didn't, not that I wanted to found a company, founding a company is the hardest thing anyone can do. Kids harder. Fair. That's, I can, I can say that too. I just, I just seen it. Appreciate it. You have a good [05:24] that world and through just like a whole series of sort of conversations and networking and meetings. Actually, one of them with... [05:32] ultimately ended up being your josh and hirsch so one of your very good friends from the browser company yeah this is a very long time ago um ended up meeting ev williams um when he had left twitter um and was founding medium and so i jumped to medium and and um was there for four and a half years and you led all their i led i was their first person in new york and their first like sort of content head and it was just like the i just felt like such a sort of seamlessness product [06:02] I loved about working with writers in just like a new way, a faster way, much faster feedback. And, and of course, it was the first time I had worked with any real technology people. I mean, the technology people I'd always worked with was an ID department. This was, you know, product managers, things like that. And then, um, [06:20] Medium went through a number of changes as as as most companies have and most media businesses have.

[06:27] And from there, I actually went to WeWork. Very nice way of saying it. [06:32] I don't know. Do you want me to go into more of it? I'm happy to share what happened. I think we all know what happened in 2016 in terms of to the larger environment. But I actually very explicitly in my next move was like, [06:46] I'm really exhausted from media. I've just like I've been in it for a really long time. And I was just in a really early stage company. And I don't like I just need to go somewhere less crazy. So I'll go to WIQ. That's literally, you know, one can question my judgment. But the company had just raised the I don't know, however many billions from SoftBank. And it was just they were just in huge expansion mode. [07:16] for almost two years, and it was like true madness. [07:21] you know grew from 4 000 to 14 000 people just like crazy crazy like three different people doing the same job because we were being hired so fast so that was um that was a really you know different experience and and like i worked with amazing people of course all these companies have such talented people [07:39] Um, but ultimately knew that like, that was just not a place I was going to stay. Um, and then I saw a job opening, a job opportunity, which was to be the publisher of Stripe Press, which is Stripe's, um, publishing arm. And I just was like, this is such an incredible opportunity. And so I applied for the job and talked to a bunch of people there and then started there actually about six weeks before the pandemic hit.

[08:09] such talented people. I think any job in general that you're going to start right before pandemic is just like, that's just hard. And I think they absolutely did everything you could want from any company in that regard. So I was a publisher of Stripe Press, which had launched sometime before. So I was taking over an existing list and then building a new list. And then I also ended up becoming overseeing Increment, which was their magazine for developers. I didn't know that. [08:39] You know, increment run by a woman named Sid Orlando, fantastic magazine print, also print, print and online publication devoted to developers, which was Stripe's really early and core audience. And I learned so much at Stripe and just one of them being like. [08:57] You know, Stripe is a payments and financial infrastructure company. [09:02] And so people would be like, why publishing? You know, I have a publisher and, um, [09:08] First of all, I would say if I have had many people and, you know, founders or VCs or companies being like, can we do a, you know, can we do a book publishing career on Mike Stripe? And I'm like, no. And I shouldn't sound that categorical. But I think what the among the unique ingredients for success at Stripe are. [09:28] founders who are genuinely, um, [09:31] genuinely, voraciously... [09:34] you know, excited about ideas and books and that that being set from the very top really sets that up for success in a different way than if it was an initiative that came out of a different team.

[09:46] Um, and I think it was like a pure marketing. Correct. It wasn't like, you know, coming up in a marketing brainstorm. This was like, this was a genuine desire to. [09:56] both read great books, to rediscover books that have been, you know, potentially lost to history, to discover new ideas, and to really help entrepreneurs. Um, the, the mission of Stripe Press is to, um, is called Ideas for Progress, and it really connects to Stripe's largest mission, which is, um, enabling more global entrepreneurship. Um, and, um, and again, that is done through various ways, through various products, but like that is that overall mission. And so, um, that [10:26] that was certainly that was a fantastic experience and I should say Stripe is probably anyone who's worked with me knows that I'm like can be really I don't know like particular about certain things and never would have copy and commas and all things like that and Stripe is a place where the attention to detail was [10:48] absolutely incredible and you know was what they're sort of known for and it was almost a little bit counterintuitive because you know if you think about it of like it's a financial infrastructure company it's a payments company why do they need to have a beautiful gradient on their home page but they do and they invest a ton in making sure it is just right and so even if it's not meant to be a household name necessarily or it wasn't at that moment in time [11:14] It didn't matter because the people who loved them loved them. And that relationship was really important. That's really interesting.

[11:22] So, yeah. And then while I was at Stripe, I met you. I basically was like, I got to find writers. I got to find people to write some books. And every had just started. This was when... [11:31] I guess it was 2020. It was like early pandemic. Everyone was starting newsletters. And I had met your co-founder, Nathan, a number of years ago when I was at Medium. [11:42] You have, I believe. [11:44] I had met him probably with Evan. And actually, I think when I met him, he came in with Hamish. That's so funny. Hamish is one of the co-founders of Substack. Yeah. And so I just reached out to Nathan and I was like, I'm at Stripe now and would love to talk. And he connected us and we had an initial phone call and just stayed in touch. And then I ultimately decided after two years to leave Stripe and go independent and get back to working with really early stage companies. And so I reached out to I guess I reached out to both of you again. [12:14] Do you need any help? [12:16] And we're like, yes. I mean, yes. And we were like, boy, do we. There was like this running joke for. [12:24] basically from 2020 until when you join. So like three, three years or two years of every year or whatever. That was like. [12:32] editor-in-chief at every is like defense against the dark arts at hot orcs like no one lasted very long [12:38] God, I didn't know that beforehand. I really am upset. Tony does at that point. I know, but that exact framing. I definitely knew that exact framing. I definitely knew there had been a number of people. You tried different configurations of that role. It was just really hard. It was and it was not necessarily the fault of any of the people that we hired. It just was like.

[12:56] we do a very particular kind of writing. And I think early on, I was like, well, if you have editing experience at a big company like New York Times, you must be able to do the kind of stuff that we do. I didn't realize different editors are just made differently and have different jobs. And editor is a very wide ranging title. That's the same thing as founder. You know, it's like, [13:17] What kind of founder that you're, do you have a barbershop or do you have like a venture back company? Like, they're very different. Um, and finding someone who has. [13:27] the [13:29] writing taste. [13:32] for... [13:33] technical stuff or stuff [13:36] in and around technology is very hard. I think that you're either like really into tech, [13:41] or you're really into writing and you're, [13:44] It's not always, but a lot of times you're a little skeptical of tech and you're coming at it from that perspective, which a lot of more mainstream media outlets come at it from, which is an important perspective, but is very different from our coming from the builder's place. [14:00] Like, I'm pretty sure the minute, like the first time we met, like the first meeting, I was like, do you want to come on full time? Or if I didn't say it, I was thinking it. And I spent like a lot of time trying to get you to come full time. And then you did. Yeah. That's been really, really, really amazing. [14:15] Yeah, it was the right time. I mean, I basically, I think I immediately started like, you know, just sort of advising, you know, you and the company and the team and then ultimately started doing some freelance editing because you needed editing help. And so I started doing that. And then I took on.

[14:37] the, um, the guest contributors. Right. Um, and I was also really aware, I basically had like three main clients at the time. And I really loved having kind of like the ability to pick the projects I was working on, the people I was working on, set my own schedule. And I just was very aware that like, I always liked working with this team the most. Um, and so then after, um, after I was spending time on, you know, guest pieces and making those great, [15:07] was like you, Nathan, Evan. So the, the, the main writers, the sort of full-time writers at the time who you all had a very, uh, seemed to have a very symbiotic like process of giving each other feedback and writing together. But I was basically like, look, we're going to do, put all this work into these. We're also going to do a little bit more work. And that was sort of the last frontier of just like what I was actually editing. And then with, as Dan alluded to, or as you alluded [15:37] know, not only just an editor in chief, which you definitely needed, but like, what are we doing in this company? There was an opportunity to come in and be like, let's figure out what this is, which is when I came in full time. Yeah. [15:49] And I've always wondered this. [15:52] because I have never actually worked at a great company. I have worked at companies and I'm not casting aspersion of those companies, but by great, I mean, [16:04] a really high performing startup or company that used to be a really high performing startup. Um, I've only just like kind of dinked around on my own stuff more or less, um, or worked at an acquiring company or whatever. Um,

[16:16] But you've had a lot of experiences with [16:19] some of the big name companies of the last like 10 or 15 years, like EDM and WeWork and Stripe and every. [16:29] And I feel honored to be, you know, that that's you've added that to your resume after all those. But I'm curious. [16:38] how your experience here compares or like what you've taken from all those different things that [16:44] and how that relates to every. Yeah. I think in some, to some degree... [16:49] My own career experience has been like every every job or every every experience has been a sort of reaction to what has come before. And so, you know, when I left book publishing, I was like, I want something fast and, you know, nimble and and just immediate feedback just because that's the you know, the book business just works on much longer timelines. And I was really, you know, eager for that. [17:11] Um, and then... [17:12] I basically was like, oh, I need a much more established company when we were. And then leaving WeWork, which was a disaster as a company for many reasons. Stripe is an incredible company and operationally is incredible and also grew enormously while I was there. [17:42] knew that I wanted to get back to in, well, first of all, I should say,

[17:48] after I left, you know, when I left Stripe and I was doing projects of my own and I was like, I love the sort of like control. And I was like, I'm never joining another startup and I'm never joining another media company ever again. Oops. Never the never. You know, famous lost words. But I really felt that at the time that that's like I wanted to, you know, that's how I wanted to do my career. And I think ultimately what I find I get the most satisfaction out of is like, I personally just like working in small teams. [18:16] better. Um, [18:19] or just find that more personally satisfying. And I also really like working, as I like to say, it sort of sounds, doesn't sound like much, but like, I like to work on the thing that is the thing. [18:30] which basically means like, I want to work on the thing that is the main thrust of the company. And even if, for instance, media is not the most remunerative, let's say, or, you know, as, as let's say as a, [18:43] software product or something. It's the thing that is the driving heart of the company. And so what I am doing is what is contributing directly or not contributing to the success of the company in a way that feels for me much more direct and satisfying than being. [18:59] Doing marketing, even very good content marketing in some other capacity for, you know, for a much bigger company. I want to get into this moment that I feel you've had because I think of you, I think that you can be a real person. [19:15] maybe ambassador is the word for particularly people out there who maybe have some resonance in their career trajectories to you, or they're like, I'm in book publishing, or I was in book publishing, or I'm an editor, or I'm a writer, or a lot of the

[19:34] a lot of the strands of experience that you have that are incredibly important at every, but if you're from not from our world or not working at every, you might look at a lot of what we do or a lot of the AI stuff. It's like, holy shit, this is like crazy. I don't know if I like this and I don't know how this might. [19:51] how I can really [19:54] incorporate this into my work or whatever. So I just, I kind of, I'm curious about, um, [20:00] I mean, we can talk about the breakthrough that you had recently, but also maybe more generally how it has. [20:06] why you went in to do this at all and how that has changed over the last couple of years for you, like incorporating AI into your work and then where that is now for you. Yeah. [20:20] So... [20:20] It's interesting. So when, you know, three, three and a half years ago, whenever it was, when I joined the company full time, I think I was the fourth person. [20:26] fourth or fifth full-time employee yeah it was like i mean uh i mean at the time that you joined full-time there was like three people full-time including me yeah now there's 20 yeah um and it will be probably 23 or something soon so yeah it's got a lot bigger it's it's grown a lot certainly percentage-wise um we have grown an enormous amount and i think when i so when i was just starting here and thinking about you know [20:52] what I wanted my career to be. [20:55] it's kind of a little, maybe a little counterintuitive that I was like, Oh, like joining the small startup. I sort of like, I, there was a, there was an aspect of it that I was like, this is a world I really know. And also can sort of really, it will really work with my life, which at the time I just had a second kid and needed startups known for great work by balance. I was like, I'll turn startup now. A little bit, a little bit of strange timing potentially. But I was like, Hey, whatever you want to do, I don't care.

[21:25] A tremendous amount of flexibility, which I am always grateful for. But I think I knew coming in, I was like, I know, I really know what my zone of expertise is. And I really know that, again, there were only three other people at the company. So it wasn't like, oh, there's 50 other people. I was like, no one else can do what I'm doing. I just knew it. And so I felt like I had just a really clear sense of like, I'm going to come in and I know what it takes to get my work done every day. And it was very much going back to being an IC, an individual contributor. [21:55] I was like, nope, I'm just going to come in and do it all myself. And like, I can control everything. [22:00] what I can do. And I think you wanted that. You really wanted that. Yeah. Yeah. I was like, very happy to be like, this is what I'm doing every day. Which I think most people who have that, they've like managed large teams. It's like, [22:11] a little bit of an ego hit to like go back to being an IC rather than being like, this is a refreshing like way to just like get back into the basics of what I do, you know? Yes. It's really interesting. Yeah. It felt really satisfying. And I think because I'd had, you know, I ended up [22:26] I think I was sort of consulting for about a year and a half before I came here full time, I basically threw other clients as well was like, well, I guess it's just me to do the work. Like I have to figure it out. So I feel like I was really able to get inculcated that way. But, you know, when when coming into every at the time and again, having started full time, right, literally when we were launching Lex, like it was both clear that there is a new technology happening,

[22:56] we were going back to being a news, a newsletter company and really having to figure out like, what did that mean and what, what made us different and differentiated us. And I think there's always been, you know, the focus on Hickolity writing has always been at like the beating heart of the company and, and was from the very beginning. And that's been really, really important, but it's been really interesting from the, from the AI perspective to be here, [23:26] It's just, I know that when GPT-3 came out, you were just really excited about it. I was like, Kate, I can automate your job. Well, we can talk about how you've been trying to automate my job for the past three years, and I don't, yeah. But no, you just really wanted to write about it, right? You were like, I just really want to, like, I'm just going to spend the next three months writing about it. [23:47] Chachi Buti. Only Chachi Buti. Okay. And that was like that fall. And you did. And that's when we ended up launching the podcast. And it just was so clear that there was like such a richness there of things to talk about. I don't. [23:59] know if it was clear to you then what was going to happen to like actually building products and how AI would impact. It was on my mind. Yeah, definitely. It was on my mind. And like, I have, you know, an email to our investors, like, [24:11] around the around the time when when Lex happened, like being like, this is the we're going to do more of this. This is the plan. Right. But also the practicalities of we're running a business. We have to publish every day. I'm writing a lot like it's moving super fast.

[24:27] It took a little bit for that to actually like... [24:31] play out. Right. One of the things that I think is I love about this business is I was just like, I'm pretty happy doing that. Like, that's actually great for me. I love writing, you know, so I wasn't like, [24:42] upset to be in the news, in the, in the, just deep in the writing newsletter stage, you know what I mean? Yeah. Right. And I, I think one of the really important lessons of Lex was like, [24:53] yes, technology-wise, the ability for two people to build this product who, you know, weren't necessarily hardcore engineers was really meaningful. [25:04] And the fact that we were able to distribute it to a list and get that was a huge deal and sort of having that like distribution built in, which has just obviously or not obviously, but like has distribution is so, so, so important. And so really maintaining like we have to sort of cultivate this list. Yeah. But I think, you know, I can say when I started. [25:25] full-time, um... [25:28] We had very lofty goals. And I remember those goals and like metrics. And then let's see what happened. Elon Musk took over Twitter. Oh, my God. That was big. That was bad. Traffic cratered. And it was like, oh, wow. We have to figure out what this is going to be. Are you remembering that moment? It's all going back to me. I'm like sweating a little bit. That was so many crises ago. It was so many crises ago.

[25:58] this isn't the trend that we were seeing before yeah the link sharing it just like they did decaritize links and it was like oh that's our whole business it was really really tough so do you remember what our strategy was it was um [26:11] Emphasize bottom of funnel. [26:13] Yes. And that we started doing a lot of more courses. So it's like, if we're not going to grow Top of Funnel as much, we're going to... [26:22] emphasize the relationships we have with our best subscribers. We started doing courses and that was also a big impetus to start doing more software because it was like, [26:31] we can monetize even if we have a smaller number of people we can monetize them better with with software and consulting right yeah right no it directly led i mean that is what what directly led to the you know the way that area is now along with of course the changes in technology such that like now you can we can run you know four or five products with with one jam each which just like would not have been possible um a couple of years ago so it's been real so so in terms of like your [27:00] felt [27:01] being both a part of this journey and AI, et cetera. I mean, I always like, [27:07] I always have felt like media and technology has been like the real nexus of my career. Um, um, [27:14] You know, I, AI is certainly, you know, [27:18] It has felt like a huge sea change in a way that probably like people who were in early mobile, you know, were experienced it as well. But what did you feel? What did you feel the first time you saw like the words come out of GP3? What was your reaction? I'm trying to remember. I think I was I I don't I'm trying to remember where it was if I can. And I don't know if I remember the actual first time.

[27:48] Yes. You made it so we could make a prompt that would do your copy edits like three years ago. Why did you feel that? [27:55] So, so Dan has been trying to automate my job for, for like three plus years and you're getting closer. It's slowly but surely. You have, you have really been like, there is, you know, because what. [28:09] an aspect of what I do of editing is like repeated patterns, things I'm sort of catching all the time, a certain standard we want to hit and rules to follow that is like very well suited to AI. I don't can't tell you how many times you've deleted the, um, the word very from my drafts. And like, I'm just trying to save, I'm just trying to save your brain. [28:31] Um, correct. Um, and so I think like initially, I don't know that I felt, I don't think I felt the [28:39] wonder that you felt. I think it was more like, wow, this is really interesting. And I think in general, the way that I have watched it has been like, [28:49] I feel like I get to have a front row seat to everything that's happening. And I'm so grateful for that. I am riding along with it and, you know, trying to, I basically feel like I get to learn every day from you and from colleagues about what's happening because the attitude very much is, you know, as you said, it's not like you have to try these tools. It's you get to try these tools. Like you work whenever you get to try these things. By the same token, I think, you know,

[29:19] Like you're the early adopter, very much on the edge, pushing forward to the nth degree of what these tools can do. And I'm really that like knowledge worker who works in tech or is savvy about tech, who wants to use AI, knows I need to use AI. But like I also have been doing a job a certain way that has not been reliant on AI. [29:49] versions of a copy editor, it's been like, well, I don't still fully don't have to use this because it's not nothing is as perfect as what I have been able to do myself. Yeah. And is just going to in the short term, at least is just going to create more work. And then in the past couple of months, again, I'm not about to say that you motivated my job. But in the past couple of months, you [30:11] I have seen and I have been able to use AI to do aspects of my job that, again, maybe aren't fully yet along the editing side, but are in, you know, absolutely administrative, operational research reports, anything like that. [30:41] to the settings and figure this out myself. I can just tell an agent to do it. And that, um, [30:49] Time is precious and I try to accomplish a lot every day and I don't want to be spending time wrestling with software. Can you say one or two specific concrete big light bulbs for you? I think one of the light bulbs was around Notion and hiring. I had opened two or three headcount positions earlier this year. We don't have an HR department. We're doing all the hiring ourselves.

[31:19] efficient way to just, um, manage the, uh, manage the whole process. We use notion for our, our, not just the job descriptions for, but for the applications and then how do we filter them and how do we post them and how do we, how do we do all these things? Um, and I was really dreading it. I really was like, this is just going to be such a manual process. And, and, um, it was just going to be hours of, of, of, of sifting and going through things. And what we did was, um, I think [31:49] list was the first was, um, [31:52] OpenAI's browser, basically you were like, just have it tell Notion to like, [31:58] first of all post the job because notion is not my friend i maybe i shouldn't say that actually [32:09] notion first of all i mean maybe at some point in this in this conversation like google docs is just the thing that i have lived in for years and as as for you know there's a reason that google docs is still [32:22] has the privacy that it has. It is so easy and so good. But really found that like, if I was going to be having to spend an afternoon figuring out how to replicate a Notion page and get it up and get it in our database, like this was just going to be [32:36] a huge waste of time and you were like just tell atlas to do it and i was like what do you mean and i like sat and i am that person who needs to sit next to you and you're like open the computer type in it and then and you're like just type that and you're very good you don't do it for me you you're saying like type in this and then i know what i'm doing you you lead me you lead me to water and um and i did and it was like

[32:59] I mean, I watched it think and I watched it like it needed access to a few things, but you were just like, and then if it doesn't work, I sort of assume that if something isn't working, that it's my fault always like, oh, I did something wrong with technology. And you're like. [33:14] oh no, just ask it again or say you did it wrong, do it again. And I was like, oh my God, you could talk to us and just be like, you were wrong. Try again. And it really, it made that process so much more seamless, both like just getting the jobs up, which again, probably sounds super easy and mechanical, but like was, was just taking me a little bit of time. And then, you know, very grateful to have hundreds of applicants. And I, you know, needed some ways [33:44] and understand, you know, what I wanted to be looking for. I will say I did end up going through almost all of, I did actually end up looking through all the applications myself, but it was very different being able to be like, okay, these are sort of 10 flags of things I should look for, not just like, okay, I have to go through 100 and I'm really going to spend time on each of them. It's a first pass. It just gave me a first pass. And then it gave me a way to sort of evaluate, like here is everyone I'm looking at and I can evaluate it against this. [34:14] And it, you know, made it, [34:17] possible to hire people while at the same time doing my job. And I think that's, that is a pain point for startups, um, where we all hire our own people, but you have hiring is so important and you never really have the time to spend on it that you wish you did. This should show you something about like the way my brain works and how much experience or lack thereof I've had at big companies. Like I really had not thought about the fact that HR is the one that recruits people.

[34:46] and like i worked at a big enough company and hired a team with a company that i probably i must have done that in the past but i i didn't really think i didn't just didn't think about that yeah um and yeah there's always that thing where you're like i can either do the job or hire the person to do the job which is a there's a very similar ai thing here yeah where it's like i can either do it the way i know how or i can like potentially waste time trying to figure out [35:16] And yeah, it's been really interesting. [35:20] seeing that light bulb for you because i actually have a very similar light bulb with [35:24] agent browsers, [35:25] in particular around just like settings panels. [35:28] I'm just like not good at those. Interesting. And I'm get asked all the time for like, can you have me to this role? Or can you do this thing? And I'm just like, really? I'm going to spend like 15 minutes of my day, like transitioning from whatever else I was thinking about to like the this console dashboard thing. And it's like not intuitive. And it's like, Alex just does it. I'm like, great. This is fun. Yeah. Yeah. And it's very actually freeing for me to see that, like, you're like, I'm not going to I don't know what the settings thing does. I'm just asking that. [35:58] Like I am someone who I clearly I know myself. I obviously learn by by doing and I also learn by like modeling. So like it's important for me to I'm not necessarily going to be like I can think of the five use cases by myself right here. But I will be able to be like, oh, we're using for editing and we're using it. How's the company using it? We're using for operations and we're using it for that. How do you use it for writing and editing currently?

[36:22] Great question. So I should say that as a team, we have ways of using it. And then my usage has shifted as... [36:34] I have actually been able to have more of a team. So when it was, um, when I was the only editor, um, and working with our writers and, and maybe freelance editors, um, I really used AI as a way to try to enforce our editorial standards, um, more, more consistently across the board. What I was finding is I worked with, um, super talented freelance editors, but, but like, but sort of the, [37:04] came back at a different level from each of them just because they were different people and you know in very different backgrounds and not all sort of inculcated in the in the company and i was having to just spend a lot of time sort of get each draft to a certain level and um it turns out that again ai if you can you can tell it your your patterns and your roles we basically created um i don't know if it was initially a chat then it was a project it's and then it was an artifact basically [37:34] thing came out but basically a project where we um we for for every draft before it came to me if it was an internal writer they would need to just like run it run it through our editor which is you know we have a style guide I'd written a style guide when I came here it was like 400 rules and um you know have it have it essentially transfer to this to this um project and then

[38:04] have the writers, uh, use you and, and editors use this, this guide such that hopefully by the time it gets to me, the sort of floor has been, uh, [38:15] lifted. Yeah. [38:17] As we have... [38:20] been able to to build up a little bit more of a team and infrastructure i actually do that very little now but the editors who are on the team um and the writers do that themselves um so that's that is one key way um and then uh [38:37] I mean, and then there's lots of other ways, I guess. Well, I guess I'm curious if... [38:43] you know, someone is listening to this or watching this and they are, [38:48] Maybe they're an IC and they're like, I need to apply this to my writing and editing. Or maybe they want their IC now, but they want to manage like teams of writers and editors in the future. Or maybe they're currently managing teams of writers and editors in the future. Like, [39:02] What do you think are the skills required that might be different from... [39:07] what you might have learned in English class or journalism school. And how are you setting up the editorial arm of Every differently now than you might have before? And what does that look like? [39:20] It's not really a newsroom, but what does a writer's room look like in this age that is different? Well, I think we... [39:31] You know, we publish a newsletter daily and we are a [39:34] pretty small team. Um, the team consists of myself, managing editor, um, a senior editor writer, and then, um,

[39:44] another writer who also does our, um, who does a lot with, with AI. And so, um, we all have to do a lot. And I think one thing that has been, you know, really clear is that a, like, we need to know what are the standards that we're trying to enforce. And, um, [40:04] We have to, the thing is that if you are going to do that with AI, you have to sort of, [40:09] um you have to be the one to teach the ai and you have to be the one to enforce it with ai and so i think something that we've really learned over the past months is like actually how let's say claude can read a style guide is maybe not in the lane is like not in like how i would normally write a style guide but actually just needs to be like structured slightly differently to actually get the work done the way we want it to there's been an incredible amount of trial and error of like okay it might not work this way the first time it needs to work this way the second time we [40:39] our writers, again, the writers who are here, not necessarily freelance writers who are involved in our day-to-day, to also be using these systems themselves. They're going to work with an editor, but they should know that there are both things that we are looking for from the perspective of, again, our style, but also what makes a good every piece. And having a really good sense,

[41:09] what makes a great every piece? You know, is it from a, you know, first person perspective? Is it from, you know, sharing expertise that has been earned, um, all sorts of things. And, um, [41:20] So, yeah. [41:21] They, they, I think there's been a lot of like in, in, in, [41:26] newsrooms or traditional newsrooms it's like you're a reporter you're an editor and you're and the the you know [41:33] A reporter is never writing their own headlines. We are testing headlines. We test them with AI. We use AI to generate lots of ideas and then riff on them from there. I, again, not necessarily for writers who are not full time with us or involved with us on a daily basis, but like that is something we work with together as a team. [42:03] but then you have to make sure people use them. Yeah. [42:08] There are different things. That's really different. Yeah. And sort of being the person who's like, okay, okay. [42:15] you know, can you confirm that like you use this tool before you send it to me? Because I don't want to you know, I don't want to be looking at the same sort of changes that I always want to make. [42:25] I think another thing that we have done is each week at our editorial meeting and we've actually done. [42:32] done this a little bit less now more recently, but particularly when we just had, you know, we're sort of getting up to speed with this and had newer people coming on board, we would go through every piece we publish that we we would look at the this the head, the subject.

[42:46] the, the, you know, the headline, the deck, the subtitle, and the lead, the introduction, and be like, let's, let's break it down. And, you know, we, you know, email is a primary delivery system for how we distribute our content. And so what that subject line is, and, you know, what that first paragraph is, that's going to keep you reading, are just really, really important. [43:16] um [43:18] you know, head, deck and lead of each piece and be like, was this good? Could it have been better? Did it work? Did it not work? And basically feed fed all of that feedback into our cloud project to be like, this is what we're looking for going forward. So we're really trying to train. [43:33] As I like to say, like... [43:35] You don't have to. It's not about accepting what AI says blindly at all, but it's like this is not it's not generic and it's not random. It's trained on our stuff and it's trained on what's worked. And so your job as a writer or editor is to. [43:52] Consider it. Yeah. And to again, not that the expectation is you have to accept it, but it's basically a stand in for another editor. And so you have to consider it. And if you don't think this suggestion is good. Yeah, that's your prerogative. But like you, you need to wrestle with each of them. [44:10] Totally. A way to summarize some of that is... [44:14] There is a way to use it where you're constantly pushing the best of your taste and your learnings into written artifacts that the AI can consume and that that can be used by writers and editors at the moment when they might need it. Not as a never think for yourself tool or as a like just push out spam tool, but like actually just as a.

[44:40] here's the best of what we know from the past and what we've been able to write down. Like, does this help? Because a lot of times you're making repeated mistakes that it's like really helpful to have a little helpful friend that's like, hey, like Kate would probably say this, you know, without you having to do it. And I think that's really, really cool. And watching that kind of like blossom internally inside of the editorial organization is really cool. One thing that one experience I had recently that I think is is cool that we haven't really had a chance to talk to you, but we do these vibe checks. [45:10] And so new models come out and, um, [45:14] we get them before they come out and we often have very little notice and so it's a scramble to do that and recently we had two models come out at the same time Codex 5.3 and Opus 4.6 and we were like [45:27] testing them just super intensely for like a week and then had like a very short amount of notice for like we had to write the vibe checks and one of the things that we do when we test these things that we just have a discord channel where everyone's throwing their thoughts in and i just had co-work like go in and summarize all that stuff into a notion doc that it was like constantly updating and then i was able to like kind of vibe code that into a web into a website for the vibe

[45:57] to one of these companies and it caused it to like fact check. It wreaked havoc. Unintentionally. It was like at 2 a.m. Here's our initial testing. I just want to give you a heads up. It was very funny. Sorry about that. If you're listening, I will not say who it was, but sorry about that. [46:16] But [46:16] Yeah, I think we can go a lot faster and there's a lot more... [46:21] There's a lot more like multimedia aspect of this. Like, you know, I was, you know, I was looking at, I was looking at the first article I ever published about AI, which is AI in the age of the individual 20. Yeah. 2022. November 2022. And this was like a big deal. So first of all, I think this is actually a really great articulation of exactly what I feel like right now. [46:51] So I'm very happy about the call. And like, if you look at this, it's like, I don't like, uh, uh, you're going to need vision, taste, ability to prioritize. It's like, that's pretty, that's pretty good. Um, but the, the more, the more important part of this is like, um, [47:08] uh, [47:10] Having these kind of images was a big deal. And look, it has the prompt in it. A man with four arms typing on four typers at once hyper-realistic. That was a whole thing for a long time. And now that's just the old hat. I think there's something about the new expectations for publishing that is...

[47:28] not just about the processes, but also what you can do. [47:33] What you can produce at the small team. Yeah. [47:36] And in what time frame. Yeah. And I think there was a whole era, like the Gawker era, where it was like, every writer has to publish once a day or whatever. And it was like kind of... Once a day? My gosh. It was like... Maybe a couple times a day. Yeah. And it was like shitty stuff, basically. Yeah. [47:52] And, [47:53] I think that... [47:55] Obviously, we publish once a day, but... [47:59] Um, [48:00] I think you can raise your standards for... [48:03] what is possible for a single person to do well in a short amount of time, maybe not in like the gawker sense, but just in a like you can if you're the right kind of person and you've you've spent time with this, you can like write the article, make a really great image to go with it, vibe code a website for it and then like do a little video. And like that's all possible now in a way that it wasn't in the way and then the way that organizations support those kinds of people is different. And that's also really interesting. [48:31] yeah i think i think as a company we all really felt step change like between late december and then early january and um you know we've written about this and we've talked about it with um opus four five and um and cloud code and co-work um but i will say that like i feel like we as a company have been sprinting from january 2nd or whatever day we came back and just like an absolute sprint

[49:01] like we're just doing more. [49:04] You know, even this vibe check that you're talking about, you know, to be able to do, [49:09] do all of that for two major models so vibe checks are our version of breaking news like that's it that's like you know breaking news the newsroom is you know on fire so to speak that's us model is coming we know the embargo we've got a plan for it we know what we have to pull together and it's not always easy we have a far-flung team with really different perspectives and most of whom are not professional writers and aren't writing is not their full-time job so you're [49:38] how we get good reviews because it's like people that use this stuff for their job but also it's like [49:43] You have a job. You have a job. And now we're going to be like, tell us. We need you to test it. And then not only do we need you to test it, we need to go over how we phrased it, and make sure it actually captures what you said, and all the rest. And it just is like-- it really is this monumental sort of cross-team effort and cross-functional effort. [50:03] you know we we've been really investing a lot in our vibe checks and want them to feel like a special experience so feel you know not just be the average like our average post page feel like a really different website experience and much more interactive and we probably did that all in like 24 hours for two major models that released at exactly the same time you know normally we're just doing it for one um and so you know certainly had its hiccups along the way and we we learned a

[50:33] huge effort. [50:34] Are you using your cloth? [50:36] I am not using my cloth. Did you ever use your cloth? [50:40] No, I have one though. What's his name? Strunk. Strunk. I think you told me that that's hilarious. That is the best. [50:50] Okay. We've got to find some, I have some claw ideas for you. I didn't talk about it. I haven't used it, but. I think that Strunk could actually be a good copy editor inside of Proof, this markdown editor I've been building. Along with your, with the style guide. But anyway, that probably comes in a little while. Yeah. Okay. [51:10] to get you to your claw magic moment? I need my claw magic moment. Um, I, uh, yeah, I have one. And again, I, it's like, I have been taken by the hand helped to set it up here. I have it. I know what to do. I know technically what to do with it. And, um, as soon as I asked it to do something and it needed to integrate with one of my apps and it, it need, it was like needing something that I didn't have immediately. I was like, I'm not going to focus on this right now. [51:40] uh i mean i don't certainly higher higher odds this year than at any point i think the challenge is is both the challenge is sort of twofold one is like just consistency right you've always had trouble with just like is it consistently capturing and not just like making errors yeah and then the other is the judgment yeah um you know i we haven't

[52:06] talk that much about taste and that's been obviously like such a such a a big word important word loaded word but often at least when I am reading a piece and I'm reading everything before it goes on the site like it's the first time I'm seeing it which is by the way great that's like because I used to be like in involved in every piece of the in the heavy editing of it and that just wasn't sustainable for for after a while and so if I'm seeing it for the first time after it's [52:36] edited like [52:37] And it's basically been through our process. That's a huge win. But I am reading it. [52:44] not just for, oh, how does it, is the copy correct? I'm reading it like, how is this piece? Does it fit? I want to be super clear. I'm talking very specifically about, um, [52:56] this that repeated little copy edits that you do all the time that i think you shouldn't have spend your time on right yeah right no i appreciate that you've long said like there are other things you could be doing um what i end up doing though is like i'm basically reading something for the first time and then doing a top edit at the same time and so it's just like those steps are conflated for me versus like if i have read the draft and then i'm like great just like have a mechanical copy editor copy edit i'm gonna say june [53:24] Yeah. [53:25] Hi, it's what, February? June of this year. [53:28] That's my goal. I think the models are almost there. I mean, they're certainly closer than they have been before. Closer than they have been. Absolutely. I mean, I will say you were hoping for this vibe check, this double vibe check that we published. You were like, we're not going to use Google Docs. And I was like, absolutely not. This is not the time. I'm going to get what I want at some point. Maybe not on something that might be. My only problem is that Google Docs is not AI friendly. And I don't like that. That's fair. That's fair. That's not.

[53:58] final words before we sign off. [54:00] Any final words? Gosh, I don't know. I'm just really excited about the future. [54:08] Um, [54:13] You know, the recent... [54:17] time, the more recent time, and whether that's six months or nine months, I don't know exactly what the time period is. [54:24] It's been really sort of fun and exciting because I haven't, [54:28] felt like from an early stage perspective. Now, my previous companies were incredible companies, but like, you know, Stripe was extremely established, but like an early stage company of being like, is this working? You know, hi, is this working? And for a while, you know, you're just sort of like fumbling along, trying to figure out what will. And for the, you know, this past moment in time, it's been like, okay, like it's working. And so, you know, [54:58] and that has been really, uh, [55:01] really you know just really exciting and really and really fulfilling and and and I have had to learn oh you know whole different ways of doing my job um both because like I do now manage people so I'm not the IC anymore um and I have a lot of opinions about how I want things to be and so I need that to be clear and articulated and instituted and then also like [55:25] You know, every time you're like, there's a new model coming, I'm like, oh, my gosh. OK, let's check the calendar. What do we have going on? You know, like just try to keep up. Just try to keep up with everything that's everything that's happening. Well, it's been a pleasure doing this with you and excited for what comes next.

[55:41] Me too. [55:50] Oh my gosh, folks. You absolutely, positively have to smash that like button and subscribe to AI&I. Why? Because this show is the epitome of awesomeness. It's like finding a treasure chest in your backyard. But instead of gold, it's filled with pure, unadulterated knowledge bombs about chat GPT. [56:12] on the edge of your seat, craving for more. It's not just a show, it's a journey into the future with Dan Shipper as the captain of the spaceship. So do yourself a favor, hit like, smash subscribe, and strap in for the ride of your life. [56:27] And now, without any further ado, let me just say, Dan, I'm absolutely hopelessly in love with you.