How to Run a $100M Company with AI: v0 + Devin Tutorial from Gumroad CEO, Sahil Lavingia

1. 0:00 – 1:18

   Intro

   1. AGAakash Gupta0:00

      Sam Altman said one person will be able to create a billion-dollar company

   2. SLSahil Lavingia0:04

      Such that one person could ship the entire product. The nice thing about Gumroad is it's basically just me. It's a dictatorship, right? That's kind of why we can move super fast.

   3. AGAakash Gupta0:12

      Sahil Lavingia has one full-time employee, and he's taken Gumroad to a $10 million ARR and a $100 million valuation. From Slack to code.

   4. SLSahil Lavingia0:21

      It'll go and build this, and then you can actually see, based on your PRD, what did the designer think they were supposed to be designing. All of a sudden, this is now a feature in production without us having to really do anything.

   5. AGAakash Gupta0:30

      That's crazy. Weeks of coordination. You guys just decide, "We want to do this. Devin, address it."

   6. SLSahil Lavingia0:35

      We are leaving CSS behind to migrate every single one of our CSS files and delete them all.

   7. AGAakash Gupta0:40

      You can go straight from Slack to Devin.

   8. SLSahil Lavingia0:43

      This is, like, the first step I tell everyone to do if they're moving to start to use AI more.

   9. AGAakash Gupta0:47

      Oh, wow. Honestly, I think that's the new dream, though. Everybody wants what you have now. And how many dividends have you paid yourself in the last year? [beep] Really quickly, I think a crazy stat is that more than 50% of you listening are not subscribed. If you can subscribe on YouTube, follow on Apple or Spotify podcasts, my commitment to you is that we'll continue to make this content better and better. And now on to today's episode.

2. 1:18 – 2:09

   Meet Sahil Lavingia, Gumroad CEO

   1. AGAakash Gupta1:18

      Sahil, welcome to the podcast.

   2. SLSahil Lavingia1:19

      Ed, thanks for having me.

   3. AGAakash Gupta1:20

      I was telling you off-screen that I have been fanboying about you for a while. I once saw you in person four years ago, but I was too shy to actually come up to you, so this is a dream come true to be recording with you.

   4. SLSahil Lavingia1:31

      Nice.

   5. AGAakash Gupta1:31

      What are we gonna do today?

   6. SLSahil Lavingia1:31

      We are going to be cooking. We're gonna be making three things, walking through how we, you know, use AI, I guess small, medium and large. The first thing will just be kind of how we sort of use it to small, like, you know, to solve, like, a really small issue that comes up in Slack where we have Devin just kind of take care of the issue so that we don't have to get bothered by some small fix, some small feature, feature request that normally a engineer, designer, PM would have to tackle. The second one is going to be, like, sort of how we walk through building a feature, scoping out the PRD for an existing product of, of ours called Flexile. And then finally, we'll sort of try to build something totally from scratch, v01, and see how that goes. And we don't know what we'll build, I guess, so TBD on that.

   7. AGAakash Gupta2:07

      Truly live, everybody. Awesome.

3. 2:09 – 4:12

   Demo 1: Slack to Devon to Production

   1. AGAakash Gupta2:09

      So can you show us this first one? I was pretty mind-blown when you even just told me about this. From Slack to code.

   2. SLSahil Lavingia2:15

      Yeah, sure. This is, uh, an example of how we would use something like Devin to solve a problem. So, you know, we, we have Slack, like many companies who use Slack. We have an internal customer support tool that we've built called Helper, helper.ai, and we had a feature request from somebody basically adding this kind of assign yourself button when something is not assigned. So if you use something like GitHub, it just basically, instead of two clicks where you have to, like, first click unassign and then click your name, it's one click, so you save, you know, your customers one click, which is awesome.

   3. AGAakash Gupta2:42

      Yeah.

   4. SLSahil Lavingia2:42

      Uh, and so this was reported by one of our customer support staff, and she explained exactly why. This is very great. This is, like, exactly, like, the perfect feature request, which is, like, what you want and why you want it. And so I just said, "Hey, Devin, solve this. Address this." And Devin reads the Slack thread, sees the two images, and then sort of explains its planning and keeps you up to date, and then eventually opens a pull request. And, uh, here you can read the code, as you should. You should probably test it, make sure it did what you want it to do. Obviously, uh, if you have tests, you should make sure the tests are up to date, all that kind of stuff. So Devin, this is kind of, you know, kind of goes through and creates this, and then I do it and review it. So this is a good example of something that, like, no one had to really spend too much time thinking about, but it was solved. It was deployed. It was... And all the deployment and all that has long been automated, so all of a sudden, this is now a feature in production for people to use and benefit from without us having to really do anything. So yeah, that's one, one example.

   5. AGAakash Gupta3:37

      That's crazy. So the original process at a s- larger company that most people are living through right now, they are going to get that request in Slack. The PM is going to take note of it. They're gonna investigate it. They're gonna see, oh, how many people is this affecting? Is this a big enough request? Then they're gonna prioritize it. Once they prioritize it, there's gonna be a request for, create a PRD about this. They'll create the PRD about it. Then engineering is gonna create their engineering spec about it. Then, finally, it'll get prioritized in a sprint, and an engineer will actually work on it. Weeks of coordination. You guys just decide, "We wanna do this. Devin, address it."

4. 4:12 – 7:59

   The Dictatorship Model & AI Speed

   1. SLSahil Lavingia4:12

      Yeah, well, you know, I have a big reason that companies like that are, they do that, is because they need buy-in from so many people, right? So the nice thing about Gumroad is it's basically just me. It's a dictatorship, right? Uh, so that's kind of why we can move super fast, is you don't have to kind of, like, put together all this sort of work that's really just to get, like, internal buy-in. It has nothing to do with actually shipping the change. The, the hard part is, like, you know, aligning all the people to kind of b- get behind that specific decision. Uh, one other reason I really like AI is because I think it sort of is quite good. It's sort of like 99th percentile at many things. Uh, and so generally, the decisions it makes are pretty good ones, and actually, like, things like design and code, it's, it, we just kind of defer to the AI, right? And so it should come up in that sort of PRD. Something that's important should, should show up in the spec conversation itself, and if it doesn't, then it's not that important, and the AI can, can decide however to tackle the problem. You know, as long as it meets the criteria in the spec, it's going to, it wi- you know, it's "correct," quote, unquote.

   2. AGAakash Gupta5:07

      But could bigger companies adopt this for features that are a little bit less contentious, that don't need buy-in? Like, I feel like there's gonna be a future world where some companies are taking more of how you guys are working, maybe creating a bar for certain features and all, that way shipping much, much faster.

   3. SLSahil Lavingia5:25

      Yeah, I mean, I think the way that a, you know, a bigger company would probably do it, I don't have a lot of experience working at big companies, so I don't know, to be honest, like, the mechanics of them. But basically, you would just do it on a team basis, right? So you would pick a specific project. Uh, you know, microservice architecture's quite popular in these big companies for this reason, right, where you can sort of... People get frustrated that they have to move at the slowest pace of the large organization, and so they think that, oh, if they, like, factor off and they just have, like, a small API microservice, that they can now all of a sudden move much faster. They can make different technical decisions. They can make different architectural choices. It's true to some degree, but that introduces, like, a lot of other-... complexity that you then have to think about, and I think my bias is, is to tend towards, like, what's the right approach such that one person could ship the entire product? Like, so a lot of the way that big companies move faster makes it impossible to get smaller. Like, their process actually is probably quite fast, but requires, like, a number of people that you can't... Like, you need a PM, you need the designer, you need the engineer. Like, the way they work, you can't, without refactoring the process, you're not gonna be able to get... And that's largely why these roles even exist. People talk a lot about automation and unemployment and stuff like that, but most of, most of that unemployment, like it's, one, it will happen to these kinds of jobs. It'll be mostly these kind of like producer-like jobs. But two, they're going to persist a lot longer than people think because the production process is quite rigid in most of these organizations. You're not going to see companies all of a sudden decide to get rid of, like, three steps in their process. It's very difficult once you have process to get to unwind it, unless you have some sort of, like, like what happened recently with, was it EA, I think, which just went private. They were acquired. So in the case of EA, EA is a good example. Like, they probably have built up, like, a ton of bureaucracy over, like, the last 30 years that in the next five years is gonna get chopped, right? Like, the bureaucracy is gonna get massively sliced. Better or worse, it's going to lead to a much higher volatility in what EA as a company is going to ship instead of just, like, FIFA '26, FIFA '27. People, you know, like, they're not doing much, right? Like, they're, they're really beholden to this bureaucracy, which basically exists because FIFA was such a successful game that they basically haven't had to do anything innovative for 30 years. That will now sort of start to happen, and which I, I'm very, I, I, I think, I think that's, like, sort of a... I'm actually more interested in, in that. Like, how does AI, can you revitalize these existing businesses and brands, not necessarily new businesses and new brands? I'm actually, I have, like, zero interest, frankly, almost in, in, like, new things right now.

   4. AGAakash Gupta7:47

      Mm. So we saw the dictatorship workflow. Maybe not portable to other companies. I think the second workflow might be portable, especially some of those microservices, those smaller teams who want to pilot it. Can you show us this workflow?

5. 7:59 – 10:09

   Demo 2: GitHub to Feature Launch

   1. SLSahil Lavingia7:59

      Yeah, sure. So this is for Flexile. Flexile's open source, so anyone who visits this URL will be able to see. But we have a GitHub issue here, which is basically for this feature on Flexile. Flexile is a contractor payments tool, so you can invite people, you can pay them, and then they can accept the payment. They can type in their bank account, all that kind of stuff. And so there's a feature we have which the admin of a company could issue a payment, right? So contractor does some work, they want to get $400. You can give them $400. You can, you can sort of specify what it's for. And in our case, we've built equity into this. So in this case, you can sort of choose an equity range, and then the contractor can sort of say, "I want 10% or 50%," or, or what, but they have to kind of pick within this range. This is very complicated. It's reduced, like, a ton of complexity, and so we're basically going to simplify this and say, "We're gonna get rid of this." The money, like, when you say, "I want to send someone $400," you're done. They just get $400. If they want equity, that's going to be specified ahead of time. It'll just show up as a note here. You know, for example, like, they will receive 10% of this, $40 in equity and stock options. But this way, be able to delete probably a few hundred lines of code and, like, it'll be much simpler because you don't have this, like, two, three-step process. So anyways, all of that I just put in this, which is not super helpful, right? If you're an engineer and you're like, "Okay, great, Sahil wants me to do this," I can guess roughly what this is going to be, but, you know, I would love to know more. And so people have questions. We have a conversation about this. You know, this engineer asks a question. You know, I flew somebody else in. We kind of have this sort of discussion, debate, et cetera. And then, you know, typically from here, what you would do is you would sort of create a PRD. Uh, so in the case of this, I pasted this into GPT 5 Pro, which you probably don't need something like Pro, but it kind of comes up with this. You know, now a human would kind of go through this and, and read, uh, you know, edit this down. To me, this is, like, way too long, uh, and so often I sort of... That's why I sort of think the PRD is dying. Like, I don't think you really need it anymore, uh, because the PRD is sort of, you can think about it like this, like, internal resource for a company. Like, you don't really give it to the customer. You're just kind of using it to get the sort of PMs and engineers and designers aligned. So if you can sort of get everyone closer together and have more overlap in the skills, you can, like, operate off a much leaner and tighter PRD.

6. 10:09 – 12:19

   The PRD Is Dying

   1. SLSahil Lavingia10:09

      So this is the PRD in my case, right? Like, this truly is.

   2. AGAakash Gupta10:11

      Mm-hmm.

   3. SLSahil Lavingia10:11

      And so as people have questions, I would sort of go up here, and I would sort of, like, flesh this out, you know? Uh, you know, like for example, uh, always use the equity percent that was selected by the contractor. Note this in the modal, et cetera. So you kind of, like, basically, like what I love about this process is you learn where your PRD is lacking, and your PRD is only as dense as, like, what can basically cannot be inferred, you know, just naturally, right? So for example, I could sort of take this and go into v0 and say, you know, for Flexile, build this fe- you know, build this feature, and imagine this PRD was much longer. Imagine it was, like, this long. You know, you could, like, build this. It'll go and build this, and then you can actually see based on your PRD, like, what was, what did the designer think they were supposed to be designing? What did the engineers? And if, if they're wrong, if they, if you sort of think they took a different approach, then you go back and reread your PRD and you say, "Oh, crap, I forgot to mention that, like, do not update," you know, whatever, right? So it kind of-

   4. AGAakash Gupta11:11

      Okay

   5. SLSahil Lavingia11:12

      ... allows you to have almost, like, this, like, fake conversation with, like, the engineer and the designer such that when you actually give your PRD to the engineer and the designer, which may still exist in your company, they get a much better PRD, right? Like, they're getting a much tighter, much more accurate to what you want, what's in your head. 'Cause you're s- that's sort of the problem is there's something in your head that you're trying to express in English and you're failing, right? Basically. There's a lot of stuff that you, that you wanted to have accomplished that you didn't necessarily state. You know, for example, there's like, there's this whole bit that I, I, uh, it was in my head, but I didn't sort of state it, so now I have to go back and state that.

   6. AGAakash Gupta11:45

      Mm-hmm.

   7. SLSahil Lavingia11:46

      Yeah. So that's, that's kind of my flow.

   8. AGAakash Gupta11:47

      And are there any tricks to getting v0 to, like, respect your design system, or is it fine if it's just whatever it comes up with?

   9. SLSahil Lavingia11:53

      I personally don't care because for me, v0 is, like, an internal tool just to get the UI/UX right.

   10. AGAakash Gupta11:58

       Mm-hmm.

   11. SLSahil Lavingia11:58

       And generally, the code base, sort of that's, that will be tackled in sort of the code base step. And if you have a w- a, a well-architected code base, then it, you won't have to worry about it. It'll happen perfectly every time. If you don't have a well-architected, architected code base, then you're gonna have a lot more issues. But yeah, generally, I kind of have this-

   12. AGAakash Gupta12:12

       So for people who may not exactly know what that means, what is a well-architected code base? What are the elements that they need to establish to make it that way?

7. 12:19 – 15:25

   Architecture: Why CSS Kills AI Speed

   1. SLSahil Lavingia12:19

      Yeah, I mean, this is-Sort of like the now most valuable skill in software engineering, which is to architect the code base such that making incremental changes to it is really easy and intuitive. Uh, but, you know, for example, uh, there's, there's, there's, there's, for a web application, there are several parts of the code base. One will be the design, right? Like, uh, typically the design used to be handled with CSS, so you have these massive CSS files. Uh, so for example, I can show you in Gumroad's case, which is also open source, so people can follow along if they want. But if you look up our CSS files, you'll see that we have tons of CSS files that define the, the style of the application, right?

   2. AGAakash Gupta12:59

      Yeah.

   3. SLSahil Lavingia12:59

      Uh, and they all live within this, right? So the calendar, the car, the carousel, the car, like, there are all these things. This is a terrible way of building stuff today because what it means is all of this code, to make a single change on gumroad.com, this code will be included on every single page, right? So if your code has a pro- if your page has a progress bar on it and you change the width to size 7 instead of 5, size 5, that'll change the progress bar across all of Gumroad, right? So now you have to test every single page in which a progress bar shows up, right? So then you have to search for this specific thing, progress bar. Search for it in the code base, and then check, m- update each one of these places and test it to make sure. Like, for example, here you're overriding the width of the progress, right? Um, and the display. So you have to test. You can kind of notice in a lot of these places, the width is .75em, so maybe that should be the default. But there's no way to... The default is set to size 5. What is size 5? I don't know. It's coming from somewhere else in the code base that I have to go, to go discover, right? So there are all these, like, misdirections where it's really hard to start to build out this, this thing. The new way, world of building things, we use Tailwind. Tailwind, if you go to Flexile, uses Tailwind. There's no CSS files. If you search for CSS, you only have a single CS- CSS file in the entire thing, which basically just defines, like, the overall themes. And this way, you never, as, like, the AI never has to think about what's the foreground color, what's the background color, what blue do we use, what red do we use? The entire design system literally basically is here. The visual design system is here. Like, the font, the color. What else is there, right? Padding. There's not much else, right? And so sort of super, super easy to understand. Like, like, think about how much context is here, right? It's 181 lines of code, and you can basically build stuff within Flexile. But if you wanted to understand the design system in Gumroad, all of a sudden you're not 100 lines of code. You're at, like, 100 files [laughs] of which each-

   4. AGAakash Gupta14:47

      Wow

   5. SLSahil Lavingia14:47

      ... each one is going to have, you know, lots of code. Like, you go to one of these files, you have all of this code that you have to think about, right? And so, yeah, so j- so sort of Tailwind. You can, you know, you can have this conversation with ChatGPT, like, why is Tailwind better than Global CSS for shipping features? By the way, nothing to do with AI, right? Like, and if you really want a much faster response, you can ask ChatGPT with, without Pro turned on. But anything that basically is good for humans is good for AI and vice versa. So this is, Tailwind is good for shipping humans. It's just much more noticeable because the rate of speed increase with AI is much more, right? So

8. 15:25 – 17:23

   Deleting 5,425 Lines of CSS

   1. SLSahil Lavingia15:25

      if you make this change, and, and we, we basically, like, this is, like, the first step I tell everyone to do if they're moving to start to use AI more. Like, they need to de- delete their CSS.

   2. AGAakash Gupta15:33

      Oh, wow.

   3. SLSahil Lavingia15:34

      You're not gonna be able to make these changes. So for example, like, this addressed this, right? Like, why was this so simple and easy? Let's see the code. The code was one file change, right? Because there are no CSS files, there's no CSS to change, right? So in the ca- let's say we were using CSS. Well, you have to change this file, which is the component, the equivalent of HTML. Then you have to change the CSS file, right? There would be an equivalent CSS file, right? Well, that CSS file is referenced from every single TSX file. So now you have to, effectively, it's like as if you edited 300 files, [laughs] right? So in this case, only one file. You only have to... And all the CSS changes are right here. Class name, flex, item center, gap 1. All of these changes are, you wanna make a button, but with this color, and you wanna underline it on hover, and you want the text to be small, all here, you know? And in reading this as a human, I can see all the changes. There's only one file, right? Uh, it's, it's, like, infinitely simpler, uh, to work on. Uh, so, so that's one, sort of one thing about architecting the code base. And you can actually see with, with Gumroad, we are leaving CSS behind, and we have a task open to migrate every single one of our CSS files and delete them all. So all these files will be deleted, uh, soon. Uh, and to sort of calculate that for you, let's see if I can open up this. Find the file. So yeah, so, like, in this, you know, I could just k- ask AI, how many lines of code are all CSS files, and see what it does. You know, thousands of lines of code, right, will be deleted. So-

   4. AGAakash Gupta17:01

      Yeah

   5. SLSahil Lavingia17:01

      ... a huge amount, like a book's worth of content will be deleted. Uh, so all this. And the, you know, number of bugs that can exist will, you know, like, everything will improve. Although some of this code may not even be used, and we don't even know.

   6. AGAakash Gupta17:12

      Mm.

   7. SLSahil Lavingia17:13

      You know? It's, it's hard to say because it's, you know, uh, it's, it's, it's all these, it's all hidden in all these different files. 5,425 lines of code.

   8. AGAakash Gupta17:21

      Crazy.

9. 17:23 – 19:34

   Ads

   1. AGAakash Gupta17:23

      Today's episode is brought to you by Vanta. As a founder, you're moving fast toward product market fit, your next round, or your first big enterprise deal. But with AI accelerating how quickly startups build and ship, security expectations are higher earlier than ever. Getting security and compliance right can unlock growth or stall it if you wait too long. With deep integrations and automated workflows built for fast-moving teams, Vanta gets you audit ready fast and keeps you secure with continuous monitoring as your models, infra, and customers evolve. Fast-growing startups like LangChain, Writer, and Cursor trusted Vanta to build a scalable foundation from the start. So go to vanta.com/aakash. That's V-A-N-T-A.com/A-A-K-A-S-H to save $1,000 and join over 10,000 ambitious companies already scaling with Vanta.

   2. AGAakash Gupta18:15

      AI is writing code faster than ever, but can your testing keep up? Testkube is the Kubernetes native platform that scales testing at the pace of AI-accelerated development. One dashboard, all your tools, full oversight. Run functional and load tests in minutes, not hours, across any framework, any environment. No vendor lock-in, no bottlenecks, just confidence that your AI-driven releases are tested, reliable, and ready to ship. Testkube, scale testing for the AI era. See more at testkube.io/aakash. That's T-E-S-T-K-U-B-E.I-O/A-A-K-A-S-H. And so the powerful thing here is you define your design system using Tailwind, and then v0 can just pick up those components, and v0 will, once it gets put into your code base, it'll have your design system in place.

   3. SLSahil Lavingia19:06

      Yeah, exactly. Exactly. Uh, and it's just a industry standard, right? So just think of the difference of like teach... You know, you hire an engineer, and they're used to two-by-fours, right? They understand the sy- like, the AI knows how this stu- you know, it's, there's so much training data out there about Tailwind, et cetera. So it's just, like, really easy to get started and make progress.

   4. AGAakash Gupta19:26

      Amazing workflow. So that's the second workflow. Now let's look at the large workflow. How does it look like end-to-end if we're building from scratch?

10. 19:34 – 27:00

    Demo 3: Building Kit Competitor Live

    1. SLSahil Lavingia19:34

       Yeah. What do you want to build?

    2. AGAakash Gupta19:35

       That's a great question. Let's build something. Let's say we were building a competitor to a Kit or something like that, like a email provider for creators. How would we get started with that?

    3. SLSahil Lavingia19:45

       Yeah. So normally I would frankly start in Apple Notes. That's like my default. Something like Google Docs, right? And you just decide what you want to build. So you want to build a Kit competitor, so this is like an email newsletter kind of-

    4. AGAakash Gupta19:57

       Exactly. Really focused on-

    5. SLSahil Lavingia19:58

       Email newsletter

    6. AGAakash Gupta19:58

       ... like, creators as the audience and really helping them, like, be easy to send stuff that delivers.

    7. SLSahil Lavingia20:04

       You know, so this is our PRD.

    8. AGAakash Gupta20:05

       Yeah.

    9. SLSahil Lavingia20:06

       Right? Basically. Uh, I use v0. Uh, you know, people have their own preferences, so, you know, use whatever you want.

    10. AGAakash Gupta20:11

        So you go straight into v0, getting the front end started. Not really a step needed in between, like, the four lines and v0, like creating a PRD or something like that.

    11. SLSahil Lavingia20:21

        Yeah, in my opinion. You know, a lot of people, they invest in the PRD because they're gonna give it to a human, right? And they wanna make sure that they don't waste that human's time.

    12. AGAakash Gupta20:27

        Yeah.

    13. SLSahil Lavingia20:28

        If you're just giving it to an AI, like, why not, you know? And, you know, by the way, like, ChatGPT now supports this, right? I think. So, like, I think if you turn on canvas, you know, do... I think that's how you do it. Oh, no, canvas is a-

    14. AGAakash Gupta20:39

        Just give me a prompt.

    15. SLSahil Lavingia20:39

        I think it, I think maybe if you just say, "Build this." Or so n- it created a PRD. I don't want a PRD, just a prototype.

    16. AGAakash Gupta20:45

        Oh, let's see what ChatGPT-5 does here. I haven't tested it.

    17. SLSahil Lavingia20:48

        Uh, yeah. I've heard, I've heard generally good, pretty good things about it, so...

    18. AGAakash Gupta20:51

        Hopefully it's powered by Codex, which people are loving.

    19. SLSahil Lavingia20:55

        I guess preview. So yeah, something, you know. So you can iterate on it, right? You can say, "Okay, I don't really like," like send test or, or whatever. But you basically, it forces you to reckon with, like, what are you at? What's unique about your idea, right? Like, what are you-

    20. AGAakash Gupta21:06

        Yeah

    21. SLSahil Lavingia21:06

        ... building? You're forced to sort of, like, not just say, "Oh, it's an email newsletter thing," but like, okay, great. Like, what, how does it look like? What is the l- name? What is, uh, what is, like, the core functionality that is, you know, different about your system than somebody else's system? Things like that. So that's sort of ChatGPT, you know, say like, you know, "Make it look like Stripe built it," you know, something like that. See what it does there. And this is often what I do. I have, like, several open, you know, 'cause it takes some time.

    22. AGAakash Gupta21:30

        Yeah.

    23. SLSahil Lavingia21:30

        I'll often have, like, three or four different windows open and just kind of, like, run through them and just kind of, like, juggle different ideas and, like, it's almost like having an army of assistants I'm kind of, like, riffing with, trying to come up with, like, an interesting idea.

    24. AGAakash Gupta21:41

        Yep.

    25. SLSahil Lavingia21:41

        Uh, and I just kind of hop around and-

    26. AGAakash Gupta21:43

        Yeah, just think about this more, like, some ways we can differentiate, right? It basically makes it really easy to, like, create good landing pages, make it look really good, but they start with, like, a paid plan. So I think we're gonna, like, want to start with, like, a free plan or something like that. It's gonna be the big thing.

    27. SLSahil Lavingia21:58

        Yeah.

    28. AGAakash Gupta21:58

        And I guess, like, Kit is, like, really, really good at, uh, driving, like, a lead ma- like, allowing you to spin up lead magnets or different introductions to your newsletter send versus, like, a Substack is like, there are no lead magnets or anything. So I think that'll be the direction we take it.

    29. SLSahil Lavingia22:14

        Gotcha. Yeah, it's kind of slow. This is the biggest problem. Things just can take some time.

    30. AGAakash Gupta22:18

        Yeah. And you mentioned you've tested out a couple different tools. What other tools have you tested out for this? Why do you end up landing mainly on v0?
11. 27:00 – 31:11

    Working on 5 AI Projects Simultaneously

    1. AGAakash Gupta27:00

       and do you use Devin in Cursor?

    2. SLSahil Lavingia27:02

       No, I use it, I use Devin through Slack-

    3. AGAakash Gupta27:04

       Okay

    4. SLSahil Lavingia27:04

       ... or through their website.

    5. AGAakash Gupta27:05

       Okay, nice.

    6. SLSahil Lavingia27:06

       So normally I use, I, I use Devin for like small stuff and Cursor for big stuff. That's kind of like the simple binary.

    7. AGAakash Gupta27:12

       Yep.

    8. SLSahil Lavingia27:12

       But yeah, you get, you know, it talks about, oh, you know, border radius, for example. Our theme, you know, indicates stuff about, uh, how round things should be and all the different colors. Oh yeah, here the radius. You know, the color of our links should be blue. You know, s- uh, design system is actually like pretty small. There's actually not that many things that define a brand, right? If you think about it. The colors, the font, the padding, uh, the shadows, the borders. It's pretty cool. So this, yeah, it'll, it'll take some time, but it'll, it'll, it'll run through it and then, uh, yeah. I mean, you know, pretty interesting. Let it continue doing this thing.

    9. AGAakash Gupta27:43

       Very powerful.

    10. SLSahil Lavingia27:44

        Yeah, so you can see it, it didn't do much because we probably already have like a similar theme, but it cleaned up these things. You know, if you want, you could... There's another thing people, you know, you can do. You could literally just take a screenshot of like, you know, Flexile.com. Say, you know, let's say you wa- it's like not blue enough. Make it look like this. As a parent company, right, you want similar branding, right? Again, like state why you want something. AI's good at reason, so if you state why you want something, it'll actually like help it think through the right approach.

    11. AGAakash Gupta28:11

        Mm-hmm. Presumably v0's probably using some Claude models since those are the best Claude models, but they probably have some mixture of models for different tasks, I guess.

    12. SLSahil Lavingia28:20

        Yeah, exact- Yeah, exactly right. They have some generally framework for... It's not that much more than a sim- a simple model, but there's a little bit more of like, you know, planning step, executing step, wireframing. Like they, they have a few of those. And some models they can use a cheap, fast model. Other things they want like a slow, smart one. So yeah, I mean, you, really the, I feel like these models are effectively as intelligent as I need them to be, not that much more is necessary for me to feel like they sort of have replaced most kinds of software engineering. Uh, the biggest thing is as you see, the speed, right? Like humans are faster. [laughs] So-

    13. AGAakash Gupta28:57

        Yeah

    14. SLSahil Lavingia28:57

        ... uh, if this could be instant, like if this like, really I feel like that's the, the core problem, is I, I, I actually, it's just slow. So things just take a lot of time. It's like, okay, see now everything is blue. It, it did a pretty good job. Like it looks much more like Flexile now. Still not perfect. I would still mess around with, you know, uh, for a while. Uh, you know, for example, like this thing, it would be cool if this like counted down. But again, the re- the, I'm always doing like five different things at once, right? I'm answering email, I'm answering Slack, I'm PMing something, I'm do- doing something on Twitter. I'm like, that way, like all this dead time, I'm not waiting, right? I'm going out and doing something else and coming back like, you know, five minutes later.

    15. AGAakash Gupta29:34

        Yeah. I found that-

    16. SLSahil Lavingia29:34

        And the demand for these AI services is so great that it's kind of like LA traffic. There's just a lot of demand to use these tools, 'cause they're very useful.

    17. AGAakash Gupta29:41

        Yeah. I found that, uh, time blocking on my calendar doesn't work nearly as it well as it used to, 'cause I need to be doing like four things at once [laughs] with these AI tools.

    18. SLSahil Lavingia29:50

        So here's a good example of a fade PR- failed PRD. It's doing what I asked it to do. It added a countdown. I think that's kind of cool. But really I meant for it to be animated, [laughs] and specifically with a specific React library called Number Flow.

    19. AGAakash Gupta30:02

        Oh, okay.

    20. SLSahil Lavingia30:03

        So I'm gonna ask it to... So that was in my head, animated using Numberflow, but I didn't say, uh-

    21. AGAakash Gupta30:09

        Yeah, in the old world, you're paying people, like-

    22. SLSahil Lavingia30:10

        And a really good, a really good front-end engineer would know that I meant this. Uh, so-

    23. AGAakash Gupta30:14

        Yeah, it didn't work with the-

    24. SLSahil Lavingia30:15

        ... AI is not as good as, like, a really talented sus- you know, front-end person who would understand that, like, when I say countdown, I mean, like, a metaphorical countdown, like something that's... I'm not adding it just for the function, I'm adding it to get the customer excited that we are good at software engineering, and good software engineers can do animation. That's sort of what I'm... All the stuff I didn't say, um-

    25. AGAakash Gupta30:35

        Yeah

    26. SLSahil Lavingia30:35

        ... and it looks like it failed because there's some error, uh, so we'll ask-

    27. AGAakash Gupta30:39

        It to fix itself

    28. SLSahil Lavingia30:40

        ... you know, generally, you just have to click this Fix button. Another part of being a software engineer. Okay, like, it looks like this works, so let's run this version. Then I'll say run locally and see, let's see what... So yeah, it created a Next.js app, but, uh, I'll run it locally and see. Yeah, there's some... Yeah, it's doing some animation, but it's kind of weird.

    29. AGAakash Gupta30:56

        Oh, yeah, it's not as good as the example.

    30. SLSahil Lavingia30:58

        It's not as good as the example. It's not as good as the example. Why not? All you gotta do is copy and paste the example. Come on.
12. 31:11 – 35:14

    Ads

    1. AGAakash Gupta31:11

       Today's episode is brought to you by the experimentation platform Kameleoon. Nine out of 10 companies that see themselves as industry leaders and expect to grow this year say experimentation is critical to their business, but most companies still fail at it. Why? Because most experiments require too much developer involvement. Kameleoon handles experimentation differently. It enables product and growth teams to create and test prototypes in minutes with prompt-based experimentation. You describe what you want, Kameleoon builds a variation of your webpage, lets you target a cohort of users, choose KPIs, and runs the experiment for you. Prompt-based experimentation makes what used to take days of developer time turn into minutes. Try prompt-based experimentation on your own web apps. Visit kameleoon.com/prompt to join the waitlist. That's K-A-M-E-L-E-O-O-N.com/prompt. Today's episode is brought to you by The AI PM Certification on Maven. Run by Miqdad Jaffer, who is a product leader at OpenAI, this is not your typical course. It's eight weeks of live cohort-based learning with a leader at one of the top companies in tech. As you know by now, the future of PM is AI, and this certificate will give you the learnings plus the hardware to show you are ready for an AI PM role. I myself took the course and recommend it. Put on by the amazing team at Product Faculty, including Mo Ali and Pavel Hern, it's worth it. Former students come from companies like OpenAI, Shopify, Stripe, Google, and Meta. The best part, your company can probably cover the cost. So if you want to get $550 off, use my code, AAKASH550C7. That's A-A-K-A-S-H-5-5-0-C-7, and head to maven.com/product-faculty. That's M-A-V-E-N.com/P-R-O-D-U-C-T-F-A-C-U-L-T-Y.

    2. SLSahil Lavingia33:06

       Well, we're gonna try one more time, and we're gonna literally say, like, this. That's the code. The code needs this import number flow. So now it should get it. So you can see that, like, you know, that's what PMing is, right? It's just, like, being incredibly explicit with what you want.

    3. AGAakash Gupta33:20

       Yeah.

    4. SLSahil Lavingia33:20

       And if you are incredibly explicit with what you want, then the sort of software engineering intern, et cetera, like, will do what you want it, you know, want it to do. Yeah, so this is Flexile's growth. Pretty cool. Things are working. You know, the way we work is starting to work. We grew 50% [laughs] month over month for some reason. I don't know, I don't really know what we're doing that's working, but that's, that's cool. It tried, uh, so let me, uh, let me just quit Co- Codex, say NPM write Ella. Where did it cr- oh, yeah, create a newsletter, NPM write dev, and it created it, so I'm gonna open it up and see what it did. 'Cause this is why I don't like Codex, right? It's, like, a lot of trust. You don't know until you open it up. So this is what it did.

    5. AGAakash Gupta33:58

       [laughs]

    6. SLSahil Lavingia33:58

       Wow, that's pretty impressive.

    7. AGAakash Gupta34:00

       I think it did the most work.

    8. SLSahil Lavingia34:02

       It did a lot of work, yeah. And that's what people say. It's sort of very industrious. Like, it does a lot more than you would expect. Like, it iterates on itself quite a lot. So this is pretty cool, I would say. Like, it's all in one page, step one, step two, step three. I don't really like the UX. It kind of has this, like, very, like, old school, like, template vibe.

    9. AGAakash Gupta34:19

       Yeah.

    10. SLSahil Lavingia34:20

        But it's sort of pretty functional. It's, it's, uh... I assume this will update automatically. Yeah, so the-

    11. AGAakash Gupta34:24

        And if we build in Codex-

    12. SLSahil Lavingia34:25

        ... all this stuff works

    13. AGAakash Gupta34:26

        ... presumably, like, for Codex, we could have linked it up to the code base at the beginning and the Tailwind CSS file that Flexile or whatever, if we had an existing product, had, if we were doing that-

    14. SLSahil Lavingia34:35

        Totally

    15. AGAakash Gupta34:35

        ... we could have done that.

    16. SLSahil Lavingia34:36

        Yeah, exactly. Yeah, you know, all, the way you, the way you use, you use Codex, so, like, is you just open it in any terminal, right? So if I'm, if I'm working on Flexile here, right, I just run Codex, and then all of a sudden I can... All of a sudden I'm using Codex, right? Whatever I want to do. So, like, for example, like, I have that task of, like, uh, you know, remove the figure, you know. I can just say generically, uh, actually, let me, let me see just for fun how, how, uh... Yeah, and this is a good test, is how good the spec is, right? Like, you can just say, like, in theory it should just do it, right? If it's AGI, it should do it. If it's not, then, you know, there's something I didn't state I need to connote.

    17. AGAakash Gupta35:08

        Mm.

    18. SLSahil Lavingia35:08

        So yeah, something like this would take days, now would take hours, so that's a pretty significant boost in productivity.

13. 35:14 – 36:11

    Three-Tier Workflow Summary

    1. AGAakash Gupta35:14

       Love it. So I think that's it for the demos. I think we showed everybody a really comprehensive set of three workflows. You can go straight from Slack to Devon if you're in a dictatorship. You can talk about it more if you're a more mobile team or a bigger team in GitHub on an issue. You don't even need to create a full PRD. Or you can iterate on your PRD with AI and a prototyping tool, then connect it into your design system with something like Tailwind and make it much better, much faster to reveal what are all those things inside your head that you actually wanted to create.

    2. AGAakash Gupta35:46

       Amazing. That's the demo section. This is how Sahil is running his company. Now you guys got an inside look. This is how you power an AI company with AI tools. I wanna talk a little bit about Gumroad because it's just, it's gotta be one of the most fascinating companies out there. Tell us a little bit about how you guys are structured. Uh, you have only one employee, but you have contractors. How do those contractors get work? Do you pay them? What is the structure for getting work done at Gumroad?

14. 36:11 – 39:00

    Gumroad Structure & Open Source Model

    1. SLSahil Lavingia36:11

       Yeah. We've been through a lot. You know, we started out as, like, a sort of traditional venture-backed startup in 2011 up until 2015 when we failed to raise our Series B. That's when I sort of did our first sort of major pivot as a ... The product stayed the same, but a sort of operational pivot, so we downsized. We went to about three, three or so people, got to profitable, and sort of ran the business myself for a while, did customer support, product support, you know, fraud risk, customer support, fixing bugs. Not that many new feature shipments, uh, but, you know, basically just tried to kind of keep the site up and respond to customers and stuff. Uh, then we started growing a little bit. COVID happened. That was a big boon. Crowdfunding round, so all of a sudden we had some more cash. We started growing the team again, but this time sort of differently, primarily contractors, hourly, uh, part-time, remote. Uh, and we have a core team now that's, like, full-time, so me and Irshad, and, and we have, uh, maybe three or four people who work, like, 30 hours a week or so, and then, uh, then we have the bounties that we do with the open source community now. So all of our code is open source, which means we don't actually need that many engineers on the team because you don't have to be on the team to work on the code, right? So in theory, you need a lot less, uh, less people, and that's kind of how we work now. So, so it's sort of a s- you know, a small team. Uh, and it's hard. We have an office that could fit 11 people, but it's just tough to, uh ... You know, our bar is really high, and AI is so cool, and there's so many, so much opportunity for everybody out, out there that it's hard to, hard to find the right, the right people, so.

    2. AGAakash Gupta37:30

       So you aimed, it sound like, at one point to build a billion-dollar VC-backed company. What is your current goal for Gumroad?

    3. SLSahil Lavingia37:38

       I mean, very practically, very boring, uh, but we're trying to get to 10 million in EBITDA. It's just nice to have a goal, nice to have a short-term goal for the year where the team and everyone can align on, like, what we're trying to do and the changes we're trying to make and the sort of order of magnitude changes we need to make to, to make that goal happen, which I think we'll get to. It'll be tight, but that's a good goal. I think pushes you at the very end to, to, to work a little bit harder. Uh, beyond that, I haven't thought about it too much. Like, open sourcing everything was a goal, but to be honest, like, we don't have a lot of goals 'cause, like, if we have, if we wanna do something, we just do it. You know? Like, there's not, like, a long lead time to doing stuff, so, um, we just started doing more frequent on-sites. Uh, so, you know, trying to sort of one week a month where we get together in person and build stuff together. Uh, so, you know. Well, I think a lot of my focus over the next couple years will be on, like, now I have a kid. I have a five-month-old, so balancing the two and structuring the company and the culture such that I can be really productive and, and, and still do stuff with the team, uh, but then also, like, not have it be as intense and as, as all-consuming as it can be sometimes. Uh, and so that's kind of, like, the, the sort of balance that I'm trying to strike. But yeah, nothing too crazy, to be honest. I feel like for the first time in, in, in a long time, which is the nice, the nice thing about having a profitable business doing dividends, is I'm not, like, looking at the future and the, like, I get dividends every year. So I'm, I'm not, I'm actually s- I'm trying to, like, operate the business, you know, slightly better, like incr- like, almost, like, as a perfectionist. Like, can we just perfect the current business, uh, instead of, like,

15. 39:00 – 45:00

    $10M EBITDA Goal & $2M Dividends

    1. SLSahil Lavingia39:00

       do other stuff all the time? So we'll see. Flexile's fun to work on. It's growing. You know, chip away at that. You know, fix bugs, improve the product. I just wanna, like, work on software, make the software better, have people use it, you know, and just, like, be proud of the work we do, have some cool people to do it with so I have some structure in my life. You know, stuff like that.

    2. AGAakash Gupta39:18

       Honestly, I think that's the new dream though. Everybody wants what you have now, like, a business working towards 10 million in EBITDA. Where are you guys at so far in the progress towards that?

    3. SLSahil Lavingia39:27

       Yeah, I think we're at, like, seven, seven or eight million.

    4. AGAakash Gupta39:30

       Wow.

    5. SLSahil Lavingia39:30

       So, you know, still, still work to do. Like, it's not gonna be a walk in the park. But, uh-

    6. AGAakash Gupta39:35

       That's pretty amazing. I'm sure people-

    7. SLSahil Lavingia39:36

       Yeah. I think, I think we'll get there by the, sometime next, next year.

    8. AGAakash Gupta39:39

       And how many dividends have you paid yourself in the last year?

    9. SLSahil Lavingia39:41

       Um, last year I got about 2 million bucks.

    10. AGAakash Gupta39:44

        That's [laughs] really the life, right? Living, uh, building a profitable software business, having a five-month-old, controlling your schedule, paying yourself $2 million. I think a lot of people wanna aspire to that. When you talk to people, you know, outside of, when they're not in the VC-backed realm and they're just hiring like crazy, but they are trying to stay bootstrapped, profitable. They are growing. They're trying to take on AI tools. What are the things getting in their way? Where I see you very freely just using AI tools, what are the unlocks those founders need to hear?

    11. SLSahil Lavingia40:10

        Yeah, I mean, maybe their businesses are just doing super well, uh, and so they don't need to reinvent themselves. I think that's a big part of it, you know? Like, I think people underestimate. Like, if you're doing fine, like, there's no reason to, like, investigate the future. So yeah. So I, I would say that's the biggest one. You know, just, like, P0, right? Like, there's only so many priorities you could actually have. So, you know, if it's not your number one priority in the business, like, you have one more priority, like let's say it's hiring someone, you're gonna be spending all your time on that, so that's kind of just, like, the way it goes.

    12. AGAakash Gupta40:36

        Mm. So you have some other things you're working on, like Antiwork. You talked about Flexile. I think you guys acquired, like, Small Bets. How does it all fit together, the picture for you?

    13. SLSahil Lavingia40:44

        Yeah. How did it fit together? Um, I don't know. Honestly, I just kinda did, like, a retrospective on, like, COVID and everything else like that and just felt like, just felt like, uh, why not make a bet, and what was, like, the best fit? It wasn't like, "Is this the right fit?" It's just like, "Well, let's just make a bet on the best fit." So that's what we did. Uh, and now we're in, you know, nice to have a community. You know, nice to have a place that's kind of, like, our default, like, community of small-time creators, and so that's kind of, like, why we did it. And, you know, I was thinking a lot on, like, okay, we built this, like, big, you know, like, last month, like, 35,000 people sold at least one product on Gumroad, so there's, like, a huge number of people. But Gumroad has kind of plateaued, and we sort of have what we want. We're incrementally making it better, but I don't think that's, like, more features isn't really what creators want. Like, community advice, stuff like that I think would be better, and so that's kind of, like, where we, uh, where we kind of decided to say, "Oh, maybe there's this, like, community, and we can upsell." And so there's, like, the base Gumroad SaaS, and then there's, like, the community that can ac- actually help you, like, with legal, finance, business, marketing, sales, tax advice, hiring. That's where I think people, creators sort of-Tend to need to need help get- getting to the next level, going from, like, one person to a team, for example. Like, even me, you know, I technically only have one employee or whatever, but, like, that's ... Who cares, right? Like, what matters is, like, are you, you know, doing the things you wanna do? Are you ... You know, do you have a great... The reason we don't have $1 billion single-person companies is not that we can't, it's just that we don't want to. [laughs] Like, like Google has 3,000 billion-dollar companies in Google alone. There's, like, probab- I, I can't imagine how many PMs run a billion-dollar company by themselves, you know? So it's probably pretty common. Uh, within FAANG, there's probably thousands of examples. It's just outside of FAANG, if you're doing it all yourself, like, then, like, I have to do the bookkeeping, I have to do the ... You know, like, it's just not, not what most people want to do. Uh, so it tends to be more of, like, a marketing gimmick, like, oh, one person, you know. So yeah, that's kinda how I... I mean, maybe that's different, like, when you have true, true AGI, where you really don't need any accounting, like, ev- all sort of tasks you can sort of have a robot do. But in my mind, if that's possible, then, like, the robot will automate you too. You'll kind of have a zero-employee business. You won't have, like, a one-person business, right? You'll just go straight to zero. You'll still own the business, but there will be no people working on it, you know? It'll be kind of like a smart contract or some- something like that.

    14. AGAakash Gupta42:54

        Yeah. I feel like that's what I discovered too scaling my business. [laughs] After a while it's like, what, what was the point of being a solopreneur? Like, I, just so I could put the label solopreneur? I think I'm gonna switch over to entrepreneur and hire somebody to help me with taxes, with sales, with all these different activities that you don't wanna do. There's... You don't have to stick to a solo business.

    15. SLSahil Lavingia43:14

        Yeah, exactly. I, I think that's a big part of it. Like, I think there's just so much... There's sort of, like, the pro-VC anti, you know, like pro-VC cult, and that sort of a lot of the indie hacking is kind of, like, a response to that. But I think it also is a cult in a way, right? Where it's just like there's nothing wrong with having a team, there's nothing wrong with, like, working in person, there's nothing wrong with, like, being acquired or there's nothing wrong with having a job. Like, there's just like we're all basically trying to s- manage our time and money equation, and, like, the way that ends up looking for people is going to be different, and it, it's very dependent on, like, your personal background, right? Like, like, for example, I don't invest in public stocks. Most people would say that you're insane. Like, that's, like, crazy. Like, I just got 2 million in d- dividends, I just hold it in cash, 100% in cash. But the reason I do that is 'cause I have access to deals that no one has access to. Like, the public does not have access to startups, right? So it's sort of like a lot of these, you know, these things are sort of dependent on that. Like, like, why would I invest in Figma as a public company if I could invest in Figma as a private company? So all this stuff matters when you're making decisions for yourself that, you know, that's just kind of part of things. So yeah, I, I, I think you don't wanna be extreme. Like, one person, you know, company is not necessarily the goal. Um, it's sort of a vanity metric, right? It's kind of cool fun. It definitely gets eyeballs, so, like, if there's ever, like, a, a Time cover of me, it'll be something along those lines just 'cause that's kinda, like, what people read and click. I think having a team is great. Like, the, the... one of the nicest things about having profitable business, you can hire people, right? You can be like, "Oh, I really wanna work with this person." Boom, I now can talk to this person for an hour every month because I hired them to do this task, and I get to talk, talk to them about this specific thing, and it's, it's, like, kind- almost like a, like a hiring a therapist or something, you know, to, to, to, to some, to some degree.

    16. AGAakash Gupta44:45

        Yeah. As soon as you build that one-person company that is making, like, 500K or a million, [laughs] I think everybody will cross that point. They'll realize it as well. Being a, a solopreneur business is kind of overrated, but I don't think what's overrated is embracing AI

16. 45:00 – 49:00

    AI Won't Create More Engineers

    1. AGAakash Gupta45:00

       in every part of your business, and being at the front edge of how to use AI in your business. Like, someone I was talking to, he has a huge consultancy. So what he's done is he's created a bunch of AI agents that take the notes, that take the advice he's given his prior mentees [laughs] and then simulates for him a draft of what he would want to say after a meeting. So there's all these different workflows, whether you're in a consulting business or a product business like we just demoed all day today, where I think being on the edge of AI is still really useful.

    2. SLSahil Lavingia45:29

       Totally. I mean, at the end of the day, it, it is, like, the technology of our time, so, like, I would have learned to drive in 1920s. Now you have to learn to use AI, right? So if you wanna be in tech, that's, like, what being in tech is, definition.

    3. AGAakash Gupta45:40

       So you have a legendary Twitter account. That's what made me a fanboy back in 2021 as we were talking about. It's a lot of people. I think you have over 360,000 followers at this point on Twitter. So it sounds like that's helped you in a couple different ways, right? It's helped you with your deal flow, it sounds like, for your angel investing, which is the major thing you're doing with those dividends. It's obviously helped you grow Gumroad, and you probably get some creator monetization. Are those the three main areas of benefit, or how else do you think about that?

    4. SLSahil Lavingia46:08

       I, I, I think it's g- really good feedback. So I think most people don't have access to the number of people to give you feedback. So, like, imagine you have an idea. Your idea's gonna be somewhat reasonable, but, like, if you have 100 people give you feedback on the idea, whether it's a book or a chapter or web application or a product idea or a startup or a business or philosophy or politics or whatever, I just think I- my ideas are better because I just get so much more feedback about them.

    5. AGAakash Gupta46:32

       Mm.

    6. SLSahil Lavingia46:32

       So I think that it's sort of like market feedback in some kind of way. Um, you just have a lot more of that. Uh, and so that's a real, really helpful thing. I think it's really helpful if I move to a new city, I sort of have, like, a built-in social network, which is really nice. I generate a lot of really cool, smart, hardworking, successful s- people. Um, I can sort of, like, pre-filter them just based on what they say on the internet, like, whether I would like to hang out with them or not. So that's cool. But to be honest, that's about it. Like, I, I, I actually think it's, like, never been less useful. It's actually really... And that's because the algorithms are really good these days, right? Like, you can sort of think about it like the better the algorithms get, the, like, more useless your audience is because, in theory, a really good algorithm means that, like, a new person with a really good piece of content but no audience will get significant reach, uh, which can happen. So I think now more than ever, like, the sort of incumbent advantage is very low for content. Uh, for other things it's still high, but, like, for, like, being... Like, for example, if you're a new mu- musician today, it's, like, the best time ever to be a new musician. It's never, never been a better time. Mus- film- m- filmmaker, are you kidding me? Like, a new filmmaker today. The only problem is you have so much competition because so many people also have s- access to the same tools. So just, like, the number of people, but, but most people frankly don't actually compete that seriously, so you're not really... Like, the pool of people, like, doesn't grow as, uh, proportionally to the, how much easier it is, right? So even if it's 1,000 times easier to build web applications today than, like, five years ago or 10 years ago or 20 years ago, whatever, you don't see that m- number of people doing it, right? Like, roughly a similar number of engineers today than a hun- than 10 years ago, even though it's much easier to deploy something to the internet today, like, a lot easier, and that's because people just have personal preferences what they wanna do, what they wanna work on, how they wanna spend their time. Uh, so I, I think a lot about that too, like, how does AI, like, change things, but also how does AI not change things, right? Like, if AI makes it in- incredibly easy to, to code-You're not gonna see more software engineers. You may- you're actually probably gonna see a decline in software engineers because the people who wanted to become software engineers already did. So, like, actually the hardness may be something that attracted them to it, that, like, if it's no longer hard, they may not be attracted to it anymore. Uh, so yeah, fun stuff to think about.

    7. AGAakash Gupta48:34

       I was gonna ask after that, and since we're at time, this is pretty much our last question here. Would you recommend people who are trying to create, you know, a really lean AI-first team to be investing in building a personal brand and building in public?

    8. SLSahil Lavingia48:46

       I mean, I think the most important thing is to get to know your customers, right? If your customers there, are there and that's a way to reach them, fine. Uh, if you're doing venture, fine, because your customer is the VC to some degree, so that's important. But I would say most people doing it, like, they, they, they're, they're doing it 'cause

17. 49:00 – 51:00

    Building Reputation vs Follower Count

    1. SLSahil Lavingia49:00

       they want to. They're not doing it because there's, like, a business reason to do it, right? Like, I'm not tweeting because there's a business reason. I didn't, like, come up with my, you know, like, some sort of strategy, right? Like, I was just having fun. I was, like, a young kid in high school, and I wanted to talk to other people building stuff, and the only way to do that was Twitter at that time. And I was living in Singapore. There's, like, no way to hang out with these people in Singapore. So, you know, I created a Twitter account, and there's, like, 100 people from Singapore on Twitter or something. And, uh, and may- maybe it's not Twitter today. Maybe it's something else. I mean, the most important thing is you wanna get to, you wanna build a reputation of someone who is helpful and smart, right? Like, like, this person helped me in this way, right? Uh, that's generally, hopefully, why I have followers. I don't have followers because I said something witty. Hopefully, I said... I assume, like, people are like, "Oh, Sahil, like, wrote this essay, and it helped me clarify my thinking in this way, and, and I'm gonna follow him, and maybe he has some other interesting stuff," right? Like, so I think it's sort of like that's kinda how I think about follower counts generally is just, like, this person helped me. And so I, I, that's kinda how I think about it. It can be in-person help. It can be offline help. It can be parental help, friends. The goal should be just, like, to help people. And, like, if you help people, then you will have an audience, right? You may not have a follower count, right? Like, you may not even have... Uh, but, like, more people will show up to your birthday, right? Like, more people will show up to your book launch. Like, I think people, like, you know, there's, there's a big difference in, like, pe- the followers and, and, and, like, genuine, like, fanatics, you know? Like, like, if I walk down the street at Times Square, no one's gonna notice, uh, even though I may have a lot of followers. And then someone with very few followers but is famous, you know, like, will, you know, people will notice, right? Like George Clooney, I don't know if he has a Twitter account, but, like, he would probably be recognized on the street. And so I th- I think, you know, sort of there's, like, fame. And I, I think people sort of proxy, like, followers on Twitter to fame or something like that, but there's zero correlation, right? It's, it's, uh, it's kind of a, like, weird... It's s- I, I think of it like you're kind of, like, on the other side of, of the Vegas strip. You're, like, on the productions line you're not supposed to see, but there's, like, all the creators are hanging out.

    2. AGAakash Gupta50:51

       [laughs]

    3. SLSahil Lavingia50:51

       And, uh, all the, you know, business owners and all the, like, grabbing a drink at the bar, all the cab drivers. Like, that's kinda like Twitter. Um, it's like, it's like how many of those people

18. 51:00 – 52:26

    Outro

    1. SLSahil Lavingia51:00

       know who you are, right? Like, how many, uh... It's sort of this, like, meta, the meta world. Uh, it's sort of like the s- it's sort of the deep state almost. Like, it's just the people on Twitter are the deep state, basically.

    2. AGAakash Gupta51:10

       [laughs] Or maybe they're all just really shy like me, and when they do see you in person, they just don't go home out of value. [laughs]

    3. SLSahil Lavingia51:19

       Yeah, I, that could be true to some degree. Uh, I, I do notice, like, if I wear a Gumroad shirt, uh, I get recognized a lot more. But my guess-

    4. AGAakash Gupta51:26

       Yeah

    5. SLSahil Lavingia51:26

       ... is I get recognized somewhat similarly, but, but people feel comfortable because I'm wearing the Gumroad shirt that it's kinda like it's okay to say you know me 'cause I'm the one saying I'm okay with this sort of, you know, this identification. Uh, so yeah.

    6. AGAakash Gupta51:38

       Yeah. Totally.

    7. SLSahil Lavingia51:39

       Yeah. It's funny.

    8. AGAakash Gupta51:39

       We could talk for hours. Thank you so much for your time today, Sahil.

    9. SLSahil Lavingia51:43

       You're welcome.

    10. AGAakash Gupta51:43

        I really appreciate it.

    11. SLSahil Lavingia51:44

        Hopefully it's helpful to people.

    12. AGAakash Gupta51:45

        All righty. See y'all later. Bye. So if you wanna learn more about how to shift to this way of working, check out our full conversation on Apple or Spotify Podcasts. And if you want the actual documents that we showed, the tools and frameworks and public links, be sure to check out my newsletter post with all of the details. Finally, thank you so much for watching. It would really mean a lot if you could make sure you are subscribed on YouTube, following on Apple or Spotify Podcasts, and leave us a review on those platforms. That really helps grow the podcast and support our work so that we can do bigger and better productions. I'll see you in the next one.

Episode duration: 52:36