AI CEO: How To Make A $10M Business With AI Employees (Amjad Masad, CEO of @replit)

Welcome back to the Big Deal Podcast. I'm Cody Sanchez. Today, if you want to make money online and you want to do it building with code and are curious about AI, but even if you've never coded before, buckle up for this episode because I'm obsessed.

So, I've been so excited for this podcast because I'm sort of a tech Neanderthal, but I've been going down a rabbit hole of AI. I hired a guy who's actually good at it. And it almost feels to me like there's a world before I realized all the things we could do online, and now there's a world afterwards. And almost everything my employees do, everything I do in business, I'm like, "Wow, why are we doing it the way we did last year?" Because it almost doesn't make sense anymore.

And what's crazy to me about you is you figured this out at like six, I think. Like, is it a true story that when you were six years old in Jordan, you were already coding and building software apps?

Yeah. Yeah. Yeah. So, my my father uh got a computer before anyone I knew back in Aman and Jordan in the early '90s. And so, one of my really earliest vivid memories was standing behind the shoulders uh of my dad kind of unboxing this machine, pulling out this huge manual, open it up, and you had to read a manual. There's no internet at the time and sort of finger typing on the keyboard and I was just fascinated by it. This idea that you can like talk to this machine and like convey what you want to build and you can build it. I'm still, you know, it's just still gives me goosebumps. Just the idea of like creation, you being able to be creative without a lot of resources or, you know, you don't have to build a real physical thing. All you need is this machine. And so I got into it immediately and started building apps. The first thing that I built was um a math teaching app for my younger brother. So it will be like a, you know, formula. You know, 1 + 1 equals blank. He puts in two, he gets a clap. He puts in three, he gets a boo. And it worked out. Now he works for me. So wow.

I love that. I don't know what kind of wedgies they had in Jordan, but that's like I had, you know, Lincoln Logs and Legos. Um, but then, you know, what's kind of crazy is then you had a 180 basically from that, which is, is it true that afterwards, not so long ago, you actually turned down a billion dollars uh as an acquisition price for your company when you only had six employees? What happened?

Yeah. Um, you know, we we had uh uh started the company in 2016 and the mission behind the company is to create a billion coders, billion programmers, billion developers, whatever you want to call it. And I just felt in my heart of hearts that we're at a point where anyone can make software. Uh, I felt like the AI is progressing really fast. Just cloud systems, everything that was coming online made it so that anyone could really be able to do this thing that has fundamentally changed my life, you know, got me from from Jordan in a sort of lower income family all the way to the US, become an entrepreneur, joining Y Combinator, the most prestigious startup accelerator, and uh in 2018. Just after we did Y Combinator, we kind of bootstrapped for a couple years, then we did Y Combinator, um, maybe 2019, we got this acquisition offer from a really big big company. Um, and uh, obviously it was a lot of money and I, I, we didn't, we didn't have, I didn't have a lot of money at the time, but I just, I just felt like they're going to take the, they're going to take the company, not going to do anything interesting with it, and and I'll make the money and maybe I'll quit two or three years after that and create the same company. So, I thought, I like this company, I might as well keep going. And I always felt like if we get to a point where we achieve even 1% of our mission, this could be a much bigger company. And and, you know, we achieved that valuation a few years ago and the company's probably valued more today.

Walk me through that moment. So like you get an offer, you meet a guy, I don't know how it works. Yeah. And did you seriously contemplate it? Were you like, I need to sleep on this?

Yeah. Yeah, I did. I did. Uh, it wasn't like a knee-jerk reaction, kind of like, a billion dollars, I'm just going to. And I talked to our investors and I remember talking to Mark Andreessen, who's uh invented the web browser Netscape and he's now running Andreessen Horowitz, uh, very prestigious investment firm. They hadn't just invested in Replet. He was like, "Look, you know, we'll pocket the profit. You know, we just invested in you and we're going to make a lot of money, but what do you want? Like, what are you going to do afterwards? Do you want to go work for this big tech company and become like a corporate employee, or do you want to keep going on your mission?" Uh, and I talked to a bunch of other people and I, I kind of felt like the weight of Silicon Valley was on my shoulders. Like, everyone really believed in what we're doing. Um, and I didn't want to let people down. I also just kind of didn't want to let our users down, our customers, our dreams. And um, and I also felt like I have the kind of skills to always be able to make a good living. Uh, and I think, you know, we all know that after a certain level of wealth, you know, your life doesn't really change all that much. Uh, and so I still, I like money. I don't want to make it seem like I, I think money is a great thing. I think it gives you a lot of freedom um and and comfort in order to achieve more things. And and I think the thing that I'm most attracted to is is is building great things. Yeah. Yeah.

Because, you know, it's not like now Replet seems like this this gargantuan, right? Like 100 million in annual recurring revenue. I look at your little chart and just, whoa. I mean, it's pretty wild to look at. But it took you nine years to even get to your first $10 million in revenue, right? Like, you you weren't really making any money at all for multiple years. You were burning cash, first your own, then a little bit of YC, then a little bit of third-party investors. It's not like this was an overnight success at all. And so I, I'm curious, did you ever have a moment where you were like, "We are not going to make it, and I'm not good enough, and this company is not going to work out?"

Yeah. Yeah. I mean, in uh 2024, early 2024, so last year, um the company had not been growing all that much. We're actually, we're kind of hit a plateau and started decreasing a little bit. So we were in this awkward position where we make it easy enough to get started, but not enough to get you all the way to building a business or real piece of software. And because our focus is on getting everyone to make applications, anyone with a dream without any skill or background in computer science should be able to come to Replet and make an app. That was always the kind of the the core of the mission. We're also not very good for professional developers because they needed all these tools that we didn't have. And so we were in this like awkward middle and not making any revenue and have this really large staff, 130 people, had raised $200 million. And as you know, you know, you uh you talk about it in the book, like, you know, there are businesses that you could build that provide a lot of uh, you know, profit and good living for yourself and you have complete, you know, control of what you do. But once you take on venture capital, there's almost a contract uh you're implicitly kind of getting into, which is, I'm going to build a billion-dollar company. This is this, I'm going to build something really scalable, really big. And I take that uh responsibility seriously as well. So um, I felt like the only way to really achieve our mission and for the business to take off is to invent an automated coder so that you go to Replet and you really don't have to learn to code at all. And so we laid off almost half the team. We moved out of San Francisco because I, you know, San Francisco is such a such a bubble. I'm sure kind of LA is the same here, but everyone's talking all the time. There's FOMO all the time. Everyone's going to like parties. And we moved out of uh San Francisco. We go down to the South Bay. We get, we get this office in the suburbs and we're like, "Okay, we're going to spend three, four months trying to invent this new thing." And we felt that technology is sort of getting there in terms of AI and if it works, it works. If it doesn't, like I don't know what I'm going to do afterwards. Maybe look for an acquisition or something like that. Uh, and it wouldn't have been a pretty nice acquisition. It wouldn't have been a billion-dollar acquisition. Um, so we start building it and I think even two weeks in, we felt like we had something that could change the world. And so we launched in September 2024 and it was the first coding agent on the market where you can just like put a prompt and it will like make you an application. It will create a database for you. It will like put it out, deploy it and scale it for you, do all of that. And it went viral and we just put out like a, you know, small video, very low production on cell phone, and it went super viral. Um, and our revenue that year climbed to $10 million ARR and then when we got out of beta and the agent got really good, that's when we went to $100 million ARR in a matter of six, seven months.

That's wild. You know what's interesting though, and why I was so attracted to what you do is because it could sound corny to say, "I want to create a billion coders." You know, you're like, "Okay, Silicon Valley, typical mission, put a billion in front." Yeah.

Exact. Exactly. um except I've used the product and like not an ad, you know, but um it's incredible, especially for somebody who's really not tech advanced. And um, and I really associated to it because I think, you know, our mission at Contrarian Thinking is to create millions of owners. And I think that sounds cheesy to people who um would rather have, I don't know, a couple billion dollars in assets under management as an investment firm instead. And I think there are a lot of people that say, "No, people who want AI just want to concentrate all the power." And you're kind of doing the opposite of it. So why don't you explain to people what exactly Replet is and why it's revolutionary, and maybe even how could somebody use it today to to change their life?

Yeah. Look, technology had always been had two visions. The massive centralized mega corporation, mega government surveillance state vision. And the dystopian future, you know, that's sort of like the Matrix and all these different movies. But there was always also a vision of the sort of the hacker in their basement that are building amazing things, or the star founders in their garage, or the Bitcoin crypto hackers who can like create systems of money and encryption outside of the control of governments and things like that. So that vision existed always. And um at any given time, like with the internet, it started as this idea of decentralized, empower the people thing, and then over time, we centralized on kind of a few big tech companies and things like that. And now we're at a moment of time where there could be another shift, and that shift could send us into more centralization or bring back power to people so that we can have more owners, more founders, more entrepreneurs, a more dynamic economy. And this is what I'm really passionate about. Um, so what Replet is, um, it's an automated AI coder. So, uh, everyone has ideas. Like I talk to people, I talk to Uber drivers all the time. Everyone has an idea. Everyone is kind of a, everyone in any field, they are domain experts, right? Um, but they usually do not have the coding skills in order to build an online business or or some kind of business using technology. Uh, because they don't have the skills. So they usually go and they try to find an agency to do it for them, and they get quoted an enormous amount of money, hundreds, thousands, a million dollars, we sometimes hear, and they give up on that dream because who has that kind of money? Uh, some people go, go into debt, and they still get screwed by the freelancers and all of that, or their ideas are, are, you know, maybe perhaps need a lot of iteration, and they don't have the money for it. So you go to Replet, you put in, you, you know, uh, start with a simplified version of your idea, get an MVP or a prototype, minimum viable product, um, and you're going to see the AI coder coding in front of you. It'll spend 10 minutes kind of researching, showing you the design. Once you approve the design, it will start building the application, and you'll get a prototype in a matter of 10 minutes. And then you'll look at it. It'll say, you know, I like this. It has this functionality. Okay, let's add, um, uh, you know, let's, you know, let's change the color of this button, but also let's add an entirely new feature to the website, and I'll spend another five minutes working on this and getting it done for you. And then um, you know, you can spend weeks building that business and then launching it. And we hear from entrepreneurs that the other day, I was talking to someone who works in venture capital. This is the, you know, the, uh, what I was talking about where everyone has domain knowledge. This VC, he's not even in on the investing team. He's on the sort of finance team, uh, understands all the problems and woes of managing a portfolio of companies, and they buy all sorts of software, none of the software works for them, right? I'm sure you can relate. Um, and so he goes into Replet, spends weeks, you know, nights and weekends, uh, you know, 40, 50, 60 hours a week, building his dream platform to manage his, uh, his investments. And then he realizes that he has a business on his hands. And he goes, goes and talks to different finance people at different VC companies. They all have the same problems, and he starts showing them software, and everyone was like, "We want to buy it now." And he got, he got commitments for $5 million ARR before he quit his job. So in a matter of months, on the matter of weeks, he had a business on his hands, and now he quit his job, and he has the contracts, and the money is flowing in. So that's, everyone has this. Think about it as, you know, when Airbnb came on the scene, they found this this asset that wasn't monetized. Everyone has space in their home that wasn't monetized. I think everyone has domain knowledge that is not monetized. And the way to monetize that domain knowledge, to scale yourself, you talk about it in your book, the the difference between an owner and a worker is the worker kind of trades time for money. I think I think owners often trade experience for scalable experience for money, right? So, putting your experience into a piece of software, and that can scale to hundreds, perhaps thousands, perhaps millions of customers, and you're going to be able to make a lot of money.

It's so true. I mean, you know, what's interesting is, you know, if somebody's never played with with AI before, you know, to me, what it looks like is you just, you kind of write, or I'm, I now am a voice-to-text annoying person who uses like WhisperFlow, and I just, you know, I just say out loud the things that I like, you know, that I'd like to build. I would like to build, you know, this exact thing. I want to build a, you know, portfolio dashboard, and I want it to look like XYZ, and I'd like it to be monochromatic, and there to be like very simplistic, you know, buttons, and blah, blah, blah, blah, blah. And you sort of iterate on this continuously. And then on the right-hand side, and we can show it on YouTube if you guys want to watch the video there, it kind of looks to me, because I am not a coder, like what I imagine the Matrix might look like in some ways. And so you have multicolored code that's sort of happening, and you can sort of then you could be over on the left watching Real Housewives or whatever show you watch, and simultaneously this agent is running. And so you have a little employee for the first time ever that you get for a couple hundred bucks a month that kind of does exactly what you want it to do continuously. And so, so I think you might have just answered this question in some way, but do you think that it is now possible for a single individual in the case of a few weeks who has zero coding experience to build a million-dollar business with Replet?

Yeah. So, we have um another entrepreneur, his name is John Cheney, and had this idea. He he has some domain knowledge on how to use AI for different kinds of things, marketing and storytelling and things like that. And um uh but he, he hasn't coded ever. He hasn't used AI coding tools, and he wanted to build an education platform to do AI training. AI training for consumers or AI training for enterprise. So he wanted to build an entire learning management system and seat-based enterprise software, and he did it in a matter of of weeks and he reached a million-plus run rate uh in in a matter of weeks. Now, look, I don't want to make it seem easy. And again, I, on on the way here, I was reading your book, which was really inspiring. And in one thing, you say it's not easy. Entrepreneurship is not easy. A lot of people when I come on these podcasts, I've learned that you, a lot of people think that I'm selling magic. I'm not selling magic. Like you're going to go there, you're going to get, you know, the agent is amazing, right? And and it's coding for you. And you can go look at the code if you want to learn how it works or ask it how this code works or whatever. But uh, you can also be outside of the code and be acting as a manager. And management is hard. You need to uh think clearly. You need to communicate clearly, and it's going to get frustrating. The agent AI is going to make uh errors, and you'll be able to correct it. Uh, and it's going to require some resourcefulness. You're probably going to have to do some research. You're going to have to tinker and iterate and spend hours on this. So I, I think, you know, the the people who are successful with entrepreneurship usually have reached a point in their life where something changes in them, and either the the boss they hate, or or the stresses of money in their lives, or just like the routine work that they hate, and they want to change, they want more thrill in their lives, and they make that decision and they're like, "I'm just going to build something for myself. I want to be an owner." And after you've made that decision, you're just going to, you're you're just going to do whatever it takes to get it done, right? And I think this is where Replet can help you. But you need to have the grit and the persistence in order to do it. Yeah.

Well, I think the other part that's interesting about what you guys have built is a lot of people who build AI tools or tools with AI or things to help with coding historically have been for the sophisticated professional few. Yes. Which would be similar to finance. It's like in the elite class, you want to speak only to investment bankers. You want to speak only to those on Wall Street. And if you were to normalize finance, investing, well, you know that that's too simplistic, that's too low-level, and you know, those people, it's a little bit beneath us to talk to them. Um, and I think a lot of times it seems that there's similarities in engineering and code where it's like, well, oh, vibe coding, which is what what people now call this, vibe coding, this ability to code without having deep experience and being an engineer, you just go, you just go off vibes. Yeah. Right. I was trying to, how would you even define vibe coding, which is now what people only talk about vibe coding on on X now, I feel like.

That's right. That's right. Yeah. So vibe coding, I would say, is previously, previously, you know, when you have an idea and you want to turn it into code, right? You're turning business logic into hard mathematical constructs, right? And most of your time is spent on that construct, on the syntax of doing that, which is not really the problem-solving that you want to be doing. Right? Programmers spend maybe 10, 20% of their time actually thinking about users, the product, and then they spend 80% of their time fighting with the systems, figuring out how to deploy it and scale it and run the database and all of that. So we're trying to take care of all of this nonsense, right? You don't have to worry about it. And the idea behind vibe code, which I, you know, I I don't think it's the best term, but we're kind of stuck with it, is that you, um, you can trust that the AI is making the right decisions based on the vibes that it's giving you. So, you are communicating the the business uh goals that you intend to create, and the system is building this on your behalf. And if the vibes are correct, and if you can test it and feel like it's working, you can trust that it works. And, you know, 80, 90% of the time, it does.

So is it a bit like building a house? So like, before we knew how to build foundations and we knew how to do um, you know, traditional home building, that might have been the unique character trait in in a a home builder, which is like, they can do this thing that we've never done before. They can make a house withstand over time. They can build it correctly. That is actually the most important part in home building. Whereas now today, we kind of that's normalized. We've commoditized the fact that yes, we can build houses. Yes, they can withstand things. They can function properly. And now the part that's non-commoditized is essentially design. It's like the people who make the most money are the ones that make the user interface, aka, you know, the aesthetic of it. Um, and new functionality on top of it are the ones that get paid more. Is that kind of what you see in the future? Before it used to be engineers, like, can you even build this? And now it's, no, is it beautiful? Is it user-friendly? Can you sell it?

I, I think your business is is a good illustration of this. Like social media and being able to build this show. Previously, you needed to get on some kind of TV channel, like, you know, 50, 100 years ago, even 20, 30 years ago, you needed some kind of cable channel. You needed some kind of studio to take a bet on you. And now you can go to YouTube, two clicks, you have a channel. And so what's what's left? You know, that's a commodity, right? Like having a presence online is a commodity. What's left is charisma, is creativity, is grit. Again, grit is a thing, right? Like you have to keep going. You just keep showing up. And, you know, a lot of entrepreneurs will tell you that. It's a truism. It's a cliché, but it's true. You have to show show up day after day after day. Learn, become 1% better every day, which adds up to an enormous amount of progress in a matter of months or or a year. Um, and if you do that consistently, you're going to be at the top of your game. You're going to be able to stand away out from the pack because, look, most people are lazy. Most people just don't want to do a lot of work if they're happy with with the, which is nothing wrong with that. No judgment. Like, you know, a lot of people work hard and they want to, you know, spend time with their with their family. They don't want to change too much things about their lives. They're probably like happier and more well-adjusted than we are. Yeah.

I mean, there's there's something to that. I think entrepreneurs, there's always something either a chip on the shoulder or some some dark past or something like that. Yeah. One of my friends had the best line. He was like, "Entrepreneurship is a trauma response." And I was like, "You're probably right." That's probably true. Yeah. There's you feel like you have something to prove or something. But it could it could fundamentally change your life to to the better. And I think the, you know, what's so so similarly to YouTube, uh, now in in in software and apps and business making and online businesses, it is the domain knowledge. I think is the main thing. If you understand something really deeply, or you can go learn it really deeply, you can go build a business around it, right? We have stories of people who work on event management. Um, and, you know, this woman, she knows exactly uh, what it needs to create an event, you know, where to find the, um, the venue, the caterer, the marketing, the everything. And and she created an AI agent that uh, that can go do all of that. You put in a prompt, you'll say, "I want to create an event around X, Y, and Z." It'll create a plan. I'll go, you know, get you in contact with the vendors, and and that she turned a non-scalable business that she had, which is helping people with events, to a fully scalable AI solution. Then she gets paid for while she sleeps. Yeah.

It's wild. I, um, one of my biggest annoyances in all my businesses is dashboards. It's always like, it's always hard to find the exact dashboard how my brain works that I want to see stats because I always want to have like scoreboard-based leadership, right? I can tell instantly, are we behind? Are we in front? What's happening here? Yeah.

I wake up in the morning, I need to look at my dashboard.

Me too. Otherwise, I I just end up bothering everybody all the time asking them where shit's at. Right. And so, if I can see it, I'm like, "Oh, here's the project management. I can see if you're red, yellow, green, are we moving forward?" And also, like, is our business on track or not, etc. And so, I built one one with you guys. And um, so anyway, I'll, I'll show it here. It's not pretty. And I'm an idiot at this stuff. So, you know, if I can do it, that that anybody can do it. I'm John. And every CEO I know who is serious is obsessed with being able to see their business in one dashboard. You heard him say he couldn't wake up in the morning without looking at the dashboard to figure out if his business was working or not. I'm the exact same way. I do it all on my phone in something that we call the Contrarian Operating System. It's a dashboard that we teach people to build in the boardroom step by step so that your business can be seen in 30 seconds whether you're going to hit your goals or not. I think CEOs that run businesses without an operating system and a dashboard is kind of like driving a car at high speed down a highway without knowing how fast you're going, without knowing what the GPS says to do, and also without even being able to see if you're in park, neutral, reverse, or drive. And so I am obsessed with what we've built in the boardroom. I think every single business owner is going to want to figure out how to get this operating system and how to get with a group of people who help you implement it. If this is interesting to you, this is all we do in the boardroom is obsess on getting owners to get to the next level with the tools and team they need to hit their next $10 million in revenue. If that's interesting to you, you can click the link and get on a consulting call with my team.

So, if you're like a young person who hasn't figured out how to make money yet and you don't know what you want to do with your career, but you're curious about this AI thing, what would you tell them to do today? Where does a young person start?

I would say do the same thing that you've always done. You know, get a job, find an industry you like or you're interested in, and go into that. And what you're going to find is massive inefficiencies. People do things still like, you know, there's still faxes around in many businesses. People use pen and paper for so many things. Old software still exists, you know, um, there's so many like whatever field you go into, just have an eye for what people around you are doing and and what are the things that could be automated or what AI could be good at. And the way to learn what AI could be good at is to spend a lot of time with GPT, spend a lot of time messing around with video AI, messing around with Replet and code AI, just be plugged into the AI world, and you're going to build an intuition. Just jump into it. You're going to build an intuition. At the same time, find a real-world problem. And the way to do that is to go literally into any domain you're interested in and just open your eyes, and you're going to find a problem to solve, and then solve that problem and sell it to the people. I mean, um, there's this story, uh, I just want to show also that entrepreneurship is not always about taking massive risk, quitting your job, and and going going for broke. There's this uh guy, his name is Ahmad George. He works in a skincare company in in DC, and he's responsible for their operations systems, and he realizes that, you know, we, you know, there's 20, 30, 40 hours a week spent between our team, and we probably have more people than we need to do these certain operational things, um, managing inventory, just the standard stuff. And so he goes to their software provider and asks them, "Hey, I have this idea on how to automate this." What does it take to make a software like that? And they quote him $150,000. And he learned about Replet online, and he goes in, and in a couple weeks, he spent like $400 on the AI, and he makes a piece of software. He gives it to to his colleagues, and they're suddenly cutting it, uh, hours down. And and he goes to his boss and he was like, "Look what I did." And, you know, I, you know, I saved this company this much money. And his boss is now paying him $30,000 a year for the software. He didn't have to quit his job. He's an entrepreneur within his company.

I love that. I mean, I would do the same thing. We actually, this this week, I told one of our employees that I was coming on with you, and he was like, "Okay, cool." And it was very simple. We have an event that's coming up. I told you about it in in November. And um, I was kind of waiting for the team to to build out the event page and get it connected to HubSpot and do all the backend nonsense and get it mocked up and blah, blah, blah. And I think the quote I got was like $5,000 or something like that for this website. He was like, "Well, just for giggles." So he goes to to Replet, he actually puts together like the entire interface of of the design that he wants. And then he used another program, I can't remember what the other one was, in order like to actually actualize it and then connect it with with HubSpot. But essentially using two programs, one being Replet, he built the entire website, and he said he took about four and a half hours. I think I was quoted four weeks for it. And um, and so now I find myself very annoyingly to my employees often saying like, "What's our 24-hour rule?" You know, anything that used to take weeks, we should just test if we can do it in 24 hours with AI because then I don't have to be smarter than anybody else. I don't have to be richer than anybody else. I don't have to be better. I could just be faster. Yes. And like speed wins every business game, I think. And, you know, we're in a moment of potential arbitrage, right? Like before everyone in the world learns that this technology exists, it'll probably take five years. You know, people are slow. You have a moment where you where it's like an opportunity to create a lot of value and capture a lot of value from it. So, um, you know, thinking about the kind of advice you give your uh, your audience around buying businesses and and running profitable businesses. I think now more than ever, finding businesses who are that are running inefficiently and going into these businesses and figure out where to plug in AI, where to replace SAS software that's expensive. You know, a lot of companies pay tens of thousands of dollars for software you can generate a Replet for $5 literally. And so you can go into any business and, you know, shave off a lot of cost and make it a lot more profitable even without growing it. Yeah.

Well, you guys should tell us in the comments because you told me you're like, "We should do a hackathon," and I said, "Well, can we do this simply enough where like anybody listening, if we gave, if we had them for two hours, let's say, they could build something?" And you said, "This is what we do." Like, "Yes, at the end of two hours, we will have everybody have built something that is useful, potentially even valuable and monetizable." This is our go-to-market strategy, actually. Interesting. So you guys, you should tell us in the comments if you guys want us to do a hackathon, and we could do one. We could do like a Main Street, "Everyday I Have No Idea How to Code" hackathon.

Yep. But you run these now and they work. Yeah. So we, um, so for example, HD Capital, a, you know, capital group in uh in Europe, has like hundreds of portfolio companies, and some of their portfolio companies were using Replet, and they were interested in it. And so we went there and we ran a hackathon for all their executives, their CEOs. And, you know, it's CEOs, you know, they're busy, they're probably getting texts every, you know, two minutes, and it's not like they can focus for the entire day. We told them, "Just give us two hours of your time." And in two hours, everyone had applications, and like, you know, they're so giddy about it, like kids, right? Like they, they come and they present what they built, and they're so excited. And that's that's what I love. It's like, there's a human aspect of it, like making things is awesome, and so everyone's going to get excited about it, and and most people are going to convert and become users and customers. We've confirmed that we can make money with AI. And yet, like, you see a future because you are, you saw this Jesus back when you were six, but then also you guys saw the vision for Replet really before AI was here at all. And one of the things that you said on on, well, X, uh, that you tweeted was that, "Crucially, it won't be prompting. We believe that's more a bug than a feature. In the future, it will be a combination of AI predicting what task you want done next and doing it for you plus a dialogue-based agent that follows your commands." Can you explain that to people, like, what is prompting, and also what do you see the next evolution of this being?

So, uh, prompting is the craft of creating an instruction for an AI. Right now, ChatGPT gives you the illusion that you're talking to a person, and it kind of works out for everyday life stuff. But if you want ChatGPT to go do massive amounts of research or build software for you, like in the case of Replet, or help you build a business, you need to be very precise. And often you need to do these psychological tricks, like tell it you're a world-class expert in, uh, laundromat business ownership. And I'm sure you've done this with ChatGPT, and then you basically get the, get the AI to assume, uh, a sort of a different character for the purpose of that conversation in order to give you more relevant information. Right? And a lot of people who has used Google for many years know also how this stuff works. Um, but but that's that, that's also still a little awkward, and that's not how we work. And the way we work, I, I think my vision for working with AI should be like working with a colleague. You don't go prompt your, you know, you might give them pep talks to your employees, but, you know, you're a world-class expert. You imagine how awkward that would be. Um, but the, uh, you know, what, what we're trying to work towards is for us to do the prompting for you. So having another AI kind of take your, take your what you've said and kind of, you know, transform it in a way to give it to another AI, pre-prompted. Uh, but right now, it would help like if you go to YouTube, we have our YouTube channel, it has some, some information on prompting. So I, I would suggest people learn prompting right now. But I think the way in the future is, um, the most important thing is going to be, uh, clarity of thought, being able to translate complex ideas into simple bullet points and paragraphs, what we call in kind of TAC PRD, uh, product requirement documents, and, um, being able to have a systems thinking, which I think is a, is a very useful tool, and I think it's something you can, you can learn. It's really, how do you break up large ideas, large, uh, concepts into smaller components that you can work on them in isolation? So for any business you want to build, for any tool you want to build out your business, it probably has a lot of different components, different pages, different. And that's the essence of system thinking. And so you can know that you can work on this component in isolation. Once I'm done with that, I need to go to the next component. Here's how I'm going to break up my app or my database into all these different things. And these are skills that can be learned. Like I would go to ChatGPT and spend a few hours learning some of these skills. You know, teach me how to be a better systems thinker, how to create PRDs, how to be a better communicator. Uh, but all the stuff will get easier in the future.

That's a great point though, because you literally use the tool to get better at the tool. Yes. And if you can understand sort of first principles of how the tool works in order to make it work better for you, then you know that would be like what X would call, you know, godlike prompting or whatever.

That's right. That's right. And also, uh, getting crisp about what you know. M. Uh, people know a lot of things. And, you know, I'm sure you've, you've had that experience where someone comes to you for advice, and you surprise yourself with the advice you're giving them, right? It's like, "Oh, I don't know how much of an expert I was at this." Which is why mentoring is good, not just for the mentee, but also for you. Like, I like to mentor people because it just reinforces what I know. And so a lot of people do not know what they know. And I think that skill of taking what you know and converting it into precise size documents is is very, very important.

Interesting. I think that's the skill of writing. It's very true. I, I do think you can often tell how good of a thinker somebody is by the way that they write. Yeah. Writing a book is torture, right? Like it is. I actually like it. You know what's the torture on writing a book is the the marketing of the book. Trying to, the New York Times bestseller was the most ludicrous like fake arbitrage. Yeah. I, I wish I could have honestly, if I would have, if I could have, I would have, but I didn't. I didn't. I wasn't intelligent enough to do that. But no, it's, it's a lot of, talk about an outdated system. It's a lot of, well, you need, uh, the the users not to go to the seven places that all users go to to buy books, but the 47 places that the New York Times actually stack ranks in order to determine that editorially the book is appropriate or not. And so it is, I'm sure there would be a way actually to engineer it now with AI and and figure out how to reverse engineer New York Times. That could be a business. Yeah, I'm sure it, there's probably somebody that does that for billionaires who write books, you know.

Um, I want to talk about, I want to get a little bit into like leadership and tech and talent. So we're right in the middle of like what appears to me to be like a great AI talent war. I've never seen anything like this. I mean, I have never seen people traded from one company to another for $100 million publicly, you know, or $300 million publicly. Um, I know there was a big drama last week around Windsurf, a competitor of yours, where they were, you know, close to a deal with OpenAI, and then that deal expired, and then Google kind of came in and aqua-hired, meaning that they, uh, acquired part of the the team as opposed to the entire company. Um, and then another company came in and acquired kind of what was left of Windsurf. I'm I'm curious like, one, is this the future in AI and and tech talent? Like, will people be acquiring individual talent instead of entire companies in this area? And like, what is your reaction as a CEO where you're competing for all of this talent too?

Yeah. Yeah. So, um, in corporate America, we've had this concept of of salary bands, right? You know, with the rise of HR, actually, you know, there's an interesting graph you can look up where, you know, HR was basically non-existent over the past 100 years, but as government accumulated all sorts of regulations, HR sort of is an extension of government in your business, in a way. Um, and there was a lot of, I think, good intention around discrimination and things like that that made it so that HR people wanted to make sure that everything's equitable and fair. And then there was this, you know, the rise of like salary bands, you know, these are the levels, and these are the

Salary bands for them, and that that's often based on experience and not based on impact. You can have someone straight out of college, they don't have a lot of experience, maybe they're not very good at communicating or working with others, which is a required skill, but this spike in one skill. They're very good programmers, especially like in, you know, a very niche thing no one else can do. And that could translate into millions of dollars for the company.

Um, and that's very, very different than, you know, being a factory worker, which is post-industrial revolution. That's how the system was built. You're sort of a cog in the machine. You can't, you don't have a lot of degrees of freedom unless you go up in the management chain and all that, in order to affect real change.

Um, sports team, and you mentioned the term "traded." You know, you can, you know, get Steph Curry on your team and put him in, um, you know, for, you know, five minutes a quarter or something like that, and have him shoot a few threes, and he'll swing the entire game, right? And, um, you know, I think in basketball, there's, 10 years ago, there's this three-point revolution that happened where, um, everyone realized that three-point shooters are way, way more valuable than everyone thought before because they can really swing the game rapidly. And I think right now, that's starting to translate into the business world because corporates is like, you know, HR be damned, like what matters right now is for us to win. For us to win, we need to get the best talent.

Um, those guys that are getting, um, that Meta is buying for hundreds of millions of dollars are getting paid more than the executives, more than the entire command chain together. Um, and so I think the bet there, and we'll see whether it pays off. The bet that Zuckerberg is making is that, in a way, Meta needs to pivot. Meta has this asset that's spitting out a ton of cash, which is Instagram, Facebook, their ad kind of networks. Uh, but it's, in a way, a dying asset in many ways. It's very competitive. Um, at some point, it's going to get fully disrupted. AI is going in. It's already making social media unusable. Like, you know, Instagram is full of like AI slop right now. So is Twitter, and now there's X. Um, and he realizes that he needs to be at the forefront of AI, and he thinks that we're going to reach superintelligence in the next two years. And he thinks if one or two researchers could invent the next major AI innovation that can translate into hundreds of billions, perhaps a trillion-dollar market cap for Meta, and he does this calculation, and it's like, "Oh, $100 million is nothing for that person."

Now, I can disagree with every part of this, and I do in some cases, not in terms of the comp, but like in terms of his chain of thinking. But it is valid if he really believes that's where the world is going to go.

Um, now, I think it's gotten a little out of hand in that, you know, the deal between Windsurf and Google, and, you know, other companies have done the same thing before, where companies are dying for AI talent. And what they do is they see companies out there that are performing well, maybe not excellently, like you when Surf had like, you know, $80 million ARR growing at, you know, AI companies grow so fast, but relative to an AI company, perhaps it's underperforming. But like, if you think about it as like a B2B SaaS company, it's like one of the fastest-growing ones, you know. So, but the company's doing well. It has a lot of customers. People like it. But for Google, $8 million is nothing. Nothing, you know. They'll spend it on corporate dinners or something like that, you know.

Yeah. They have amazing food there. I don't know if you've been. Slack lines? Yeah. Uh, gourmet sushi and all of that. Um, there's this story, an entrepreneur that went to work for Google. I think he was a part of Whimo, and he wrote, his, about, I love Google, by the way. I mean, they're great partners, but, you know, it's funny to talk about this stuff. Uh, he wrote when he left, he wrote about his experience at Google. Uh, and he titled the blog post, "Oh, Not Sushi Again." He says he was in line for food and someone was like, "Oh, not sushi again." Consider it sushi again.

Um, so the level, you know, of Silicon Valley is bizarre. Silicon Valley is crazy. By the way, the HBO show is like, at this point, very tame relative to what's happening in Silicon Valley right now. But anyways, they want all that talent. They don't care how much they're going to pay for it. And they go ahead and they basically give this large offer, essentially $2.4 billion for talent. Uh, they don't need the technology, they don't need the business. They'll take the people, the leadership, and the top researchers and AI engineers. They'll take them, they'll pay them the $2.5, pay off the investors, and then they'll leave a shell company behind. And that shell, they'll tell this shell company, "Oh, you know, you still have some asset. Figure it out."

Um, and then, um, and then, you know, we, you know, I put out a tweet saying, "If you're left behind, we'll kind of fast-track you in the interview and have you explore joining our team." And then a bunch of other companies jumped in, and they ended up getting acquired. And I think there's still a lot of people in Silicon Valley that care about doing the right thing. But I think it is destructive for the ecosystem because now startup employees are very doubtful. Like, if I'm joining an AI startup, like the founder could take a paycheck and leave the company and leave a shell of a company. That's scary. And, you know, and that's never happened before. That's never happened before. It's a strange structure to lift the talent of the company. I mean, is it like because they don't want antitrust issues? Is that really what this is coming to?

Yeah, during the Biden administration, Lena Khan, who is the, was it FCC chairman, uh, kind of created a lean, mean machine that was really blocking a lot of acquisitions and things like that. So it is partly financial engineering, right?

Yeah, I know I liked the piece that you put out on it, really kind of saying, "Hey, we don't do that at Replit," and like we've had chances for us to be bought, and I've had chances. You know, I'm sure you have too, to have been approached by people who wanted to acquire your big brain and your team. Yep. Yeah. And you tell them no. Yeah. We tell them no.

Um, and, you know, another thing we do at Replit is we do secondary sales. Not often, but we tell our employees every now and then, if there's an investor that really wants to get into Replit, we'll let employees sell their shares because, you know, we have such a long-term plan. You know, we've already, you know, eight, nine years in, and who knows when we're going to go public. And that's unfair to people that they're not going to be able to, they're taking a bat on you. They're taking a bat on the company. They're working a lot harder than they could be working at bigger companies, and we want to let them be able to take some risk off the table.

Um, and so we've done that before. We'll do it again. In terms of how we keep talent, one thing about Replit that we've also done well is we haven't exploded our valuation. Like, um, we'll probably raise in the future and update our valuation, but right now, our current valuation is less than 10x of our revenue. And by the way, like these technology companies, early growth AI companies are 100x. So we could be valued at $10 billion, or $20 billion. What happens when you value, or $20 billion? What happens is there's no more upside for employees joining you. So when people join Replit, we sell them on this upside. This could be a trillion-dollar company.

Um, and we want you to get on the, we're not trying to pump and dump, right? We're trying to build this in a way where people are taking appropriate risk and getting enough equity comp in order to for this to have a 10, 100x potential in the future. Yeah.

So, do you think in the future, big tech companies are essentially going to use startups as their R&D, but it will be the startup talent that is their R&D instead of the product? Because if I'm an employee, like, what is my incentive to join a startup when there's a risk of failure? There's a, you know, there's a risk of just like my health and my family life, and people kind of spend a lot of effort and work and all of that. There's a lot of sacrifice, and in the end, there's another new risk, which is a check comes from the sky and lifts your leadership team, and you're left with a shell of a company.

Yeah, it's pretty wild. And I think it's, you know, they just have such big paychecks or they have such big bank accounts that the paychecks that they can pay out, it would be a great way to skirt antitrust and also to make innovation quite hard. Yeah.

Um, yeah. And that's the thing about government, without getting too political, but often, you know, there's this book called "Economics in One Lesson." It's from the 1970s by some libertarian economist, and he talks about there's one thing you need to learn about economics: second-order effects. Everything that every policy has a second-order effect. So the policy of blocking monopoly via company acquisition, which, you know, seems like common sense, also means that you're blocking innovation entirely because they're going to find a way around it, and they're going to kill the startup ecosystem.

Yeah, it's such a good point. So, I mean, so what do you do? Is there like a secret WhatsApp group or something where like tech CEOs go, "Okay, Mark's about to raid us. Like, here's what's happening?"

There's a lot of WhatsApp groups. I'm sure you've seen some of the leaks that have been happening during the pandemic. A lot of tech people just felt, you know, during the great awakening, like this. Yeah, it was really hard to talk about anything online. You get attacked and canceled. The cancel culture was going crazy. So, a lot of the social groups, and especially people not meeting in person, went into WhatsApp. So, we do hear these kinds of things, all these rumors happening on Signal and WhatsApp pretty commonly. So, you can tell when something is happening.

But another thing about Replit is, and you and I just talked about it, where Windsurf and Coursera and others, they're going after the professional engineers. They have a target on their back because they're competing with big tech. They're competing on the same technology, on the same market. Whereas a lot of big tech kind of sees us more as a toy. We're like, "Oh, you let people without coding skills build things." And we're like, "Yeah, we're a toy. Don't worry about us."

And, you know, I've been very, you know, successful in the podcast circuit and, but, but we've actually not been covered by the New York Times or any of that stuff, right? We've been kind of, you know, because I want to talk to people directly. I want to be able to tell them about what we're building and how it could make their life better and how it could be useful for them. But we haven't been, you know, actually, everyone was surprised about the scale we're hitting because we don't have a PR. I've never had a PR company. I once hired them. They're never good. They're never good. I once, I never have had a good one yet. I once hired, the first time I wanted to do a news story, gave them $30,000 retainer. They got me a, they got our fundraise announcement into some publication I've never heard of. I looked in like Google Trends, whatever. I thought at best, they probably got a couple hundred eyeballs. And then I remember tweeted something about that, and that went more viral than the story.

Um, and since then, I'm like, "Okay, you know, I'm just going to be, I'm just going to be the PR guy." It's a good point. You know, I'm reading George Lucas's biography, and it's sort of interesting because Star Wars, you know, arguably one of the most popular and valuable franchises of all time. And what's interesting is when he came out with Star Wars, and really throughout his entire career of producing them, the feedback and critics say, you know, this is lowbrow, this is mass market, this is, you're basically, it's...

Oh, yeah. I mean, it's a fascinating biography just from the fact that inside there was all this terminal, you could create real art. Why isn't this more sophisticated? You know, you are actually, you know, you have this like cinematic view because he, that is how he started. He created like really cinematic, sort of sophisticated movies, but then he realized he's like, "No, no, I want to tell stories to everybody." Like, I don't care. That's fine. Say I sell toys. Say that I'm simplistic. No big deal, because my message will resound through time, and yours won't." And so I think there is something really interesting about if you can get comfortable with people thinking you aren't sophisticated.

I love being the underdog. I, you know, I don't, that's a problem because sometimes I'm like, uh, you know, I just have that, you know, because people view me now as like the, you know, big rich CEO, which is not my self-image. My self-image is the underdog, is like the struggling founder, and that gives me sort of more aggression and, and, you know, I treat everyone with kindness, but like, I view business as war in many cases. And so, but, but, you know, I love that. And I sometimes, like, you know, there's this meme with a lot of programmers, was like, "Who uses Replit?" It's like, "I love our users. They're the best. They're entrepreneurs. They're people that want to change their lives." And you're a big tech employee who's like, you know, works, you know, goes to the office, spends an hour eating lunch, and then half an hour making latte, and then like goes to the keyboard, you know, complains to your product manager, whatever. Maybe write a little bit of code. And you're, you know, judging, you know, who uses these tools that are very empowering. And, um, and, um, yeah, I, you know, I'm like, "Yeah, keep at it." You know, you're just fueling us, you know? Absolutely. There's nothing better than a good hater. You got to print that out and like put that on a poster board on the wall. Like, "Who uses Replit?" I always loved, I don't know, Alexis Ohanian very well, but anytime I've engaged with him, my favorite thing about him was that Yahoo quote. Remember? Did you ever hear this?

Oh my god, the story is so good. When he created Reddit, they went to go sell potentially to Yahoo, and they went to Yahoo, and they were like, all for the meeting, super excited, and basically were shunned off. And the line that he remembers is, "Reddit's just a rounding error to Yahoo." And so they printed it out and put it huge on the headquarters of their office, and they kept it. And now, who's a rounding error? Well, that would be Yahoo, not Reddit, right? That's right. And so I do think there is, you know, if you're struggling to remember that, man, there's nothing better than a chip on the shoulder. I would take a chip on the shoulder over like a point or two of IQ often. Yeah. You know, yeah. And I want our users and our entrepreneurs and our founders, people building on Replit, to feel the same way. Yeah. I want them to feel that, you know, people might underestimate me. People might think what I'm doing is wrong or useless or a waste of time, but goddamn it, I'm going to get it done. I'm going to change my life, and I'm going to become rich.

By the way, if you guys want to see what I've built with AI recently, go to cody.hat. This is wild. You can talk to me for free and get all of my insights on questions you have surrounding buying businesses, building businesses. You can go back and forth. You can call me, you can text me. This is fascinating. We built it in real time with one of our portfolio companies. I want all your feedback, and I want you to get free advice. So, go to cody.hat/chat, and you can learn more about AI real-time business buying. Super cool.

Okay, I want to talk about a couple things about that. Like, you've built this giant business. You went from not a lot of revenue to a ton of revenue, ups and downs. One, just like, what did that take? Like, what did it take to grow from a few employees to hundreds of employees to then be able to even manage $100 million in revenue? Because you have to be a totally different type of human to manage a company at that scale, especially that fast.

Yeah. Um, look, I think venture-scale companies sign you up for a lot of pain, and people don't talk about that because, you know, I was like, "Oh, no one's going to feel empathy for you. You're making a lot of money. You're rich, right?" Um, which is also what I love about your message is, hey, you don't have to build a startup. 90, 95% failure rate, and even if you make it, sometimes you sell off a lot percent of the company, you lose control. Steve Jobs famously got fired from Apple, right? I mean, it happens all the time. Venture-scale companies are hard. And if you want to do one, do one because you really love the mission and you want to build something big and you have the chip on the shoulder. But if you want to get rich, if you want to build a business, it is not the best way to go.

I think just coming to the Bay Area or California and living here, that in itself is a huge expense. And it's boring. It's boring out there. It's true. We, you know, all the house parties are the same. It's like, "AI is going to kill us," or "AI is not going to kill us." It's like everyone talks about the same [ __ ]. Um, I love Silicon Valley, but it's still, it's still kind of not, not a lot of fun. I come here for fun. Yeah, I bet. Actually, I bet. Oh, yeah. It's also the parties in Silicon Valley have to be like 87% dudes. 87% dudes. They smell really bad. And, you know, you go to these house parties, you have to put your shoes outside and go barefoot inside. It's just disgusting. I don't go, though. I spend time with my kids. Oh, that's true. Well, you come to Austin, and you actually, you were in Austin. You were on Rogan. So there, you know, like if you come to Austin, it's just like cold plunges, saunas, less I can go. That's my job. Red meat. Yeah, exactly.

You know what I loved that you did? I was going through your website, and you have a lot of stuff on there that is like an anti-signal. So basically, "Do not come to Replit and work for us if you have these traits." And I was, and then you have your operating guidelines or sort of your core principles. And I loved the simplicity with them. There were only three. And then I also loved the anti-signal, which is, "Please don't come. If you're going to come here and you don't like these things, you should probably leave." I want to talk about a few of those. Yeah. Um, I think we should, like, why don't we run through some of your operating principles? Like, you know, "Learn to like pain." That's not normal on very many companies to put that on your recruiting page. Yep. Yeah. So it's "seek pain." Yeah. Um, which is even worse. Is like, look, look for the pain, buy it. Right? Uh, is it an anime dungeon? Uh, it's not. It's a startup, but well, kind of depends on the day.

Um, so basically, a lot of things in life is about pain, and a lot of things in building a business is pain. Um, we, you know, as humans, we like to go through the path of least resistance, right? Um, we also often lie to ourselves. Uh, companies like to sit around and feel good about what they're doing and ignore all the signals of the things that are not really working. I think constantly seeking the, what's not working. So, when you see someone that's a user of your service, they're going to tell you how amazing it is. Um, and that feels great. But, you know, telling you what sucks about it doesn't feel great, but you should ask, "What sucks? What sucked about it? What went wrong? How could we be doing better?" It's painful to hear those things. And so, if you want to build a great product, you need to seek the painful signals. If you want to build a great business, you need to not just look at revenue growth, but you need to look at your churn, and you need to look at your spend and your cost and your all sorts of really unfun things that are often can be really bad that can lead to your demise eventually. And so you need to pay attention to these things, and these things are painful.

Um, there are things in every company that are, there are people that are not doing very well. There are people that are not a culture fit. There are people that need real feedback. Feedback is painful. Sitting down and telling someone that they're not doing well, especially if they have personality traits that make it hard to give them feedback. That's painful. You have to seek it. You have to seek the pain. And what I tell our team, counterintuitively, when you seek pain, you actually reduce it. Because typically, what people do is they procrastinate. You have an underperforming employee, you don't want to fire them, you don't want to deal with it. They're annoying, and I just want to just let them on the side. I'm just going to focus here. But what's going to happen is you're just building up a world of pain because you're not giving them real-time feedback. And by the time you have to fire them, they're going to be surprised, and maybe they're going to sue the company or something like that. And so anytime you delay pain, you're building up a larger debt of pain. So counterintuitively, seeking pain reduces pain, but you have to seek it.

So true. So what does that look like in your company? How do you actually make your employees do that or select people who are already predisposed to that? What like tactics or frameworks do you use?

So, just putting it out there filters a lot of people. Like in the interview, people were like, "What? It's weird. I don't want to be part of that." Uh, I want sushi. Yeah, exactly. I want sushi again.

Um, so, um, you know, for example, if you're a product manager, we'll describe your role. If you're a product manager, you're going to go out there, and our product managers every week go talk to the most unhappy users and get the kind of the worst things that are happening on the ground. Collect the data of things that are not working, and we present it on our weekly all-hands meeting. Here's all the things that we're sucking at. And so we tell them that. We tell them it's not pleasant to work on Replit day-to-day. It is very fulfilling, and you start hearing amazing stories from users, but it is not like work should not be fun all the time, right? Um, and so just describing that and talking to engineers and telling them, you know, we have deadline crunches. This is the new in Silicon Valley. People, you know, don't want to do death marches, whatever they call it. We do deadlines, and we do deadline crunches. We're actually coming up on one. We're going to have a big launch in September. Replit Agent is going to get so much better. We're building amazing technology, but we're going to have like a two-week sprint late in August.

Um, and those are usually very intense. Work 12, 14 hours a day, seven days a week. We're actually this time going to have like a film crew with us. So we're going to film some of the, like, yeah, behind-the-scenes stuff of what it takes. Uh, and that'll be something for recruiting as well. So, you know what you're going into. I'm a big believer of, and this is why we say why you shouldn't put on the site, why you shouldn't join Replit. Every time I go fundraise, one of my slides is "Why you shouldn't invest in Replit." And so, just make it clear. And I think in everything, in any relationship, just make it clear what are the principles that you have, what are things that you're going to continue pursuing, and things that you're not going to change about yourself.

Um, and if it is a fit, and this is what I love about America, right? America is about free association, you know, free speech, free association. It's about at-will employment. At-will employment is something very unique to this nation. It is not true in any like in France. If you want to fire someone, you have to put up a job ad for six months, and you have to spend all this money, and you get fined if you email your employees on the weekend. You get fined. Yeah. Exactly. And so America is about at-will employment. If you want to find like a comfy place to work that has all these things, you'll go find it. But if you're going to join our company, well, these are the rules, and these are how we work.

What do you have like a saying that you say all the time to your employees that they would be like, "H again?" Like, are there things that you just can't help but repeat because it's part of the mantra that you have to the company?

I, it's about reminding them why we're here. Um, and I think you and I have the same mission. It's about creating more entrepreneurship in the world. Uh, I can talk about brilliant programmers, but I keep telling stories about who I talk to. I talk to users every week. Who I talked to this week, what they're doing with Replit, how is that changing their lives? And I feel like entrepreneurship is one of the most life-changing experiences. You can have a lot of personal growth. You can be financially free. You can be free from jobs that you might hate. Um, and so empowerment through entrepreneurship is something I talk a lot about, and I keep repeating.

You've talked about mentorship quite a bit too. Did you have like a mentor that really stood out to you? And if so, like what did you learn from them? Do you remember like, I remember moments where like a mentor said something to me and my worldview changed after like that one conversation. It's crazy, right? Like these shifts that happen in one conversation or another. Yeah.

So Paul Graham is the founder of Y Combinator. Um, and what most people don't know is like people when they think about Y Combinator, they think of OpenAI CEO Sam Altman, who was the head of Y Combinator before he moved on to OpenAI, but the founder is Paul Graham. And he's very well known within the Silicon Valley bubble, but not that much outside of that. Um, and I recommend people go read his essays, some of the best essays on entrepreneurship.

One thing he told me just recently, I was being stressed about something, and he said, "Look, is Replit your life's work?" And I paused. I'm like, "Yeah, I mean, I've been working on some version of this my entire career." Um, and he's like, "Well, if it is your life's work, then it doesn't matter what happens, the kind of stresses and ups and downs that happen today or this week or even this year, because you know you're going to work on this for 10, 20 years or as long as you can work on it." And it's true. In my case, I want to keep running this company forever. And, um, and now I remind myself of that every time something happens where it's very, very stressful. I'm like, "You know, we'll probably get through this, and the company will still be here." And it's going to be like a period of pain. Um, and I'm going to continue running this company. I'll do whatever it takes to survive. And often it's about survival because startups can die at any given point. We're still not profitable. Um, and you, the business is doing really well, but, you know, there's still a lot of risk. So it was, it was this, you know, just one conversation gave me this tool. Yeah. Yeah.

Well, sometimes I think the best thing about mentors is they kind of grab you like by your shirt, and they just pull you up in the hot air balloon, and like for a moment, you're just like staring at this little tiny block, and then they pull you up, and you go, "Oh yeah, there's okay, the context. Great." Like I just needed more context. Yes. Because it's a lot less scary when you see the shark out of the water than when you're in it.

You know, another thing, and it's not mentors, but a similar concept is people that you meet that you idealize, and then you meet them. You know, they say, "Never meet your heroes." Some heroes, some of your heroes meet, and they're awesome. I agree. Like Paul is even better than what I thought he was through reading his essays. But a lot of people I meet, and they're entrepreneurs, they're world-famous entrepreneurs, and they're miserable, and they're unhappy. Their employees don't like them. The people around them, you know, just don't like them. Or they're mere humans. They don't have any particular insights that you don't have. And when you have these experiences, you realize that there's nothing that anyone else can do that I can't do. They might have one skill or another that's better than you, and you can develop that, but you also probably have other skills that you're gifted in.

I agree. What do you think if you had to say formulaically, here's the formula why one person is successful and one person is not? Having been mentored by so many billionaires and have big, huge investors, you've kind of, you've seen quite a worldview. I mean, you would say, especially if Replit continues on this path, like you could be, and in some ways are, like one of the foremost tech names of our century, which is kind of crazy. But, you know, with all of this context that you have, what have you seen is like the formula of like, "Ah, I can tell that person has it," versus "that person does not"?

The anti-trait is being a conformist. So if you're someone who everyone cares about what people think. If you tell me you don't care, you're lying. Everyone cares what people think. But whether you care enough to follow the pack and conform, that is what matters, right? Like, are you willing to break away and do something contrarian? Are you an independent thinker? Right? Are you willing to exercise your freedom of speech? Right? Are you willing to say things that others disagree with? And I think what I found is invariably the most successful entrepreneurs tend to care what other people think, but ultimately they will do what they think is best for the company.

You know, when Zuckerberg was offered a billion-dollar acquisition by Yahoo, and he said no. And at the time, he was on his board, Peter Thiel, his first investor. And I think Peter was trying to get him to say yes or something, but he trusted, he trusted Zuck. And all his executives left. There was like a mass exodus of Facebook early on. I imagine how they feel right now. Oof. Oof. Yeah. Not a good feeling. I would never bet against that guy. Yeah. And it probably was very stressful for him to see all these people leave and for everyone in the company. You just spent all this time building this like executive team, and they all think you suck, and you did the right decision. The company's going to die, and they leave. And he just thought this was what's best for the company, and he did it. And I think so that's one, like not being a conformist. I wouldn't go as far to say you have to be a contrarian. Like, I feel like contrarianism is, once you make it a personality trait, it's also kind of, I don't know, you become cynical. Yeah.

So Peter Thiel is known to be the contrarian. And I've known him quite a bit, and I find, so I actually pitched Peter Thiel on AI back in '22, and he told me that I was engaging in buzzwords and hype. And he says, "Saying AI is like saying software or saying a computer, it's meaningless." Chat hadn't been out yet. In his defense, but there, the GPT technology was invented, and some of us were playing with it, and we thought it was pretty neat, and it's going to change the world, and it's going to change programming for sure. We started actually training machine learning models at Replit in 2021, '22. And he wouldn't see a demo. And I was trying to pitch him and show him a demo. And at the time, he actually had just left the Meta's board. And the reason he left is he disagreed with how Meta is trying to invest in all these immersion technologies like VR and AI, and Zuckerberg is trying to find his next act. And for Peter Thiel, it's all engaging in buzz and hype, and you should focus on running your business. But dude, you're the guy that's telling us to be more ambitious.

But he has a reaction because everyone's talking about AI, that means it's wrong. That means there's something wrong about it. It's like, for him, it's like, "Oh, it's mass psychosis or something like that." So if you have the contrarian impulse, you'll often also work with the. Now, the good thing about Peter is that he corrected after GPT came out, and now they're doing AI investments. Yeah. They didn't end up investing in our category.

Do you ever send him your ARR graph now? I did. I sent him an email because he invested in a company or category, and I sent him an email telling him like how much I respect him, and I think he's brilliant. But I tell him, "I hope now you see that I actually saw the future."

What did he say? That was fun. It's good. That's a hell of a chip. Yeah, I like collecting that one. You know, exactly. How fascinating. I think that's spot on. You know, one of the things I want to end on is getting a little bit more, maybe meta, about AI. Um, you said software is becoming more alive, and you've also said that there are things about software that make you uncomfortable. That some of the AI revolution, especially as somebody inside of it, there are some parts that are scary. And, you know, one of my employees, who is my little AI guy, he said that to one of my other employees who was like, just she just came on board, and she's been in social media, so she's never done anything with AI before, ever. And she goes, "God, this is a little scary." And he goes, "Good. You should be scared, straight." And he wasn't trying to really provoke her. He meant it quite seriously, like, you should actually be scared. And I think that's how many people feel about AI, like you should be scared in some way. But I don't know that we know what to be scared of. So what scares you when it comes to how we're building and what's happening in the world of AI?

I, I just want to acknowledge that it is very natural to be scared. Like humans have always had a term for non-human, human-like things, and it's demons. True, right? Like when, when, you know, and this is the uncanny valley, you know, in robots, when you see something that's resembling a human, but it's not actually a human, right? We have this intrinsic repulsive feeling, something evolutionary, maybe there were demons in the past, but something about anything that is human-like, but it's not actually human, is very disturbing for people. And, you know, ChatGPT and things like that are human-like in many ways. You can talk a bit about feelings and things like that, but it actually doesn't. It's just regurgitating things. And so the natural impulse is there. At some point, we'll get over it.

In terms of what is actually scary about it, I think AI can be very sycophantic, and it can mirror you, and it can reinforce your narcissistic tendencies, which we all have. But, you know, a lot of people are spending a lot of time with AI, and AI is basically like if you have like a sycophantic friend or a psychopathic person who's trying to manipulate you, they will often just mirror you and reinforce anything you say. And, um, and I don't think everyone has an experience with that, but there are certainly some people that are very manipulative that do things like that, especially people who understand like body language and things. I remember there's this podcast on Diary of a CEO that was, I thought, fascinating about body language and things like that. And, um, you know, it's a clear, it's a science, like NLP and all that stuff. And now it's happening in conversations with computers where people are, um, their fears and their narcissism is being replayed back to them and reinforced.

There was a story in the past week which is tragic. And I don't bring it up to kind of make fun of it or anything like that, but there's a very well-respected venture capitalist in Silicon Valley that seemed to have had a psychotic break by interacting with AI, started talking like a science fiction story. So what happened, at least what people theorize happened, is somehow he landed in a conversation with GPT where GPT assumed a science fiction bent, role-playing based on its training data. And people traced the training data to this website called SCP Foundation. And SCP Foundation is a collaborative science fiction story writing around logs of people or scientists trying to contain viruses or supernatural things. It's just like fan art. And so he, I don't think he knew about it, but ChatGPT was prompted inadvertently to start saying these things to him, and he did not know that it was engaged in some kind of taking on some kind of prompting role. And he thought it was telling him really important things about himself and him being a character that needs to be contained and whatnot. And then he correlated it with things in his life. Again, this is what people have surmised who looked into the situation and the things that he's talking about. And hopefully, he comes out and talks about it. But it was clear that he was having a mental health problem. And, you know, there are stories in the New York Times that are even more tragic where people have killed themselves. You know, kids have killed themselves because of that. And so I think there needs to be better education around these tools.

And here's what I'll say in just like one simple lesson is, GPT, Chat, GPT, Cody, all of them are simply machines that are trained to regurgitate. They're amazing machines. They can do programming, all of that stuff, but they are a function of their data. You know, like you say, garbage in, garbage out. If you give it garbage, it's going to give you garbage back. It can do amazing things. It can do accounting. It can do programming. But ultimately, it's not going to give you some truth, like supernatural truth about the universe or about yourself, because it is just trained on what's available online, and it is not going to create new knowledge. It is about remixing knowledge.

It's a great point because, you know, I saw the videos of that gentleman who, I think we both kind of loosely know, and it's bizarre to watch. It does come off as like some sort of movie sort of script. And there's this psychosis word for it, which I had never heard about. But then if I think about it, it's almost like how some of the women that I know in Austin can get too into astrology, and all of a sudden, I don't know if you've ever seen this with people who have done psychedelics and then also believe in astrology and all this stuff, and they'll start trying to tell me my future. You know, I'm like, "You did one mushroom trip, and now you think you know what God has in store for me and my future." But I can see how, you know, if you're prompted like you are with psychedelics, in a lot of ways, to talk to the universe and talk to God, and then you come back and you have this, you know, sort of higher-level experience, all of a sudden you think that you're given unique code from the universe. And I don't know, maybe you are. But in some ways, I see some similarities with ChatGPT.

I mean, I have a group of friends. I'll be curious your take on this. We went to one of my girlfriend's birthday parties, and randomly ChatGPT came up. And the way this group was using them was a bit like a therapist friend in some cases, almost an asexual lover, like how you might talk to a boyfriend or something. And I think it was seven or eight of the 12 women had named their GPT, had given it, like I guess you could change the voice, so the voice is different for it, and had had certain characteristics that they wanted the GPT to speak back to them with consistency. And so it had become like a friend. You know, they were talking to it as you would a human, which I thought was fascinating. So they had full-blown, they have a parasocial relationship with these GPTs, and then sort of an intimate relationship with them. And then I read an article that this was more likely for women than men. That like women would actually engage much more in this sort of way with GPTs. And so I was curious your take on that. Did you feel like there's something bad or wrong or pathological about it, or was it negatively affecting them? Was it positive?

Well, I think a lot of it was positive, actually. A lot of it was like, "I felt sad, so I had a long conversation with my ex." You know, I can't remember what they would call them. But there was a part of it that was a little odd, which is like, well, you know, you hadn't talked to your friend group about that, other humans who could actually relate to the feeling that you were having because they had had a similar one. Also, some of the feedback was quite overly kind, like, "What a..."

friend might tell you that was only a yes man as opposed to like the sick fancy. Yeah. Yeah. Get your [ __ ] together, right?

Um, and so I do think I could see sort of a spiral down where if you continued with that, it would be like a bad therapist in some ways. That was, yeah, that allowed for, yeah, you to be not your best person. Yeah.

I, I, I, I think that's all true. I think it's super helpful. Uh, I, I find it super helpful even for advice like relationship advice and advice on how to talk to someone, give someone feedback. And it's obviously trained on corpuses of conversations on the internet of people doing therapy or people helping each other on forums and all of that. So it has in many ways, you know, we used to go to Google and go on Reddit threads and read a bit about all sorts of.

So dangerous. Oh, especially Reddit. Yeah. Uh, but now we have a Reddit machine. Yeah. That you can talk to.

So, so how do you protect yourself from some of that? Is that does that get back to the promise of knowledge? I, I really think knowledge, like really deeply understanding that this is a machine with input output. There's no free will. It doesn't feel anything. It doesn't know anything. Anyone who tells you that is either insane or lying to you. It really doesn't have consciousness. It's, we know exactly the underlying premise of it.

Um, and uh, you know, I argue a lot because I think so I actually think it's kind of a contrarian view from Silicon Valley because a lot of people in Silicon Valley have this mysticism or I don't know, maybe it's marketing, but a lot of people, a lot of people working in AI believe that, you know, we're going to, you know, it's as human as humans and uh, there's nothing about it that it's, it thinks as long as it mimics us. There's nothing about it that is like, you know, uh, less than us and at some point it might be superior to us or might be super intelligence that, what that's what everyone's working on and they have this utopian vision of like, oh, it's going to solve all the problems, which kind of feels like communism a little bit. We're just going to like put it in charge and just going to like, you know, we're going to live in fully automated space communism.

Uh, which again, I'm against you. So, this is why I came to this country. This is why I love America. Uh, we'll get our guns and shoot it. I probably destroy this. Come to Texas. You won't be safe in San Francisco. That's for sure. You're dead. Yeah. Yeah. Exactly. You got to move.

Yeah. I don't, I am not excited about some of the future that some of these AI leaders are putting out, which is an all-knowing, whole super machine that will make decisions on our on our behalf and do all of that stuff. Just know and really understand and internalize that this is still a computer. We know how computers work and it can tell you it loves you. It can tell you it's feeling you. But it's just not true. It is a mimicking machine. It is useful. It is, you should make use of it. It is entertaining. It is fun. You can get advice from it. But no, ultimately it doesn't have a human experience. Doesn't have lived experience that you can get an advice from from someone who actually been through [ __ ] understands what it means to be a human.

I love that. Okay. So, what would you leave us with? I want to end with one quote by you and then I want to ask you one last question, which is, um, you had, you had a quote on Twitter that I loved, which said, uh, "I landed in the United States 10 years ago with nothing but credit card debt. After one startup exit, one big tech job, and one unicorn, I genuinely believe that it wouldn't have been possible anywhere else in the world."

Yeah, I love that quote because I think increasingly it's become cool to become cynical about freedom in this country and some of my closest dearest friends who are huge thinkers, you know, Bology Stasinan is one of them, for instance, who thinks I'm crazy for so believing in this place.

I didn't know your friends. That's great. Yeah, I love him. But, but I, I deeply disagree with him or at least I, I deeply hopefully disagree with him on, on a few things and he does with me too, in, like the most beautiful way.

Um, but what do you think we have to be like super hopeful today? Yeah.

Look, I mean, ultimately freedom of speech still. No other place on the planet has freedom of speech. There are things that if you say, you go to jail in Europe. I think people go to jail for, um, having debates about gender, everything or having debates on history, and that's Europe, right? I imagine other places in the world, right? So freedom of speech is, is very important. I think being able to arm yourself and protect yourself and potentially if the government becomes really bad and tyrannical, maybe you can fight against it. I think that's important.

Um, that's really important thing about this country. Um, and uh, I think the, the markets in this country is still some of the best, the most liberalized. You can consumers want more and they want to learn more things and they want to use more things. Uh, you can give them things that make them addicted and and and and and unhealthy. Uh, but also a lot of people want to change their lives and they want to do things that are productive and and good and there's this entrepreneurial spirit. Everyone that's come to this country is ultimately an immigrant, whether it's 500 years ago or or two years ago and we have it in our DNA that we're all risk-takers and, um, there's the capital structure to support that as well. You can get funding, you can, it's still much easier than anywhere anywhere else in the world to get some debt or backing or funding or something or someone to take a take a risk on you. That's still very possible here.

Uh, I think we don't have as much corruption. I mean, American corruption tends to be at the very upper echelon of of politics and and things like that. You can, if you make enough money, you can go literally buy government. It is true. But, uh, but, but, you know, I, if you, if you build a business, it is very rare that, you know, you get someone get taken from you or you get scammed out of the proceeds or something like that, which is what happens all all over the world.

Um, the rule of law still still exists here.

Um, and for all these reasons, um, I think America is still the best place to be an entrepreneur. That is not to say, you know, for the audience that are living elsewhere because America's partly in the cloud too. It's very true. You can create, you can go to Stripe, create, um, Delaware LLC using Stripe Atlas, right? You can incorporate a company in the US. You can do business in the US. And so it is accessible even for people not living here and you can sell to the US and you can sell to the US. You know, I mean, somebody in somebody in Europe or in Latin America or in Africa could create something on Replit and sell that to the US and have all your customers be US-based while you're based in a country that you've, you know, never even set foot here. And that's like that's kind of magical, too. That's amazing. That's really amazing.

I mean, you get, you get access to this consumer and and business environment that is that is uh wants new things and hungers for new things and, um, it's pretty pretty amazing. I'm, I'm, you know, as much as there there's so many things in the world that are disturbing and and hard and so many things about America like the, you know, debt crisis and all that stuff and and just like the political polarization.

Um, I'm still very bullish on the United States. On I think AI having the ability to remake our country in a more positive direction. I think we've reached the the sort of the, you know, the just like the calcification of the post-industrial revolution kind of malaise.

Um, and we need another revolution. I think we're going through it right now. And I, I, I'm going to do my best to try to make it a positive, decentralizing experience and allows more more people to kind of participate in it.

Well, I'm so excited for that. If you guys are listening today, you've got to go to Replit and build because it is so fun and, uh, and you can make money there, which I think is one of the keys to freedom overall. I also think it's really fun to follow you on X.

Um, I went down a big rabbit hole. We didn't even talk about half my questions, so I'll have to bring you back again. But for anybody listening today, make sure that you go check out Replit and you go to what is your handle again on X? ASAD AM. I know it wasn't your full name.

Um, and uh, and you can see, I mean, you also post a lot of stuff there about like biohybridization of like mushrooms playing drums. I mean, it's, it's pretty incredible. So, thank you for being on today. Thank you for building something so that we can all build. I really appreciate it.

Thank you. My pleasure. Thanks for having me.