Moonshots with Peter Diamandis - Replit CEO on Vibe Coding and the Future of Software Development w/ Amjad Masad, Dave Blundin & Salim Ismail | EP #196

Episode Date: September 23, 2025

This episode was recorded at https://www.imaginationinaction.co/  Get access to metatrends 10+ years before anyone else - https://qr.diamandis.com/metatrends    Amjad Masad is the Co-Founder & CEO of Replit Salim Ismail is the founder of OpenExO Dave Blundin is the founder & GP of Link Ventures  – Connect with Amjad: https://replit.com/  X: https://x.com/amasad  Connect with Peter: X Instagram Connect with Dave: X LinkedIn Connect with Salim: X Listen to MOONSHOTS: Apple YouTube – *Recorded on Sep 9th, 2025 *The views expressed by me and all guests are personal opinions and do not constitute Financial, Medical, or Legal advice. Learn more about your ad choices. Visit megaphone.fm/adchoices

Throw away most of the advice that your parents and your society don't give you. Because it's such a dynamic world right now. And conforming is the worst thing that you could do. I'm John is founder and CEO of Replit, which was a YC 2018 company. I am not a programmer. I am not a coder, but I can now create software. They came to Silicon Valley to start the company, and that was Replit. We knew for a fact that there are a lot of people, a lot of ideas, and there's so many barriers in their way. The barrier of programming and the pair of how hard it is to write software.

very quickly going to get through all of the code out the legacy garbage part of society. The idea is to be able to talk ideas into creation. Programs are no longer static. You can buy a machine with zero software in it, and you can vibe code all of it by just like talking to the machine. The career of the future is entrepreneur. That's the only career that I think is going to survive. If you have access to computers and that or not, you should be able to build something great. And we're going to find out all over the world.

Those who are having the most impact will rise to the top. What a profound opportunity for everybody in the world to make a massive difference. Now that's the moonshot, ladies and gentlemen. Welcome, everybody. Allow me one second to just bring onto the stage my two moonshot mates, Dave Blundon. Dave is the trader CEO of Link Ventures, Link Exponential Ventures, ventures runs a little over a billion dollars out of MIT. No comment to Stanford here on investing in companies in AI space. An amazing guy built one of the very first, if not the very

first neural net machine learning company exited for a billion dollars and it's been uphill since then. He and I and Mike Saylor were roommates together at our fraternity. So it was quite a corner of the house. Selim Ismail over here, another moonshot made, is the creator of exponential organizations. He was my, effectively my co-founder along with Ray Kurzweil at Singularity University and has been the author of a number of extraordinary books and we'll learn more about him. I'm Jod. A pleasure. No one needs introduction of you. The CEO founder of Replit, a pleasure to have you here. Nice to meet you. It's such a pleasure for me. If I were to summarize your life, I would go, coding an internet cafe in Jordan, rejected three times by YC, finally gets in, build, Replett, and is now helping a billion people code.

Is that a good, that's a good summer? Not bad. One of the things that Dave, Salim, and I talk about a lot, because it's what we care about deeply, is helping entrepreneurs take moonshots and having a moonshot that's part of a massive transformative purpose. And we talk about the idea that an individual today can possibly impact the lives of a billion people. And what an extraordinary world to be in. And I love the fact that your moonshot, and it sounds like your massive transformative purpose, has been to enable a billion people to code.

when did you first have that vision you know I built my first so I used to go code in internet cafes back in Jordan and one thing I realized that they you know they had computers all the place but they were not using software to manage the system so everything was paper and pen and I decided to build software to manage the store and that was my first company and the thing that I noticed is that the idea was so powerful and I felt like I could build it very quickly and it took me two or three years to get the software done

and it took me another months up to a year to figure out how to actually deploy the thing how to like burn it into CDs, how to get it to the cafe, how to install it there, how to manage the thing and I felt there's so much, there's such a big barrier between having an entrepreneurial idea

and actually deploying it into the world that's what I started thinking about okay how do you make that process easier Can I show you easy? Can we get the slide up on the screen here? So this is me this morning. I am flying from Santa Monica Airport. I've got the laptop open in my lap, obviously.

You can see a starling contendentna on the dashboard over there. And I'm coding on Reppelin on the way up here. This is the future. That's awesome. I tweet this out. hopefully, you know. Did their traffic control know about this? That's an odd one.

Hopefully, Elon reaches out and talks about the Starlink Replit partnership. I think that would be perfect. But I said, life doesn't get any better than this. The ability to create on-demand anywhere at any time. Yeah, yeah. And so in college, I ran into another problem, which is every time I went to go do a little bit of programming in a different language, I would have to install the development environment again and again and again.

And I was like, you know, we're moving everything to the web, everything to the Internet, except programming. And so I started working on what would become the first browser-based coding sandbox. And that went viral on Hacker News and GitHub. And I remember one of the sort of very important moments was Brendan Ike, the inventor of JavaScript, part of the kind of Gizola, tweeting about my project. So here we are. I'm a 20-year-old kid in Jordan in my parents' basement, and the inventor of JavaScript, the language that I use,

is talking about my application. And that was huge for me, and then I started getting these job offers. Peter Norvig actually was just started to create Udacity. And he reached out to me and it's like, hey, we're trying to integrate this program you built so we can teach people programming in the cloud. And then another company called Code Academy started up on the technology that I built, and Code Academy became the number one coding school in the world.

How many people do you learn how to code on Code Academy? A lot. I thought it would be more, but we taught 15 million people how to code. And then afterwards, I sort of left the company and set back and, and, and, You know, just reflecting on my own journey, you know, if a kid from Jordan is able to kind of come to the U.S. and invent something that affects millions of lives, there's probably millions of other kids like that all around the world. There are. And so that became my mission as, like, how do you make it so that anyone can build something that could impact millions of people?

Love it. Dave. Sir. Where is yours? So, really, really interesting. The journey from Jordan to Silicon Valley, I think, is worth spending one second on because I have a follow-up question on that. But you skipped over that in your storyline there. How did you end up actually physically here?

Because I know you live here now. Yeah. So I got an 01 visa to go to New York, and I was a founding engineer at Code Academy, and then left Code Academy. And then I couldn't work. At the time, I left, I had 60 days to stay in the country. By the way, can we talk about how important the 01 visa is? And how critical, you know, one of the things I've been tweeting and saying we should be like stapling a green card on top of every degree that comes out of Stanford, MIT, or Harvard to enable those entrepreneurs to stay here and work here and contribute to the U.S. economy.

And I think this part of the story is really important with you in particular because you didn't come over to go to Stanford or to go to one of the universities here. You already had finished universities in Jordan. And so it's a different story from us. Go ahead. Yeah, a lot of my colleagues, a lot of other people from my country, typically their way into the US is via university. But I actually wasn't like a grade student because I spent all my time hacking and programming so my grades were kind of crappy.

So I had to, like I came through the open source route. I heard the story of you hacking your university. Yes. To change your grades. Yeah, so I felt like it was unfair that. Better test anyway, isn't it? I mean, that should be the actual grade. Yeah, like, so I was actually getting failed for attendance

because I wasn't showing up. Because you were doing work. Because I was doing work. Actually, one thing I tried to do, I had like a Nokia symbion phone, and I was like, can I code on their desk so I can do something more interesting than listening to this teacher, blah, I love the fact that you hacked the Y Combinator application and put a Rickrow video into the application.

That's true. degeneracy. Right, yeah, that's right. I mean, there's this, I think, this feeling of, you know, these structures, these systems that we built around civilization that is really meant to make people conform.

Yes. Put you in the box. Yeah, put you in the box. And I always felt like the first time it worked where I thought differently and did something different. And that allowed me to go through the back door, which I think is a more meritocratic way of doing it. I just embraced that. And, and

But partly, like, it kind of always works out. I don't know if I'm just lucky. But even quitting Code Academy and didn't know what to do, I just, like, got bored and wanted to do something else. I had only 60 days to stay in the U.S. And I went and applied to a bunch of companies. One thing, one company I was really excited about at the time, Mark Zuckerberg was talking about the vision of internet.org, right?

The idea that, you know, we're going to connect to everyone in the world. Everyone will have access to their internet. That's an extension of my mission. of if you have access to computers and that or not, you should be able to build something great. And we're going to find towns all over the world. So I applied to work at Facebook, and the lawyers told me, well, you have to go back to Jordan so we can apply for a new visa.

I was like, screw you. I'm not going to do that. And on my birthday, I got a green card in the mail because I had applied for one. Something worked. Something worked. So the idea of just taking a chance and kind of the universe rewards you for that. Don't be too stupid about it. But I think that is kind of a scary story, though,

because that green card could have gone any direction. That's totally random. Yeah, and here you are, and here's Replit. I mean, think about what a difference that makes for the world. And when I joined Facebook, I was sort of like a nobody, but I was really excited about this idea of Internet.org. I started working on Android because I thought that Android is the main device that most people were going to be connected to the Internet via.

And I ran into this problem with Java, Java being this really, you know, crap, no offense to any. Really crappy programming language, resource hog, really hard to program. And then I happened on a new idea, which is, okay, can we make the mobile platform more programmable? Start working on that, and it found other people of the company working on it. That became React Native. And React Native made it to that any, you can write code once, and you can run it in all the different devices. And now that also touched billions of people.

people's because there's you know part of Facebook's written in it discord is written in it and so I think they're just like this idea of not overplanning your life just going about life and finding interesting problems to solve and solving them usually will net have the right thing well so that experience that life experience is a perfect segue into my question which is you know using replet to discover talent all over the world so there's two things I'm dying to ask you one of them is the there are two meta questions I can use my own product to discover talent and to recruit, and I can use my own product to build my own

products. We'll get to the second one later. But it's really unique to products like yours that you can actually see a billion people or whatever coding. And, you know, you being discoverable in Jordan through that vehicle, you can now make that a much more scaled version because there's talent all over the world. And it's, you know, it's not going to naturally, you know, have a grade point average, a degree, or whatever. And so how are you going to find those incredible talented people. Oh, wait, they're right here using my product every day. I can see them writing incredible code, building incredible products. Our first employee, we were, it was just my co-founder and I, my co-founder is my wife, and we were kind of struggling to hire. At the time, we weren't

in YC yet, and YC rejected us three or four times. That's why they had to get Rickrolled. Wait, was it, was it three or four? You're not going to forget that. I actually kind of forgot it because I would apply every time. Oh, I see. And then I would record a video every time. You know that YC video is like a hostage kind of situation. You're sitting there. How many of you have applied to Y Combinator? Three or four times. Every week, my team and I study the top 10 technology metatrends that will transform industries over the decade ahead. I cover trends ranging from human robotics, AGI, and quantum computing to transport, energy, longevity, and more. There's no fluff. Only the most important

stuff that matters that impacts our lives, our companies, and our careers. If you want me to share these Metatrends with you. I writing a newsletter twice a week, sending it out as a short two-minute read via email. And if you want to discover the most important Metatrends 10 years before anyone else, this reports for you. Readers include founders and CEOs from the world's most disruptive companies and entrepreneurs building the world's most disruptive tech. It's not for you if you don't want to be informed about what's coming, why it matters, and how you can benefit from it. To subscribe for free, go to Demandis.com slash Metatrends. To gain. access to the trends 10 years before anyone else. All right, now back to this episode.

I love that idea that you have a discovery engine for talent, right? And I think one of the things that's very important today is that the talent is global, genius is global, and the tools are available and demonetized globally. That's right. And how fast it's changed. I remember I was talking to Dave earlier that when I was, when we were both undergrad at MIT, there was a course called 611. 611. And I would literally build my computers with NANDGates, NANGates, and NARGates, and would hexadecule

code them, you know, finite state machines. And that was coding. And that was hard. As how far it's come. You only need a NAND gate, by the way. But, you know, we eventually got to E-Proms. But it was

extraordinary how far it has come. I want to talk about the future. of coding one second. I had Imad Mustak on my abundance stage two years ago. We're talking about this, and he was saying how coding would be done by AI systems. And the front page of all the newspapers in India is Imad Mustak announces coding is dead. And you got a bunch of hate mail for that. What is coding in the future?

Yeah. So I think to understand coding in the future, we have to understand it in the past. So in the early era of computing, maybe in the 30s and 40s and a little bit into the 50s, computers were fancy calculators. In order to change the programming of computers, you had to literally rewire them. So there was no programs.

Turing, in his 1936 paper, invented the Turing machine and showed that you can build a universal computer. But it wasn't until Von Neumann in 1945 that he invented the concept of a stored program. program machine, which is what you have in your pocket today. Every computer is a Von Neumann architecture. And that was a huge leap forward. The idea that you can program a computer was suddenly a big thing because computers weren't

programmable. People forget that. And then you had another leap, which is when Grace Hopper invented the compiler. One interesting quote that is actually very much reminiscent of this era, and I'm sure she got a lot of hate mail for that. But she said, you know, we used to kind of program in machine code. It's not really programming. It's kind of similar to, like, this rewiring of the machine.

And she wanted people to program in English. And it's sort of the C program language and the high-level program languages that she invented. She called them English. And it's sort of similar to what, you know, Andrew Carpathie recently said. It's like English is a hot new program language. It's not hot or new. Grace Hopper has actually thought about it. And she was like, the specialists are not going away,

but we're going to introduce programming to millions of people. Yes. Because programming is going to be in English. Of course, you know, that vision sort of went away, and programming became an industry, and you had to, you know, go to Stanford for four years to be able to do it. But now we're at a moment, I think, as big as the invention of the compiler, perhaps even bigger and as big as the invention of the stored program computer,

which is programs are no longer static. They're malleable. You can buy a machine with zero software in it, and you can vibe code all of it by just like talking to the machine, right? And that's been the vision of computing since the start, which is why I think vibe coding sort of undersells that vision. The idea is to be able to talk ideas into creation.

So your idea generator that gets compiled in code. That's right. That's right. What ultimately matters is thinking clearly and being able to break problems down into individual components

and then being able to communicate it clearly. So this is a question I have. I grew up doing C++ and Pascal coding, etc., right? And you learn a certain structure, navigating data structures, thinking through architecting the overall software before you start building

it so you're more efficient and you go around optimizing, et cetera, et cetera. As people are vibe coding, those skills aren't there. It's the same thing like a student that's using chaty PD to write an essay. The thought process of constructing a logical flow of arguments is

lost. So do you worry about that or is it the same argument that we used to use with the assembly line yeah, who the hell wants to be writing assembly and machine learning or hexadecimal? We'll just do it at the higher level and you're then articulating those ideas at a higher and higher level. Is that essentially the vector you see

going down? I think so. And every time someone would make an argument and any history of computing where you say, okay, this is too high level, you're abstracting the, you know, the NAND gates and the registers and all of that, you're going to, you're not going to be able to do good work because it's too high level. Every time it's been proven wrong, and every time what we thought of it is too high level becomes low level. Now, C programming is low level.

For them at the time, you know, the machine code people was like, what is, you know, this is bullshit. Yeah, this is bullshit, right? Interesting twist on that, though, if you said, I didn't know a great offer. you that you weren't around when I was doing software development. That's all I'm going to say. So I didn't know Grace Hopper invented the compiler, but if her original vision was you're going to do it in English.

But then you look at where Python ended up, or you look at where high-level languages ended up, you can't really easily specify in English, you know, persistence or pointers or in-direction and stuff like that. So it kind of went to some middle ground. So now with Lovable, you can build stuff like, boom. And with Replit, you can build stuff like boom. And I'm sure the thing you built on your plane,

worked, right? It did. It was a, I built an app that would assess my mindset every morning and then would upload it, I would upgrade it based upon sort of give me a series of prompts. I have a book finishing called Mindset Mastery, and I wanted that to become something that it would just be usable for myself. And it was great. It was beautiful, actually. In addition to being great, it was beautiful. I want to hit something that I think is important. So right now, there's 150 million GitHub accounts. I looked at the numbers, and if you look at programmer's salary

over the last three years, it's up 50% on average. And at the same time that we've seen 50, I'm sorry, the number of programmers, I'm sorry, has gone up 50% over the last three years. At the same time, the number of programmers have gone up 50%.

The average increase in programming salary has gone up 24%, right, which is, If there's an over, you think it's an oversupply, which tells me that the value of programmers is massive and it's, you know, the exponent is greater than one on this. What happens when we have a billion programmers, two billion programmers? What's your vision for that world of abundance, right? You know, you read abundance early on. Thank you for your kind words. You said earlier, it is one of the mechanisms for unleashing creativity and increasing unleashing entrepreneurship. And we talk about all the time that the career of the future is entrepreneur.

That's the only career that I think is going to survive. I don't know if you agree with that or not. So what is it like what we have a billion, 2 billion programmers, coders in the world? Yeah, I think that the world will trend into a more meritocratic society.

And not because there's someone who has an ethical vision of the future and they're going to impose meritocracy. I think meritocracy will impose itself because those tools becoming more available to more people those who are having the most impact will rise to the top

naturally, right? And I think in every aspect of society I mean we talked about our YC story when we entered YC it was ultimately because Sam Elthman and Paul Graham saw rapplet on Hacker News. So just to tell the story a little bit, which I've heard

you know you get rejected three times Paul Graham and Sam basically reach out to you. They've seen you on Hacker News, and they're like, okay, you've got to apply again. Yeah. So what happened was I wake up one morning in late 2017. We had started the company in 2016 and spent a project for a lot longer than that. Overnight success after nine years of hard work.

And we've been grinding. It was like two or three of us. we were like selling to schools and whoever can buy our software. Mom, would you buy this? Anyone. Really kind of door to the door, you know, programming environment salesman. And so

I get this Twitter DM and it's Sam Altman with his lowercase and, hey, I run YC. Like, dude, I know who you are. We're interested in what you're doing. Would love to meet. And then he gives me this address. I'm like, that's not YC's

address in Manview. And so I show up there, and it is NeuroLink and Open AI. I didn't know about it either. Maybe I've heard about Open AI a little bit. So I go into Open AI, and Sam was sitting there. And he says, we're really interested in what you're doing. And by the way, Paul found you in Hacker News. Because at the time, I was building the system and blogging about it,

wrote some interesting blog post, and he says, well, you should go see Paul Graham. Unfortunately, he's retired. He's not in the Bay Area. You should go to London and meet him. I was like, okay, let me grab my private jet and go. The thing about rich people in Silicon Valley once they're rich, they think everyone else is rich like that. I couldn't even get a visa to go there, you know.

So I'm like, how about you give me his email first, and we'll go from there. So I start this email relationship with Paul. Over two months, we're writing back and forth. Paul's a great writer. He's a great writer. At some point, I'll get his permission. to publish the writings that we had. He would write me essays about our shared vision

for how programming should be. It's like the reason we started YC is because we knew for a fact that there are a lot of people, a lot of ideas, and there's so many barriers in their way, not only the barrier of capital and the barrier of getting attention, marketing, and all of that, but even before that, the barrier of programming and the pair of how hard it is to write software.

And so we go through all of that, that. And at the end, he's like, yeah, I think YC is starting in next week. I'm going on a trip, but you should talk to Sam about going into YC. I was like, okay. And so I email Sam, and he's like, yeah, hey, I would love to have you in, you know, you should join. And we have the kickoff tomorrow. And I say, okay, and then he shoots me another email. He's like, formality, you should still apply. It's like, oh, fuck. I'm not going to go through the YC application all over it again. And how long is the YC application? I mean, if you want to do a good job, it'll take you an hour. I'm very lazy.

I was like, it's a, you know, so. You couldn't have a booklet writer for you, huh? At the time, I couldn't. There was no alum's, and so I had to do it manual. But I did a very large job at it because I just didn't, like, hey, guys, you're recruiting us. I'm just not going to put in the effort. And so the video came, I'm like, fuck, I'm not going to record the video. So I paste in a link.

And then I, we go to the next. The next day is the kickoff. They do the late interviews. So they always have this tradition of letting a few start up in at the same day. So we said the whole day we're waiting. And then finally the door opened and they told us he can come in. And so we go in and I, it was like Gustav, Adora, and then Michael, CEO of Ycombinator at the time, and he was a big guy, right?

And I'm shaking their hands and I felt Michael kind of squeezed my hand a little bit. I don't know what was going on there. So the moment I sit down on the chair, his face is red. He's like, why did you Rickroll us? It turns out, as we were sitting outside, they were getting Rickrolled inside before the waiting bus. So they clicked on the video and they got this. The Rickroll video plays. If you guys haven't seen it, just go Google it.

Yeah, there's a meme. You can find it easily. And he was very upset, and he started really kind of giving us a really hard interview. Really good question. And what are you thinking at this point? I'm thinking we were screwed, like, we're done. And so I actually exit the room and ordered an Uber. And my co-founder was like, we have to wait.

It was like, ah, we're not again. We're not getting again. After we did that, we're not getting in. But then they get a phone call from Maduro Chang, and she's like, you got in. And so we go in that same day, and Sam Altman did the kickoff and all of that. How important was the YC experience to you? I mean, would you have accelerated and gotten to this point without it? I think one property that I have is I would never quit, and so I would have kept going, and I think we would have probably made it, but it probably accelerated our progress by probably years.

I mean, it opened a lot of doors, and just being around these amazing people, being around Sam and Paul, and you just raise your ambition a lot, being around all the ambitious people at Y-C. It's a Silicon Valley thing in general. I think that's what matters. The question about talent is very important. There's talent everywhere. I think the density of that work, at places like Stanford, it's like Silicon Valley.

Paul Graham, out that great quote, that being a startup founder is only about one millimeter away from being unemployed. So when you're telling your parents, you need like-minded people all around you as your support network. Otherwise, it's just really, I mean, doing it in Jordan would have been really emotionally difficult, and you obviously pulled it through. Just to put a pushpin in that story, too, an incredible story. But if anyone from Stanford administration is in the room,

I know Eric Brinjolfson and others are, like your application, process is in complete fail, but the pull process of discovery is a complete win. So now, when you're building your own company, you see all these people and what they're doing. You don't need them to apply to come and work for you. You have the data. You can analyze the data and pull. The world is clearly going to move from applications to big database to pull.

To discovery. To discovery. Yeah, that's what I'm saying. The world is going to be forced to be more meritocratic. Yes. And I think the problem with institution. Like YC, they've gotten better, and Stanford, they're, like, very much about pedigree.

Yeah. You know, where are you from? And it's, you know, I'm not faulting that far. It's a heuristic. Yeah. If your Stanford dropout, you know, Stanford did the work for us to choose you. And therefore, we're going to piggyback on that. But in reality, again, talent comes from everywhere.

Philip Rosdale, a creator of second life. I don't know if you know him. Of course. One day, Philip says, do you know why there are more startups in San Francisco than any place else? and why the success story is so high. And he goes, I want to tell you what I found out. So he ran a script on top of LinkedIn

where he measured the density of technical founders per square kilometer. And he found that San Francisco had the highest technical density. There was Austin, there was Cambridge, there was, you know, Miami starts to pop out. But the notion is if you're taking the risk of starting a company and you've fail instead of going back to your mom or dad's, you know, bakery to work, your friend down

the street has another startup you can go enjoy it. Yeah. And so it is a density issue. Yeah. But I think, you know, increasingly that is moving online. I know you're good friends with Bologi and his idea of network state and all of that. And, you know, I just, I was in Jordan last week and I saw a company there that is, you know, ostensibly headquartered in San Francisco.

They run all their engineering there, but they come here quite often. They meet with VCs. They're, you know, they incorporate in Delaware, and when they go into meetings, it's almost like their headquarter in San Francisco. And so it's like, you know, San Francisco is going to the cloud in many ways.

Yeah. Salim. So, you know, this evolution of where things are going with the future of software development is pretty profound, but at the business level, I use the analogy of Kickstarter, right? When Kickstarter appeared for the first time in business history, you could get market validation for a product without building the product, right?

That was kind of incredible. What you're doing is essentially providing an environment where people can write software and fully finish that software and then get market validation and figure out what's needed at almost zero cost. That's right. This is taking out the marginal cost of all of this stuff, which is what we talk about a lot with our, in abundance, with our exponential organizations, link ventures, et cetera.

as you accelerate that, essentially you get to a point where society is just one huge amount of code being written on a non-stop basis. How do you think about the future of society in the context of that? Yeah. So it's a great question. Let's zoom out here.

Yeah, yeah. So I think if you think about entrepreneurship, I don't just think about it in terms of like quitting your job and starting a business. It's about being able to solve problems in the world and taking initiative without your boss selling you to do that.

So we're seeing it a lot in the enterprise as well. Like we've seen cases where, you know, a product manager at Zillow being able to deliver millions of dollars in bottom-allum revenue for the company because although he didn't, he wasn't an engineer, he was able to increase conversion of, you know, homebuyers at the company. And then suddenly everyone heard about that story.

now we have 500 licenses at Zillow, and everyone is recommended to VibeCode because they want more entrepreneurial people. That person got promoted. Agency. Agency. It's high agency. And it's the biggest challenge that large corporations have today is they're so structured and they're so fragile in that structure. I mean, Salim, you talk about this all the time, right?

It's enabling your employees. You try anything disruptive in a big company and the immune system attacks you, right? The standard problem. But here's why VibeCode. is important is because you can do it anyways and you can show the results and the results will speak for themselves, right?

This is a huge breakthrough. Permissionless innovation can't happen inside the organization as well. We call this PDI, permissionless disruptive innovation because in the past, whenever you want to do with something really disrupted, you had to get the permission from the Medici family or from the church or from the corporation

or from the government or from an investor. Somebody had to bless you to do this, right? Today, for the first time in human history, you can do very disruptive things without any permission. My favorite poster child here is a Vitalik Butern, an 18-year-old kid, gets together with a few friends. They ignore their professors. And now, boom, you have Ethereum, a $600 billion ecosystem that nobody understands.

And so that's kind of incredible that you can do that. And what a profound opportunity for everybody in the world to make a massive difference. And what you're doing is now providing the scripting language to take anybody, is a massive purpose, as we call it, and sort of making it an actual reality. So here's where it gets even weirder, stranger, more abundant, is that it is not going to be

just you doing the disruption. It's going to be your agents. You can be able to program agents that are running in the organization, finding the inefficiencies, and actually making progress on your behalf as well. You can also run a startup that is producing these agents

that, let's say you're someone who's worked in HR for 30 years. You know all the HR processes. You have a gold mine in your head. You have an unmonetized gold mine in your head. And what you do in the near future, we're actually launching this tomorrow, Agent 3, heard it here first, but Agent 3 will allow you to create other agents. So they'll be able to create like an HR agent.

How long you've been working on that? It's been, you know, about six months since Agent 2. So I think every six months, you know, it commensurate with the progress. and LMs. We're trying to predict what is the next set of capabilities that would unlock the next version. So we already know what Agent 4 is going to look like. We take a bat, we try to build it, and over time, you know, the first version of it would be crappy, but then the foundation models and everyone sort of catches up. Yeah. So I always tell entrepreneurs build crappy products. So agents can basically spin off their own agents. What could possibly

go wrong? Look, I don't Look, I don't focus on what goes The other people's problems But But, but, you know, it's going to be mind bogg It's almost like a singularity moment

What's going to happen once we have agents That can also transact with each other, right? Like, so I can have an agent Which brings in crypto. Which has a crypto wallet That can go hire other agents. Sure. I'm my agent.

I'm going to raise my agent to be a trillionaire. Exactly. Yeah, we're probably going to be seeing autonomous agents with their own wallets and net worth and all of that. You know, one thing I heard you speak about is the future of the company. I want to talk about that because we have a lot of entrepreneurs in the audience and on, you know, and the people who will listen to this around the world. I mean, most corporations today have been built in a very siloed, very, I think you use the term of factory production mechanism.

And it's, you know, Saleem, you railed against this old age, right? John Hedel talks about this. All big corporations are designed for two things that are designed for predictability and for efficiency. Right? But in today's world, it's so volatile, you need to be architected for flexibility,

agility, adaptability, speed. That's right. Therefore, I actually did a talk with you on stage, Peter, a few years ago in Toronto, and the title of the talk was the death of the corporation. We're seeing corporations go to platforms, and platforms go to ecosystems. There's an economist, I think his name is Thomas Coase, and he has this theory. Ronald Coase.

Ronald Coase, yeah, the theory of the firm. Yeah, so Ronald Coase wrote a nine-page paper in the 1930s, citing that the reason we'll have bigger and bigger companies that transaction costs inside a big company are cheaper than outside a big company. He won the Nobel Prize in Economics for that nine-page paper. We, in the book that Peter and I just wrote, the second edition of exponential organizations, we actually declare Koso's law dead because the transaction costs inside a big company are way higher than doing it on the outside. Exactly, exactly.

Therefore, the big companies as a category will start to decline very rapidly. That's right. So, like, my takeaway from that paper is that full-time employment is a bug of the system, not a future. Yes, yeah. In reality, if... And doing what you're doing what you're talking about, told there's a bug. The one way you told us a bug as well. I mean, honestly, employment is a bug. And we'll tell my kids that. Let's just go so. So the reason why firms have to hire

people, train them, and then put them in this somewhat of a draconian hierarchy is because of the transaction costs of going out of the market tend to be cost prohibitive. If you lower that, for example, Uber lowering the transaction costs from like looking for a taxi like that to like a couple of clicks a button. If an agent can go hire human or can go hire another agent, then we're going to see companies shrink and we're going to see a lot more entrepreneurship

and the market is going to be more dynamic. I think more people will get rich. I think there's going to be just a fast, you know, a fast life and death cycle of companies. Dave, do you want to jump into your question from earlier about Replit coding replet? You want the easy one or the hard one first?

No, give the hard one. All right, I'd love to get your insight, actually, on this very specific topic. You know, we're very quickly going to get through all of the code out the legacy garbage part of society, right? Right now you can use Replit to build things so quickly, and there's all this garbage software that was, you know, it was cost prohibitive to hire enough engineers to do it right, now you can just do it. But if we get to the point where you can create a thousand times more code, 10,000 times more code, then it's all going to be greenfield. What's the net new thing we can build that adds value that we just couldn't afford to build before?

You have a lot more vision into that than anyone else because you can see all these, you know, millions and millions of users and what they're starting to build. So what are you seeing them build today that gives you insight into what they'll be building in the greenfield future we're coming into? I think agents. And the reason we built this, what we're calling the agent stack of Agent 3 is that's sort of the next evolution of software. When you look at most pieces of software, especially inside the enterprise, they're trying to solve a problem, a very specific problem. And usually you create a piece of software

and then someone will use that piece of software to solve that problem. That intermediate step is not needed. You create an agent programmed and imbued with the domain knowledge to be able to solve that problem. And then you run that agent in an autonomous fashion inside the organization. And that's what we're seeing our users struggle with.

It's like, hey, like, I want to build this automatic process, but it keeps generating an app. And yes, maybe I need an interface for the chat app to talk to the agent, but I want to be able to run it based on hooks, based on time triggers, based on, so we're building all of that. And that's how our roadmap works. It's like we're seeing what people are trying to do. They're looking at AI.

They're looking at age. They're seeing that LMs can make good judgment. And I think that's the main thing that happened in the past six months to a year is LMs with agents. they can do these multi-step reasoning, especially with now GPP5, the scale on the thinking of GPD-5. Like, GPD-5 base is so dumb, but GPD-5 high is so, so intelligent.

So we're able to scale that intelligence, and so right now, you know, these agents can be nodes in very nuanced decision-making. So we had some conversation on this stage earlier today about some fairly sketchy research out of MIT that indicated that, look, in the real world corporate environment, people are getting virtually no benefit out of AI-generated coding.

And you're actually living it. Do you have any estimate on how much of Replit is built by Replit or what the force multiplier is inside your own organization? The impact of Replit on our organization is more in how we run the company and even outside of engineering than it is inside of engineering, counterintuitively so. So using Replit to run. run replet.

Yes, we're using replet to run replet. So, you know, our salespeople, for example, will write some software using replet that use an LM that analyzes all their call transcripts to find the main reason why people might not buy Replit, they will take that, they will go to the product team, that like, this is the main thing, you know, we're missing, you know, this compliance thing. So I really want to touch on this topic. Yeah. Because, you know, you're probably the old guy in your Y Combinator cohort, right?

Yeah, yeah, yeah. And so when you look across all these people who revive coding or building tools, they have virtually no large-scale management experience. And they're like, yeah, this helped me writing code more efficiently. But it's actually even better at managing large organizations. That's right. You know, all the planning, all of the, you know, like who did what. And they have no experience with it.

I've got like about a $2 to $3 million a week payroll. And it's a gold mine in knowing what's going on across 1,000 people that are all scattered around. The embedded data in that payroll is a gold mine. It's a goldmine, and it's so much easier than the harder challenge of just sweepbench. And you're probably the only guy in that Ycombinator world that's actually running and has run a large organization and sees its advantage as a management tool, not just as a leaf node code development tool. Yeah, it's almost like the engineers using AI as stable stakes, and I would certainly challenge a lot of those, I think you call it dubious, which I agree a lot of those studies that are coming out.

But it's stable stakes, and there's so many products you can pick out of. I think where Replit is really shining today in the enterprise because of the vertical integration that we do, and because you don't need prior coding knowledge to do all that, and by the way, we're launching all these integrations in all the ways in which it can hook into your data infrastructure. Because all of that, it'll help you run the company, your legal team. I mean, the gentleman here, we were talking earlier about his finance team at Dolby using Replit to kind of run the finance team. And I think that's really profound, and not a lot of people are talking about that. It's really profound.

You are going to run into some data sovereignty issues and on-premises issues, et cetera, right? Because there's a lot of people that are, we know a medical CEO that got excited by ChachupD and uploaded all their patient data into Chapitubin, and now they've got massive legal issues around that. So you're going to face some of that. So that's an engineering solvable problem. Or a lawyer solvable problem. What I'm excited about is the fact that, you know, the current stack of business software ERP systems will take

certain specific, my co-founder, her age talks about this a lot, will take certain use cases and automate those. But now you can automate every minutia-level use case across a business, all the hundreds of little things that people are doing. And anybody can automate those. So the legal team will

automate contract drafting. And they'll just start replicating those and getting that out there. The best thing that happened over the last 10, 20, maybe 30 years is the digitization of these organizations and governments. And the system of record, you know, systems such as the ERPs, the HR systems, the CRM, Salesforce, whatever, that is the infrastructure for which the next set of automation and the

invention of the future of work. So we're exactly at the right time to be able to build these systems. All right. You want the hard question? Yeah, let's go. I want to talk about the vibe coding wars. All right, that's good. You want to do it or I will?

All right, we'll start. There's the easier and the harder of the two. Okay. The easier one first. All right, lovable versus replet. Okay, good, good answer. All right, first of all, if I look three years in the future, is it all boats rising with the tide,

or is one of you going to kill the other? Are you going to differentiate and become different? Do you think about it every day? Do you not focus on the competition? How's it going to play? I think it's already quite differentiated. I think that the reason people kind of conflate all these products is because you can put it in a prompt and get something out.

Most of these, most of our competitors, generate a front-end app. For you, like, connecting to the database, you need to go grab a database somewhere else. For you to deploy them, you need to find, go to AWS and deploy them. The replication actually provisions a database for you. It runs migrations of the database for you.

Provisions of production database for you. Does that separation for you. It helps you with the deploy process. So the platform, and the reason we've been around for 10 years, we've been building the depth of the platform. Because we knew that the challenge of programming and making software is not just about the code. It's about a lot of it's about the infrastructure.

You think you're ahead of... The chart I'm looking at here is there's full stack and front end only. Right, yeah. Right, and you're at the full stack side. And then there's non-technical and technical, and you sit sweetly in the middle. That's right. What about Vursell then? Because Vurcell's kind of...

Yeah, I think of them as being deployed historically. And now suddenly they're... Well, on this chart, Vurcell's down here in the bottom right. Yeah. Versailles, I think, with V0 start as a design tool. Yeah. And now, I think everyone is seeing that the value is where Replit is, because Replit is winning all the deals.

Okay, everyone's trying to come after us, but good luck. Okay, well, that is a perfect segue into the really hard question. So we were over at OpenAI headquarters a few weeks ago with Kevin Weill. And, you know, when they launched GPT5, they brought Michael Truitt on stage, made a big deal out of Cursor, but just a couple weeks prior they were trying to buy Windsurf. It's like, okay, wait a minute. It's like, you were friends today in an alternate world where Microsoft didn't torpedo that deal

because would be the worst arch enemies in the world today. So, you know, you've got these foundation model companies. Are they going to come in your direction, or how's that going to play out? The really interesting thing about our market is it's total war. They're no friends. Really, there isn't.

Like, you know, arguably in, you know, prior eras of Silicon Valley and tech, they're, like, natural allies that could form. Right now, today, you're an ally with someone else that are going to come out and attack you tomorrow. So, like, you know, you have to be super paranoid if you're starting a company in this stage. Are you super paranoid? Of course. Like, you know, I think competition is going to come from everywhere. Like, you know, Google is one of our closest partners who we spend hundreds of millions of dollars with them, and they have three competing products. I still love them for it, and Sundar still uses Rapplet instead of the other ones.

But, you know, it's a shout out to Sundar, yeah. I think of this as the Dropbox thing where you have OneDrive, you have ICloud, you have Google Drive. But just by focusing on that one product area, wholeheartedly Dropbox does very well. And I think that will be the same, no matter what other people do, their interests are so varied. There's a rising tide of near infinite demand. Yeah, there's that. But also, I remember when I was at Yahoo, you're managing across a large...

You were aware? I was on Yahoo, by the way, too, for seven months. I'm not going to come over there for a few. Anyway, you're managing a big company across 120 different web properties, and you have to allocate resources, et cetera, et cetera. One dedicated team is always going to be you. Right.

Always. You should grab all over that, by the way. If you said, yeah, we want to be just like Dropbox. Dropbox is a great company, But I think that it's a lot more brutal than those errors because... Well, you know, back, we'll say a decade or two ago, you had this politeness between the big companies, right? Google would say, Microsoft would say to Google, don't come into the office space and don't improve Google docs too much. We won't come after a search too much.

Really? It was an implicit. This is polite thing. Oh, my gosh. Now, all the gloves are off now. The gloves are off. Well, because everything's converging on AI. You know, they all settled into their swim lanes

and they said, okay, Apple, your phones are fine and, you know, Bing will suck forever. We'll just commit to that. But then, you know, but anyway, it was all detente. And now all of a sudden it's all out war because everything is converging on, oh, wow, all that matters is AI. So I 100% agree with Salim on the idea

that the customer segment focus is a superpower and that's very important. That being said, I do think there's a potential that these products will differentiate but eventually converge. So Agent 3, you can see it tomorrow, it's the most autonomous agent on the market. And

we ran it like two days ago for four and a half hours because we, right now we can test its own code. And you and I were talking earlier about is AGI here? Is AGI not here? The thing that would make me feel like

you can define the current era of models as AGI is because if they get real and good environmental feedback, you can run them endlessly and have them actually try to solve a problem. And invariably, what we're seeing in software, if you're able to spend as many tokens as you can and they're running in a good environment, they're solving these problems. So I've heard you say that it's this sort of universal expansion of coding capability

that you believe is going to lead us to AGI. Yeah, I think that's... I think that's... I think, I think AI here already, personally. I think we passed the Turing test didn't notice. I think we passed AGI, haven't noticed, and we'll see if we noticed. I call BS on this. We're talking about something we can't define, we can't

measure, and we don't have a test for. Yes, that's why I can say, I can make... At least the Turing test has had a measurable outcome, right? It was very clear. AGI, and now we're talking about ASI, and we still have no idea what any of that means. So, anyway, aside from that, I'm fine with the conversation.

I have a question, though. Just to answer that, like the insight from Turing is that the Turing machine, the computer, is the ultimate problem-solving machine. Yes. That scales infinitely. That scales infinitely. And that is the milieu in which intelligence will use in order to solve any problem. So whatever AGI system you can imagine, ultimately it's going to be writing code to solve problems.

That's my fundamental belief. Yes. Okay. Interesting. Salim. Can I take this in a slightly different direction? Maybe. Just to show events.

How many of you at this event over the last couple of days have had your mind's completely blown? Just to show vans? A bunch of you. All right, I'm going to try and go for it right now. So we have biology, which is essentially software. You have 50 trillion cells in the human body, each governed by the DNA.

Now essentially we can now edit... 3.2 billion lines of code. We can now edit the DNA like you would software, essentially a human being. is now a software engineering problem. When you apply Replit to biology code, are you seeing that happen? When do you see that happen?

What do you think of the implications? The name replet sounds very biological. Yeah, completely. Is that a coincidence? It's right there in the name. So ultimately what, the fascinating thing about LMs, which is similar to the DNA code, is how much knowledge it compresses. I downloaded this app called, I think,

Full Moon on my phone that allows you to download open source one billion parameter models. Because I was in the plane. I like to read books in the plane. I know there's Starling in the planes right now, but I pretend they don't exist because there's like 12 hours that I can, like, read a book in peace. And so I was like, but I need to look some stuff up.

So let me download, like, Deep Seek 1B. And it's amazing how good it is. I mean, obviously, hallucinates and you need to check its work and everything like that. But one billion parameters, what is that? less than one gigabyte. And that embeds a large portion of human knowledge. And if you run it for longer like R1 does deep seek R1,

you can solve problems as well. I can't wait to download it into my neuralink chip. Yeah, exactly. Like these systems are efficient. And it shows you that something about the nature of the universe that there's so much data noise at a redundancy that you can really boil things down to like very, small, like, sort of, you can boil things down,

like the essence of things that was actually very simple. Isn't my mind-blowing to you, actually, the data? Like, if you take even a big, big, big neural net, you know, you get about a trillion parameters, it fits easily on your laptop. And this is all human knowledge, everything on Wikipedia, everything from the internet, all learned, and, you know, about $100 to $200 million training process,

and it all compresses down, and it just fits on your laptop. Now, the processing is way too slow, so getting it to run on. So that's why you need to run the billion parameter model. But putting that aside, just the amount of storage that we have is just incredible. And the amount of product, and is why NVIDIA is so much. The processing is not so incredible. We're going to look back on this period and just say, God, we've been so inefficient

because I think the ultimate LM will probably be about a billion parameters. Isn't that crazy? And, you know, back to DNA and bio, like we're seeing, like, opening I had a big results. Obviously, Alpha Fold before that. clearly we have the data just trained and now with, you know, the way we train LLMs is unsupervised learning so you can like find

this compressible information out of this massive sea of a lot of noise without supervision, without data labeling. And I think LMs will get really good at that. And once they're able to generate, be able to like code bio, the thing on us

would be to create these environments. And that's what I see companies our company as a, you know, as a tech company, we're about creating environments. We're creating, we want to be the best habitat for LLMs. Right now, the best habitat for LLM's is to be in a virtual machine environment. But in the future...

One more question to close out the competitive landscape topic. So if I look at the companies that do coding, so, you know, REPLIT, Lovable, Cursor, Winsurf, let's see. You guys are probably the only one that built foundation models, because you were there early. So you actually trained these things.

I think everyone else just picked up, you know, open AI or whatever later. So then you're like, okay, well, we're not gonna spend $100 million about training our own model. It's just way too expensive. So now we're on top of the other guys. Oh wait, now our evaluation is in the billions.

You're raising tons of money, that's confidential I just found out, so raising tons of money, and you're not actually planning to consume it, right? You haven't spent the money from your last funding, but you have a war chest now that actually is plenty big to actually train foundation models. So is there a version of your future where you get back into the foundation model training business?

I think it's cyclical. And the way it works, when we went from GPT2 to 3, there was a period of time where there wasn't a lot of progress. The tricks were known. So we sort of knew how to do pre-training. The labs was busy cooking up a mixture of experts and GPT4 or whatever. There's a period of time where startups can actually use their data and add a like. on top of these foundation models and fine-tune do all of that. And so Replit Code 3B, 3 billion parameter was state-of-the-art coding model in 223, and that

put us ahead of the competition because no one else had access to a model that cheap and that easy to run. But then the foundation models did another big jump, you know, the scaling up of the models. I think we're reaching a point where it makes sense to go back into training. Cool. because we have a certain data set that we think is very helpful, especially for the kind of use cases

that Replit does. And we're starting to find places in our agent architecture because it's not just one agent. We have so many sub-agents. We have the testing agent. We were able to build a computer-use model, a computer-use architecture

that is three times faster and 15 times cheaper than the state-of-the-art from the big labs. And I think that we're going to continue finding these things. until we get the next frontier model that actually, and that's what happens is that's the history of machine learning, the bitter lesson.

Then you can pour data and compute, and you can consume all these use cases. And so just being open-minded about it, some companies were like, we're going to go, we're going to live or die by training. But for us, we're ultimately trying to solve a problem and pursue a mission, and we'll just do whatever needs to get there. I really thought in a million years,

that you would never have the guts to say that. You know, I mean, you have the capital to do it. But I could see there's actually some MIT alum, brilliant Christian Bailey here and others. Their eyes lit up as soon as you said that. Because I think that mission statement that you just outlined will attract a ton of talent that doesn't want to go to someplace boring. They want to do something, you know, super. I want to close us out going back to your roots.

So listening to us, you know, our Moonshots podcast, when the three of us discussed, taking moon shots, there are entrepreneurs around the world who are a young Amjad someplace in the Middle East, someplace in Southeast Asia, in South and Central America, who are dreaming big. What's your advice to them? What's your advice to inspire them and how should they think about their next few years? Because do you agree that the next few years are really the game? Yes. Yes. It's everything. My advice will probably get them into trouble, but

I think that's part and parcel. We've had the conversation, Skip University, go build a company. Yeah. I mean, is there, is there, is there, is there, I think all those are details. You know, if you get into Stanford, get in and then drop out. What really matters, what really matters is

throw away most of the advice that your parents and your society sort of give you. Because it's such a dynamic world. Right Don't do what you're told. Yeah, and conforming is the worst thing that you could do. Don't do what you're told. And it'll take real, you know, self-programming to actually exit that mindset of really doing what you're told and try to, try to really think from first principles about what you

find your, in your purpose. You know, there's, I love the quote from Mark Twain, he says, two important days in your life, the day you were born, the day you found out why. Yeah, yeah. And I think, you know, the meme, the midwitch curve, I think has a real true. to it in that you can really over your over plan your life, overthink your life, following your intuition, and you know being attracted to it to a problem I think there's a you you were talking about the

human spirit right there's some and I agree with you there's something essential about the human spirit that probably not going to be captured we're saying tweeted last night the one thing AI will not replicate or displaces the human spirit and I think because we're humans we live among other humans, we can see problems that are very important for our communities and the people around us that are not going to be embedded in the machine. And so following your intuition for what problems you want to solve will probably net out a truly unique and differentiated position in the world for you to have, I mean, I'm very blessed to have a long-term

mission. Like, I sort of found my purpose, and I think the way to do it is to, follow your curiosity and tuition. If it takes your university, then it's fine. If it takes you a building company, that's fine as well. The implication of this is pretty profound because take education, right? We've been doing education from a supply side perspective for 100 years where you go get a skill, you become a developer, an accountant, a doctor, whatever, and then you go to the job market to sell those skills.

All of that evaporates now, and now we have to move to the demand side and say, what problem do you want to solve? That's right. And now you get the tools and the technologies, the techniques, the replica, repository. to solve that problem. And that I think is going to be the most exciting change that we're going to see. Yeah, a place to close out

and respect the clock here. Ladies and gentlemen, I'm Jod Masad. Thank you. David Blundon, Salinas Mal. Oh. Every week, my team and I study the top 10 technology metatrends

that will transform industries over the decade ahead. I cover trends ranging from humanoid robotics, AGI, and quantum computing to transport, energy, longevity, and more. There's no fluff.

only the most important stuff that matters, that impacts our lives, our companies, and our careers. If you want me to share these metatrends with you, I writing a newsletter twice a week, sending it out as a short two-minute read via email. And if you want to discover the most important metatrends 10 years before anyone else, this reports for you. Readers include founders and CEOs

from the world's most disruptive companies and entrepreneurs building the world's most disruptive tech. It's not for you if you don't want to be informed about what's coming, why it matters, and how you can benefit from it. To subscribe for free, go to Demandis.com slash Metatrends. To gain access to the trends 10 years before anyone else. All right, now back to this episode.