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

Throw away most of the advice that your parents and your society sort of give you because it's such a dynamic world right now and conforming is the worst thing that you could do. Amjad 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. I 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, the barrier how hard it is to write software.

We're 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 the Internet, you should be able to build something great. And we're going to find talents 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 a moonshot.

Ladies and gentlemen.

Welcome everybody. Allow me one second to just bring onto the stage my two moonshot mates, Dave Blunden. Dave is the creator CEO of Link Ventures. Link Exponential 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. Salim Ismail over here, another moonshot mate, 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 John. 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 from you.

If I were to summarize your life, I would go coding an Internet cafe in Jordan, rejected three times by yc, finally gets in, builds repl it and is now helping a billion people code. Is that a.

That's Right. That's a good summary.

Yeah, 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 positively 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. So I used to go code in Internet cafes back in Jordan. And one thing I realized that they had computers all over 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 in 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 when 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 Starlink antenna on the dashboard over there. And I'm coding on replit on the way up here.

This is the future. That's awesome.

I tweet this out. Hopefully.

Did air traffic control know about this?

That's a hard 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 very important moments was Brendan Icke, the inventor of JavaScript, part of the Mozilla browser, 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, 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 was 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 Codecademy started up on the technology that I built and Codecademy became the number one coding school in the world. How many people here learned how to code on Codecademy? 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 just reflecting on my own journey. If a kid from Jordan is able to come to the US 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 is like, how do you make it so that anyone can build something that could impact millions of people?

Love it. Dave.

Sir?

Floor 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 O1 visa to go to New York and I was a founding engineer at Kotacademy and then left Kotacademy 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 O1 visa is?

Yeah.

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.

Yes.

And so it's a different Story from us. Go ahead and finish.

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 great 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 Symbian phone. I was like, can I code on the desk so I can do something more interesting than listening to this teacher?

Blabberg, I love the fact that you hacked the Y Combinator application and put a Rick roll 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 partly it kind of always works out. I don't know if I'm just lucky, but even quitting Codecademy 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 US and I went and applied to a bunch of companies. One company I was really excited about at the time, Mark Zuckerberg was talking about the vision of Internet.org right. The idea that we're going to connect everyone in the world. Everyone will have access to the Internet. That's an extension of my mission of if you have access to computers and the Internet, you should be able to build something great. And we're going to find talents 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 and 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, it 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 like I was really excited about this idea of internal.org and I started working on Android because I thought that Android is the main device that most people are 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? Started working on that and found other people at the company working on it. That became React Native and React Native made it so 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 because there's part of Facebook's written in it. Discord is written in it. And so I think that just like this idea of not over planning your life, just going about life and finding interesting problems to solve and solving them usually will not have the right thing.

Well, so that experience, that life experience is a perfect segue into my question which is using repl.it to discover talent all over the world. So there's two things I'm dying to ask you. One of them is 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.

That's right.

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 incredibly 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, 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? I mean, you're not, you're not going to forget that?

Actually, I actually kind of forgot because I would apply. I would apply every time.

Oh, I see.

Then I would record a video every time. You know, the YC video is like a hostage kind of situation. You're sitting there.

How many of you have applied to y Combinator?

3 or 4 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 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 write a newsletter twice a week. Sending it out is a short 2 minute read via email. And if you want to discover the most important meta trends ten years before anyone else, this report's 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 dmandis.com metatrends to gain access to the trend 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, how fast it's changed. I remember I was talking to Dave earlier that when I was, when we were Both undergrads at MIT, there was a course called 6111.

6111.

And I would literally build my computers with and gates nand gates and nor Gates and would hexadecimal code them for 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 we eventually got to eeproms. But it was extraordinary how far it has come. I want to talk about the future of coding. One second. I had Imad Mustaq 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 Mustaq 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 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, we used to kind of program in machine code. It's not really programming. It's kind of similar to this rewiring of the machine. And she wanted people to program in English. So the C programming language and the high level programming languages that she invented, she called them English. And it's sort of similar to what, you know, Andrew Karpathy recently said. It's like English is a hot new programming 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 because programming is going to be in English. Of course, that vision sort of went away and programming became an industry. And you had to 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.

Your idea generator that gets compiled in.

Code that's right, sorry. 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, et cetera, 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 then you go around optimizing, et cetera, etc. Etc. As people are vibe coding, those skills aren't there. It's the same thing. Like a student that's using ChatGPT 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 assembly line? Yeah, who the hell wants to be writing assembly and machine learning or hexadecimal? Right. We'll just do it at the higher level.

And.

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, you know, someone would make an argument in any history of computing where you say, okay, this is too high level, you're abstracting the NAND gates and the registers and all of that, you're not gonna 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 is too high level becomes low level. Now, C programming is low level for them at the time, you know, the machine code people, it's like, what is. You know, this is bullshit. Yeah, this is bullshit. Right?

Interesting twist on that though. If you say, I didn't know Grace.

Hopper, I didn't curse 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 persistence or pointers or indirection or stuff like that state. 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.

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 series of prompts. I have a book I'm 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 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%, which is counterintuitive if there's an over. You know, you think it's an oversupply, which tells me that the value of programmers is massive and the exponent is greater than one on this. What happens when we have a billion programmers or 2 billion programmers? What's your vision for that world of abundance? Right. 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 on 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 when we have a billion, two 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 Altman and Paul Graham saw Replit on. On Hacker News.

So just to tell the story a little bit, which I've heard 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 got to apply again.

Yeah. So what happened was I wake up one morning in late 2017. We had started the company in 2016. It's been a project for a lot.

Longer than overnight success. After nine years of hard work.

And.

We, and you know, 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 door 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, 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 Mountain View. And so I show up there and it is Neuralink and OpenAI. And I didn't know about either. Maybe I've heard about OpenAI a little bit. So I go into OpenAI and Sam was sitting there and he says, we're really interested in what you're doing. And by the way, Paul Pham joined Hacker News because at the time I was like building the system and like blogging about it, wrote some interesting blog posts 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. It's like 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, you know, there's so many barriers in their way. Not only the barrier of capital and the barrier of, you know, getting attention, marketing and all of that, but even before that the barrier of programming and the power of how hard it is to write software. And so we go through all 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, yeah, I would love to have you. And you know, you should join. And we have the kickoff tomorrow. And I say okay. And then he shoots me another email. It's like formality, you should still apply. It's like, oh fuck, I'm not going to go through the YC application all over again.

And how long is the YC application?

I mean if you want to do a good job at it, it'll take you an. I'm very lazy. I was like, it's a, you know. So couldn't have replit write it for you, huh? No, at the time I couldn't. There was no LLMs and so I had to do it manually. But I did a very last 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 we go to the next day. The next day is the kickoff. They do the late interview. So they always have this tradition of like letting a few startup in at the same day. And so we said the whole day we're waiting and then finally the door opened and they told us, you can come in. And so we go in and it was like Gustav Adora and then Michael, CEO of Y Combinator at the time. And he was a big guy, right? And I'm shaking their hands and I felt Michael kind of squeeze 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 recrawl us? It turns out as we were sitting outside, they were getting rickrolled inside before the waiting bus. They clicked on the video and they.

Got this rickroll video place. 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, but what are you.

Thinking at this point?

I think 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's like, ah, we're not, we're not getting. After we did that, we're not getting in. But then I get a phone call from Adora 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 important was the YC experience to you, I mean, would you have accelerated and gotten to this point without it?

I think, 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, Paul and you just raise your ambition a lot. Being around all the ambitious Silicon Valley.

Thing in general, I think that's what matters.

The question about talent is very important. There's talent everywhere. But I think the density of network at places like Stanford.

Paul Graham had that great quote that being a startup founder is only about 1 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. Incredible story. But if anyone from Stanford administration is in the room, I know Eric Brynjolfsson 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, 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, pull 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 you're from and it's, you know, it's not, I'm not faulting them for it. It's a heuristic. If you're Stanford dropout, you're like, you know, Stanford did the work for us to choose you and therefore we're just going to piggyback on that. But in reality, again, talent comes from everywhere.

Philip Rosedale, creator of Second Life, I don't know if you know him. One day Philip says, do you know why there are more startups in San Francisco than anyplace else and why the success stores 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, right. There was Austin, there was Cambridge, there was, you know, Miami starts to pop out. But the notion is if, if you, you're taking the risk of starting a company and you fail, instead of going back to your mom or dad's bakery to work, your friend down the street has another startup you can go and join. And so it is a density issue.

But I think increasingly that is moving online. I know you're good friends with Balaji and his idea of Network State and all of that. And I was enjoying 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 like they come here quite often. They meet with VCs, they're, you know, they incorporate it in Delaware and when they go into meetings, they, you know, it's like, it's almost like they're headquartered 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 if you figure out what's needed at almost zero cost.

That's right.

So 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, starting a business. It's about being able to solve problems in the world and taking initiative without your boss telling 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 line revenue for the company because although he didn't, he wasn't an engineer, he was able to increase conversion of you know, home buyers at the company. And then suddenly everyone heard about the. Heard about that story. Now we have 500 licenses at Zillow. And everyone is recommended to Vibe code because they want more entrepreneurial people. That that person got more agencies.

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? Standard problem.

But here's why Vibe coding 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 happen inside the organization as well.

We call this pdi Permissionless Disruptive innovation. Because in the past, whenever you want to do 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 Buterin 18 year old kid gets together with a few friends, they ignore their professors. 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, providing the scripting language to take anybody's massive purpose, as we call it and start making it an actual reality.

So here's where it gets even weirder and stranger and more abundant is that it is not gonna be just you doing the disruption. It's gonna be your agents. 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 be able to create like an HR agent.

How long you've been working on that?

It's been, it's Been, you know, about six months since, since Agent two. So I think every six months, you know, commensurate with the progress in LLMs, we're trying to predict what is the next set of capabilities that would unlock the next version. So we already know what Agent four is going to look like. You know, we take it back, we try to build it and over time, you know, the first version of it would be crappy, but then the foundation model and everyone catches up.

Yeah.

I always tell entrepreneurs build crappy product.

So agents can basically spin off their own agents and there's going to be an agent. What could possibly go wrong?

Look, I don't, look, I don't, I don't focus on what goes wrong, other people's problems.

But.

But, but, but, but, you know, it just, it's, it's going to be mind boggling. It's, 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 bring, which brings in crypto, which.

Has a crypto wallet that can go hire other agents.

Sure.

I, my agent is, 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.

One thing I heard you speak about is the, the future of the company. I want to talk about that because we have a lot of entrepreneurs in the audience and you know, the people who, who will listen to this around the world. I mean, most corporations today have been built in a very siloed, very, I think you'd use the term, a factory production mechanism. And it's, you know, Salim, you railed against this.

I mean, John Hagel talks about this. All big corporations are designed for two things. They're 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.

Right, yeah. So we, 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 Coast's law dead because the transaction costs inside a big company are way higher than doing it on the outside.

Exactly, exactly.

And 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 fix.

Yeah.

In reality, if and doing what you're.

Doing what you're told is a bug.

Doing what you're told is a bug as well.

I mean, honestly, employment is a bug.

Tell my kids. Yeah. 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, you know, for example, Uber lowering the transaction costs from like looking for a taxi like that to like a couple clicks of 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 and 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?

Replit, you want the easy one or the hard one first?

Give the hard. Give the hard one.

All right. I'd love to get your insight actually on this very specific topic. We're very quickly going to get through all of the code out the legacy garbage part of society right now. You can use replit to build things so quickly. And there's of a lot 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 millions and millions of users and what they're starting to build. So what are you seeing them build today day 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 three 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. 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 automation 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, we're building all of that and that's how our roadmap works is like we're seeing what people are trying to do. They're, they're looking at AI, they're looking at it, they're seeing that LLMs can make good judgment. And I think that's the main thing that happened in the past six months to a year is LLMs with agents they can do this multi step reasoning, especially with NAV. GPT5 which this scale on the thinking of GPT5 like GPT5 base is so dumb but GPT5 high is so, so intelligent. So we're able to scale that intelligence. And so right now 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 course 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 RELET or what the force multiplier is inside your own organization?

The impact of RELET on our organization is more in how we run the company and even outside of engineering than it is inside of engineering, counterintuitively.

So using REPLIT to run replit?

Yes, we're using REPLIT to run replit. So you know, our salespeople for example, will, will write some software using replit that use an LLM that analyzes all their call transcripts to find the main reason why people might not buy relet, that will take that, they will go to the product team, they're like, this is the main thing, you know, we're missing this compliance thing.

I really want to touch on this topic because, you know, you're probably the oldest guy in your Y Combinator co. Yeah, yeah, yeah. And so when you, when you look across all these people who are vibe coding or building tools, they have virtually no large scale management experience and they're like, yeah, this helped me writing code, write 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 dollars a week payroll and it's a gold in knowing what's going on across a thousand people that are.

All the embedded data in that payroll is a gold mine.

It's a gold mine and it's so much easier than the harder challenge of just SW bench. And you're probably the only guy in that Y Combinator 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 is still 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 are coming out, but at 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, to do all that. And by the way, we're launching all these integrations and 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, 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, etc. Right. Because there's a lot of people that are. We know a medical CEO that got excited by ChatGPT and uploaded all their patient data into Chapter GP and now they've got massive legal issues around that. So you're going to face some of that. That's an engineering solvable problem or a lawyer solvable problem. What I'm excited about is the fact that the Current stack of business software ERP systems will take certain specific. My co founder Harish 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.

Right.

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, of the future works. 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, let's go. I want to talk about the vibe coding wars.

All right, that's good.

Or I will.

All right, we'll start. There's, there's the, the easier and the harder of the two. Okay, the easier one first. All right, Lovable versus replit. 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 competition? How's it going?

I think it's already quite differentiated. I think that the reason people kind of conflate all these products is because you can put in a prompt and get something out. Most of these, most of our competitors generate a front end app for you to like connect them to a 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 replit agent actually provisions a database for you. It runs migrations of the database for you, provisions a production database for you. Does that separation for you, 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 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?

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 Vercel then? Because Vercel is kind of. Yeah. I think of them as being deployed, deploy historically, and now suddenly they're on this chart.

Vercel's down here in the bottom, right.

Vercel, I think, with V0, started 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. Everyone's trying to go coming after you, 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 Wheel, 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, you guys would be the worst archenemies in the world today. So, you know, you 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. There are no friends. Really. There isn't, arguably, in 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. You know, you have to be super paranoid if you're starting a company in this, 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. 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 Replit instead of the other ones, but, you know, it's.

Shout out to Sundar.

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.

The 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 at Yahoo, by the way too, for seven months.

Yeah. Amjad and I are going to come over there for a Few. Anyway, you're managing in 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 beat you, right? Always.

You should cross 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 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 search too much. And it was really late. This is polite thing.

Oh, my God.

Now all the gloves. 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, yeah, sorry. But anyway, it was all daytime and now all of a sudden it's all out war. Because everything is converging on. Oh, well, all that matters is AI.

So. So 100% agree with Salim on the idea that the customer segment focus is, is a superpower. And that's. That's very important. That being said, I do think there's a potential that these product will differentiate, but eventually converge. So Agent three, 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 right now it 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.

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 AGI is here already, personally.

Right.

I think we passed the Turing test. Didn't notice. I think we passed AGI, haven't noticed. And. And we'll see if we notice.

I call BS on This, we're talking about something we can't define, we don't. Can't measure, and we don't have a test for.

Yes, that's why.

So can we please add it?

At least the Turing test had a measurable outcome. Right? It was very clear. AGI. 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.

Can I just to answer that, like the, the insight from Turing is that the Turing machine, the computer, is the ultimate problem solving machine that scales infinitely. That scales infinitely. And that is the. That is a 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.

Interesting.

Salim, can I take this in a slightly different direction? Maybe just a show of hands. How many of you at this event over the last couple of days have had your minds completely blown? Just a show of hands, 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, right? You have 50 trillion cells in the human body, each governed by the DNA. Now essentially, we can now 8.2 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 name replit? Sounds very biological. Now that. Yeah, completely. Is that a coincidence?

It's right there in the name.

So ultimately, what the fascinating thing about LLMs, which is similar to the DNA code, is how much knowledge it compresses. I downloaded this app called I think Full Moon on my, on my phone that allows you to Download open source, 1 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 seq1b. And it's amazing how good it is. I mean, obviously it hallucinates and you need to check its work and everything like that. But 1 billion parameters, what is that? Like less than 1 gigabyte.

Yeah.

And that embeds large portion of human knowledge. And if you run it for longer like R1 does deep seq R1, you can solve problems as well.

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 is actually very simple.

Isn't it? Mind blowing to you actually the data, like if you take even a big, big, big neural net, you know, 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 it's 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 pro and this is why Nvidia is worth so much the processing.

I think we're gonna not so incredible. We're gonna look back on this period and just say God, we've been so inefficient because I think the ultimate LLM will probably be about a billion parameters.

Isn't that crazy?

And you know back to DNA and bio like we're seeing like OpenAI had a big results obviously alphafold before that. Clearly we have the data to train. And now with, you know, the way we train LLMs is unsupervised learning. So you don't, 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 LLMs 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 what I see company, 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 alarms is to be in a virtual machine environment. But, but in the future.

One more question to close out the competitive landscape topic. So if I look at the, the companies that do coding so you know, replit, lovable, cursor, windsurf, blitzy. 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, OpenAI or whatever later. So then you're like okay, well we're not going to spend $100 million build training our own model. It's just way too expensive. So now we're on top of the other guys. Oh wait. Now our valuation 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, is there a version of your future where you get back into the foundation model training business?

So I think it's cyclical and the way it works. When we went from GPT 2 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 mixture of experts and GPT4, whatever. There's a period of time where startups can actually use their data and add a layer on top of these foundation models and fine tune do all of that. And so replit code 3B3 billion parameter was state of the art coding model in 2223. 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. 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, from the big labs. Wow. 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 it consists then you can pour data and compute and you could consume all these use cases. And so just being open minded about it, some companies were like we're gonna.

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 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 this out. Going back to your roots. So listening to us, you know our Moonshots podcast, when the three of us discuss taking moonshots, there are entrepreneurs around the world who are 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. But I think that's part and parcel.

We've had the conversation. Skip university, go build a company.

Yeah.

I mean, is there.

I think all those are details.

If you get into Stanford, get in and then drop out.

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 now.

Don't do what you're told.

Yeah. And conforming is the word thing that you could do. Don't do what you're told. And it'll take real self programming to actually exit that mindset of really doing what you're told and try to really think from first principles about what you.

Find your inner purpose. 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.

And I think the meme, the Midwood curve, I think has a real truth 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 a problem. I think there's a. You were talking about the human spirit, right? There's something, 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 displace is the human spirit.

And I think because we're humans, we live among other humans, we can see problems that are very, you know, 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 world for you to have. I mean, I'm very blessed to have a long term mission. Like I, I've sort of found my purpose and I think the way to do it is to follow your curiosity and intuition. If it takes you university, then that's fine. If it takes you to 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, 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? And now you get the tools and the technologies, the techniques, the repositories 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 Jad Massad.

Thank you.

David Blunden, Salina Smail. 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 write a newsletter twice a week. Sending it out is a short 2 minute read via email. And if you want to discover the most important meta trends ten years before anyone else, this report's 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 dashmandis.com metatrends to gain access to the trends 10 years before anyone else. All right, now back to this episode.

ABC Tonight.

Dancing with the Stars is back with an all new celebrity cast. You have the groove.

Robert Irwin, Alex Earl, Andy Richter, Jen.

Athletes, Fleck Baron Davis, Lauren Howreggi, Whitney Levitt, Dylan Efron, Jordan Childs. Hilaria Baldwin, Scott Hoy, Elaine Hendricks, Danielle.

Fishel and Corey Feldman. This season, get ready to feel the rhythm if you got it.

Flaunted Dancing with the stars premieres live.

Tonight, 8, 7 Central on ABC and Disney, plus next day on Hulu.