Why AI Will Lead to Thousands of Billionaires and Elon Musk's

Thank you. I’m a big fan of the show and excited to be here.

I was a big troublemaker. I looked very different. So I’m like a redhead. My family is actually split right in the middle where some of my uncles are extremely dark complexions and a couple of parts of the family were extremely bright redhead. I’m actually fairly on the not too intense redhead, but we have really intense gingers in the family. But I looked fairly different and there isn’t a lot of gingers in Jordan. And at the same time, I had a bit of a temper. I had a little bit of a chip on my shoulder, rebellious angle to me.

My father, he worked in government. He didn’t make a lot of money. He was an engineer, very low level engineer in government. But he really put a lot of money behind our education. And so we went to school in a good part of town and coming from a bit of a, let’s say, a challenging part of town where you had to fight and you had to grind and hustle to just play it and be just a kid. And then when we go to school, we’re a little bit odd ones out.

And again, like I said, my appearance was a little different. And so I always got in trouble. Since I was six years old, I would get in trouble for running away from school or or inciting my whole class to do something strange strange or weird. And so I was just seen as a sort of a troublemaker in my entire childhood. Yeah. Did

Yeah. There was a lot of early successes. So my older brother and I, we found an arbitrage opportunity. The toys in the not so great part of town were actually much cheaper than the rich part of town where we went to school. And so we started selling toys at preteen, young age. My mother would come pick us up in this, like, this beat up old car, and we would stash the toys in the trunk. And so we would sell them, we would tell people what we have during school because we couldn’t carry the toys.

Then we’d tell them meet us outside after school, then we’d go to the trunk. My mom didn’t know what we were doing, so we would sell these toys. We also sold toys that weren’t necessarily the most savory, so we sold guns and we did a lot of stuff like that early on to make money. And later on when I was a teenager, also started my first software business. It just felt like I sort of had it in me. I had the hunger. I also was attracted to greatness.

Really loved watching movies and reading books about great entrepreneurs, great athletes, great history figures, and leaders. And so I always wanted to do something great.

That’s basically the biggest sign, the early sign of exceptionalism. I think it’s fairly hard for people who has taken a normal straight path to life to do something exceptional later in life. I I think there’s always signs of exceptional things. Sometimes you’d have to dig for it, but I think it’s always there, whether it’s hustling and selling and starting businesses. A incredible athletic ability and exceptional drive to achieve athletic things is also a sign. It can be very diverse, but it’s always there, I think.

Mediocrity. The idea of being normal is just very scary to me and undifferentiated is something that I really don’t like to be.

the key. If you’re really worried about mediocrity, the best way to sort of run as far as possible from it is to do something so weird and so different that is almost unthinkable. I guess in dating scenario, maybe you would not attract a vast majority of people would be actually turned off by this. But I think the people who would actually like you and would want to be with you for a long time and would actually match your sort of personality, you would attract them. I’m a big fan of being authentic and attracting allies and potentially rebelling people you don’t like anyways, repelling them.

If you read a bit of history, you find that most of history is actually built and changed by a small number of people. We’re actually living through a moment of history right now, especially given the AI revolution. You could argue that OpenAI is at the center of it, and it’s really a small group of people. There are 300 people now and they were much smaller before. They had a conspiracy and their conspiracy was that if you get a lot of talented people together, that’s what the original thing and you work in really interesting AI research, you’re gonna find something.

And then they found something and most of the world wasn’t paying attention to it, which is if you scale language models, and if you keep scaling them up and put more computing data in them, you’re gonna find that it becomes more and more intelligent to the point that it’s actually reaching some kind of general intelligence. There’s some generality in GPT based models. So that was a conspiracy. There was a small group of people with a secret, planning for a long time, and so they were conspiring.

And now they’re at a point where they’re changing the world. They obviously started this new arms race between the two of the biggest companies of the world. They birthed a new startup ecosystem. It’s changing software as we speak. I’m sure we’ll get into it later. It’s just top of mind example but you see it in history all the time. One person starts it, obviously religion is a big part of it. Jesus or Muhammad, you read their stories and they’re typically small groups of people. They fit in a room.

They have some sort of secret or some sort of message. They work really hard over many years, perhaps decades. Then eventually they succeed and they change the world. So I think startups tend to follow the same logic as these examples.

I think it’d be super intentional about creating a religion. But I think the main tenets is have somewhat of a Peter Thiel calls it a secret, but somewhat of an insight into how the world works or how the world should work or where the world is headed. And so do you work to advance that? One of our biggest insights at Replit is that software is superpower, and this superpower is actually stuck in this ivory tower of Silicon Valley Elite, and we’re trying to free this superpower and get more and more people to experience and make use of it.

It’s almost at the time like the printing press. It was the people who could read and write were the priesthood. And then at some point, everyone in the world started doing it, and that changed our world. And we think software kind of follows the same trajectory. So that’s our main insight.

Honestly, that’s not the way to do it. The way to do it is more of a selection thing. You wanna hire exceptionally entrepreneurial people. There’s no amount of processes you can put in to make someone who doesn’t want to work hard to work hard.

Yes. And it is unfortunate, I think it is the results of some decadence in the culture. There’s a feeling that life is about pleasure and paganistic pursuits. Whereas I think cultures of past, even twentieth century America was very different. The values of hard work and exceptionalism were there. It is a problem. And I think the way to find that problem partly to inspire people, inspire younger generations that actually the value of making something is very important. And I think partly we need some kind of spiritual reform because people need to understand that pleasure is such a fleeting thing and that that’s not the thing that makes life meaningful.

There’s just so much of it, consumption and just being passive. A big part of diagnosing what’s wrong with the world today is gonna be about just a lot of people are just, like, super passive, not interested in doing anything interesting.

one set of one page among a set of things that we do to filter out people. Filtering out people is the hardest thing you could do in an interview process. I’m a big believer in free markets. There are companies out there for every type of person to work at. But the type of person that we want to work at is a very specific type of person. We do a lot of work to filter where people wouldn’t be a fit out at the top of the process.

We send people a lot of material including why not to work at Replit. We have an entire blog post about that. Work here is hard. There’s a lot of uncertainty. We haven’t succeeded yet. All these things. We send them the non values page. At the end of the process, we’re typically selling people, but we’re also anti selling them at the same time and just saying, you shouldn’t join if that’s a problem or this other thing. Even our values page, we try to be somewhat provocative in the phrasing.

Yeah. And Zach said something very interesting about move fast and break things. I think a journalist was asking about it. And he said, one heuristic stick to know whether you’re saying something meaningful is that reasonable people can disagree with it. And the opposite of it is also reasonable. So move fast and break things could be like move slow and don’t break anything. And that could be the value at IBM. If something where the opposite of it doesn’t make sense, or it’s not a value anyone would hold, they’re actually not saying anything.

It cancels each other out. To be provocative, you need to say something that people will disagree with, and the opposite of is also somewhat reasonable. One of our values for example is seek pain. The idea behind seek pain is that there’s a lot of painful things in building a startup. Talking to customers is actually extremely painful. When you your product is not working or there’s some fundamental lack of product market fit or some issue you’re dealing with, founders and entrepreneurs typically don’t wanna face that. And facing that is painful.

At Replit, we’ve had some things in the past that weren’t willing to face, for example, and that created a lot of problems down the line. Most problems at companies are just delusions from not wanting to face pain. So to create a company that is more in touch with reality, you have to create a

It is a constant struggle. There’s a personal answer, perhaps there’s a team’s answer. But like a personal answer is that I’ve tried to teach myself really deeply the cliche no pain, no gain is absolutely true. Every good thing in life comes from pain, and you work out that’s pain in order to live healthier and better and perhaps have a good body. Even things like the cold shower, they feel amazing afterwards, but you have to go through a bit of pain. A lot of the things that actually feel good in life in a really fundamental way have a period of pain that precedes them.

There’s a Pavlovian sort of aspect of this. So you wanna condition yourself to really understand that pain often precedes good things. And that’s one way to go about it. I just tell my team that the longer you delay a painful sort of conclusion or decision, the more painful it becomes. You’re just building a mountain of pain that you’re gonna, at some point, have to subject yourself to.

painful because especially people who who’s been with you since the start, you kinda wanna have some loyalty towards them. And obviously, they have a loyalty towards you.

No. I’m saying that’s the natural instinct. I think it comes from this tribal way of living that is in our DNA. But in a sort of free market enterprise, it becomes deleterious and really harms you. What you wanna do is you wanna maximize the likelihood of success, maximize shareholder value. And so you’re having empathy for the individual. You’re not having empathy for the collective, for the rest of the company, for the company itself, for the customers, and for the investors. You would need to reframe that in your mind as that, yes, I’m making this bad decision in order to signal some loyalty to this individual who’s not working out, but I’m actually harming a lot more people.

From the utilitarian perspective, it’s actually a bad trade off. You have to do the right thing by the company ultimately, which is either layer them, replace them, or do something else like that.

I typically try to find places for them in the company that would be still happy. You obviously also wanna make sure that they save face and the change doesn’t look like an obvious demotion. Sometimes there are ways to find the intersection of where they would be happy and productive and where the org really needs to be. I think you can also be too hasty instead of letting that person go or replacing them, and instead, there are some ways to find a place for them in the organization that would be happy for a longer period of time.

so much credence to the misfit, and we’ve had such a good luck with the misfit. The early days of Replit, I was the only computer science grad up until ten, twelve people maybe. All the hackers of the company were dropouts and English majors and poets and all sorts of misfits that got into programming semi randomly. That was a mistake. I’m not saying that’s a mistake. I’m saying that my bias to the misfit sometimes becomes pathological. I find misfits that I know are creative and good, but they have some personality flaw.

And I try to make it work at the expense of other people on the team. It ends up wreaking some havoc sometimes. And maybe I stick with misfits too long and really try to make them work. One of my, I would say, special abilities is to be able to manage misfits. But it’s not an ability that everyone can do, and they can have a negative effect on the company even if they’re creative. The usual sort of engineering manager rather have a normal engineer who is a team player than a sort of hostile toxic engineer that’s like a 10 x engineer.

Guess I’m dealing with that right now. So I can get back to you. I would love to think you can. I think that there’s a way of doing that, but the history of companies showed Wozniak was marginalized as Apple grew and he’s like a tall misfit, still is. But he stayed there for a long while. He was working on the Apple too even when the Mac was the thing, and he continued to contribute and things like that. There’s probably some ways of making it work. However, I will submit that you can’t build a company made entirely of misfits that scale.

So you go back to the early days of computing, and the way to program a computer was to directly interact with machine code. And that’s that’s literally zeros and ones. And then we had some ways of programming computers that little easier, like punch cards. You’re still typing in machine code. That’s was a little easier. The biggest jump we had at the time, which was like probably 10 to a 100 x improvement, which was assembly code. It just created the software industry as we know it, starting with the assembly and the assembler.

And then you had the first innings of programming languages, and C was the first time you had and again, now we think of C as low level, but at the time it was the highest level thing. The sort of haughty assembly programmers would look down on C programmers, like people look down now on Python programmers, for example. But C was another 10 x jump because C now you had the concept of subroutines and loops and code you can really reason about in the same way like a human kind of thinks.

Whereas the assembly is more related to the machine. You’re increasingly going away from working on the machine, reasoning in the machine towards something that more resembles a human narrative and the way humans think about abstract objects in the world. So C I would say were another 10x improvement in productivity. And then the next big improvement was the scripting languages. So C was it was like a typed language. You had to think about the types of the values and the objects you have in the system. But then you go from C to something like Ruby, Python, JavaScript, that era that started in the late nineties was that actually we want languages that are very flexible.

You don’t have to think about types. You don’t have to design the entire program from scratch. You can just start typing and it’s almost like jazz or improv. You’re thinking as you’re typing and as you’re coding. And arguably that was another 10x improvement. But then we had a bit of a stagnant period for maybe ten-fifteen years. The next 10x improvement wasn’t a language. So up until now we had all these 10x improvements that I think is multiple orders of magnitude, let’s say 1000x improvement in productivity since then.

The next one was not a language. I think we tapped out on how much languages could actually make you more productive. I would say the next era was open source and open source packages and sharing of packages. In pre GitHub era, pre open source, every programmer in the world had to write the same functions all the time. Like, we’re all writing the same sorting functions and the same sort of stuff. And now we had this way of sharing code, not just within companies, but just globally.

And I would say that was another 10 x, perhaps more, maybe a 100 x improvement on productivity. You had a college student like, Zach build a site that reached billions of people because he used PE and MySQL and a bunch of libraries that were all written for him. He didn’t have to write all that stuff. Actually, since then, we haven’t had a big jump in productivity. My prediction is we are at the start of a new sort of s curve in productivity growth.

speculate. We actually can look at the results. Andreessen Horowitz was the head of AI at Tesla, built a self driving team there. He now, I think back at OpenAI a few months ago, he tweeted that AI is writing 80% of his code today. 80% of his code is written by a machine. That is unprecedented. On Replit, we see that 30 or 50 to 50% of code for Ghostwriter or AI, Ghostwriter users is written by the AI. Ghostwriter users report that tasks sometimes are cut in half in terms of how much time they need it.

There was a study actually done on Copilot users, GitHub Copilot, that showed that programmers are 55% more productive. Okay. This is not a 10x, but it is a start of something.

There are two ways in which large language models are gonna change software. So one is the AI assisted development. Copilot is leading in the space. I think we’re like a fast follower. Perhaps we’re gonna exceed them in the future, but we are Replit Ghostwriter is another sort of AI assisted development. Actually, by the time the episode comes out, we would have come out with Ghostwriter Chat. Imagine ChatGPT embedded in your code, knows everything about your code, can answer questions about your code, really starts to feel like another pair programmer human that’s sitting there and helping you.

That’s one track of ways that it’s gonna help us become better software developers. But there’s another way which large language models can help us make better software and make more software. They are in themselves software. Every large model is a function, essentially. You give it input, it takes output. And that function is programmable in English, what we call prompting. And you can make it do whatever you want it to do. Crawling a website, scraping a website, and getting useful information out of that. We have a lot of libraries and ways of doing it, but it’s still a difficult task.

It’s at least a many hours task, sometimes like many weeks. G p t three can extract any type of information even in a structured fashion from any website. That’s how Bing searches in. That’s how Neva, Perplexity, all the new search startups are done. So instead of spending weeks writing a parser, you can just plug in a large language model that’s trained to parse or that’s prompted to parse, and you just saved hours of programming. So not only is it helping us program, it is actually replacing pieces of software, and it’s gonna become a core piece of software in the, like, function call graph of software.

level means that you focus on the customer. It’s crazy to think, but most of programming today is really still trying to appease the machine. You’re dealing with the issues with the libraries. You’re writing library code. You’re writing meta code in order to manage your workflow. But the actual times and on business logic is probably like maybe 20% of your day to day coding, which is absurd. So you take that 20%, and that becomes 100%, if maybe 95%. Does it slim down your engineering team significantly? Yes.

If you Yes. But that doesn’t mean software engineers go out of work. It just means that smaller software engineering companies will actually be able to have a lot more impact. But I think you would need way less engineers. You asked me about a ten year prediction. I think one person would be worth a 100 engineers. One good engineer in ten years would be as productive as a 100 engineers today.

I think that’s an AGI complete problem. What AGI complete is basically you would need an AGI in order to do that. Your prediction might be that AGI will be here in ten years. Sam Altman in OpenAI thinks that. But then after that, the world changes in a fundamental way. So it’s called the singularity point because you can’t really predict what happens after that.

But ultimately, if there’s no AGI, and I think it’s still low chance that we’re gonna get AGI in ten years, then logic is fundamentally creative activity and fundamentally novel activity that it would be hard for him to entirely do that because it would go out and have to understand customer needs and perspectives and have to have some way of having good judgment about the world. It is like all these fuzzy things that we don’t know how to program today.

I think there are developers today that are commanding insane amount of salaries while actually working very little. It goes to what we talked about at the value of hard work and all that. The idea that you can go coast at big tech and still be a millionaire, I think it’s absurd, and I think that will go away. I think that’s just the effect of an inefficient market. Right? And perhaps a zero interest rate market. But I think that was an unnatural thing that happened. In that regard, yes, maybe the average developer salary would go down.

But I think the most creative, most hardworking developers will actually be able to earn a lot more. So the market value

Because their leverage will be huge. John Carmack, one of the best software engineers in the world, he made Doom, he worked on VR, and now he’s working on AI. He’s still limited by how fast he types, which is like a crazy thought. Right? If he’s not because ultimately, you have to type the code out. Right? If the AI is actually better at typing the code out and he’s actually manipulating things at a higher level of abstraction, he’s talking to not just one or two, maybe 1,000 AIs doing work for him, he’s gonna have an insanely higher leverage on the world.

And so I think the best, most creative people will have a lot of the menial work automated for them. They wouldn’t have to type a unit test anymore. The AI would do that, for example. And so a lot of their time is spent just doing more things. And think about Elon. Right? Elon is starting companies left and right. That’s probably not the end of it. He’ll probably start more companies in the future. Part of the reason is because he’s operating at a much higher level because he’s built a set of very competent people around him.

They follow him from one company to another, whether it’s like on the finance and legal side or it’s on these sort of engineering side. A lot of that could be AI and could be automatable. So we can see more Elon Musk type people that are engineers because once you abstract the kind of low level work, you can do more and more things. You can start more things and do more things.

Wealth and quality will grow a lot. How so? Because people look at Instagram or WhatsApp as these outlier companies that were able to achieve massive valuations with very little people. And I don’t think it’s gonna be an outlier in the future. I think we’re gonna see a lot of companies, five, ten people, those are gonna have massive impact, probably be valued in the billions of dollars. It’s gonna create a new crop of millionaires and billionaires in Silicon Valley and elsewhere, and these people are gonna have more and more power in the world.

And I think that will create more envy. I think humans will just continue to get richer, but I think wealth inequality is concerning not because it means that some people are getting the short end of it. Sometimes that happens, but I don’t think it happens in technology. I think what happens just creates more envy. Because people think they look at rich people and they think, oh, you must have robbed someone to get there. That’s how politicians gonna frame it as well. And when you hear Elizabeth Warren talk about rich people, they

will get more and more hated, and I think it’s gonna create more division that way. I think politicians will jump on that and will try to gain from that. Maybe it will lead to violence. Obviously, I hope not, but maybe what you’re saying is storm the gate could actually literally quite happen, people on the streets all of that. So I think wealth inequality is something to worry about. That’s for sure.

I actually think that chat GPT probably added multiple IQ points to a lot of people. So I actually think that AI is equalizer in terms of intelligence. You have people who don’t have access to the same education, and now you can have an AI tutor that is everywhere that works the same for everyone. So I actually think that the knowledge gap will shrink, and the education gap will shrink, and even the intelligence gap will shrink. So all those are good things. But nonetheless, I think the outcomes will be radically different.

That’s just because it’s a free system and free systems tend to follow power law. Like, even in nature, anytime you see free unrestricted system, you basically have this problem where the winner takes all. Right? As as a venture capitalist, like, the power law is almost like a law of nature.

I think that’s absolutely the right thing to do for Zuck. I think the metaverse bet is not fundamentally misguided. I think it’s pretty clear to me that at some point we’re gonna have photorealistic technology, whatever it is, whether it’s AR, VR, or what have you. It’d be great if you and I were actually sitting in the same room or seemingly sitting in the same room. That’s obviously true. I think they underestimated maybe how much time it’s gonna take, and they probably overestimated how much free cash in the world there’s gonna be to continue to invest.

On the other hand, AI absolutely clearly has traction and absolutely clearly has happened. It would make total rational sense to kinda shift some resources to invest more in AI.

Look. I moved to The US because I like sort of Western values of freedom and democracy and individuals. The Chinese Communist Party embodies none of that. It’s pretty obvious they say they don’t like individualism. They’re clearly at odds with our values. But part of our values is free market and free competition. And so shutting them down just because we don’t agree with them is also bad and corrosive to future investors and future sort of countries and other players all over the world that want to invest in The US.

We need evidence that they’re actually doing something bad. But if we don’t have that evidence, I would think it’s government overreach to go and try to ban them.

Do we have actual evidence that data on US citizens is actually, like, making it some way to the CCP?

Yeah.

a little conflicted on that because the Twitter files just showed that there’s incredible US government collusion with US social media companies as well.

I see what you mean certainly. But don’t you think that we need to at least give some weight to the idea that we want our markets open to foreigners to compete in?

Yeah. I would agree with that, but there’s a lot of room. If you were to make that law, how would you actually make it law? What kind of

Right. But just single out China. What about Singapore? So Singapore is very much a top down and somewhat authoritarian, but it actually they’re good ally of The US. People like to live there as well. Do you ban Singapore Singaporean apps? So it’s defining those values clearly to make it into law. I think it’s gonna be tricky.

I think there’s gonna be increasing pressure on them to move more of their operations to The US. Obviously, use Oracle data centers now. Maybe there’s gonna be more pressure to spring up more firewalls there. I think maybe there’s more pressure for more US based companies. Maybe investors in The US have to own more of the company than Bydance. I think things like that will happen just with the increased pressure. And I think TikTok will have to become more transparent. Because right now, they’re not very transparent.

They use every growth hack. There’s a very slimy app in that way, but I think they’ll just have to react to the sort of public pressure on them. And I think they might be able to do a good job and save it off and give the US government, US people some assurances that they’re not collecting data and sharing with the CCP. Banning them outright, I don’t see that happening unless there’s a full republican government and political will.

optimistic for? Equality of access. The ability for anyone in the world to participate in the software economy is just gonna be huge. How did having children change how you think and operate? You become a lot more future oriented. You actually worry a lot more about the world.

Yes. Why? There’s something in the air. I don’t know what it is, but there’s this optimism, this feeling. I don’t know what it is, but it is intangible. Some kind of community is important. I think New York has a lot more of it today. I think you can find it somewhere else, but community is important. What’s the one word that will be on your tombstone? So mine is persistent. Win. And I don’t mean that that I’m, like, consistently winning, but I consistently try to win. And I think it’s a good value I have to try to win.

So win at a sort of a command.

I would want to say a company that’s so different just to experience something so radically different, like a freaking insurance company or something like that. Just to be able to know what happens in parts of industry that I know nothing about. Something super boring like an insurance company.

I probably said a lot of things on this podcast that many will disagree with. Maybe the spiritual malaise, but that’s more in the today. I think Bitcoin is the future reserve currency of the world. Okay. Unpack that. I think if you play out the game theory of how played out reserve currencies have played out in the world, now a lot of people know that the problem with fiat currencies and just a problem with creating money out of thin air, there’s going to be some kind of a return to a commodity pegged money.

You could return to gold, but gold has all these problems with it, like not being able to move it easily. I think that a rational long term thing that will happen, and it might happen in our lives, but ultimately, humanity will discover that there’s only one commodity with all the values we’d want for money to create on top of it a lot of different national currencies, and this can be Bitcoin.

health and maybe friendship a lot. I think I was so driven that I’m paying down a lot of sort of health debts and maybe reconnecting with friends now and things like that. I’m not gonna word totally healthy, but there’s some of just getting fit and getting back into a good life habits and all of that has been a lot of work. And so I’ve also neglected some relationships and I don’t know. It’s almost like you heard the tweet recently by mister wonderful? What’s his name? Yeah.

No. Know

What he said about wife and dog? Yeah. So was like, work, be so dedicated to success that you could lose your wife and mother and dog and something like that. And then after you achieve freedom, you can do whatever you want. There’s some way to steel man his tweet. People got really angry at it. Even people in Silicon Valley and entrepreneurs. But there’s a way to steel man that is that, like, you can actually take tremendous in the same way you can take debt, tech debt.

You can take tremendous amount of life debt in order to be successful, and I’ve certainly took that. Now the question is, like, could I have not taken that? Obviously, nothing irreversible. I wish I’d spend a little more time with my mom before she passed away. Maybe I would tell myself to be like, take life a little bit more seriously.

There’s all sorts of ways that you could answer that. Obviously, financial company, all of that, but I I would focus on impact. Want, like, a kid in rural Jordan to be able to have started a startup or successful business in a rep. Want people anywhere in the world to be able to come to Replit, teach themselves how to code, go do a few bounties and generate some income, and then start a startup and put it up in our app store community and start generating income from that, maybe start growing it.

I really want that opportunity to be available for in the world. I want Replit to be a full stack service from your first line of code to your first app to your first company to your first dollar.