#171 Ditching a Microsoft Job to Enter Startup Hell with Lonewolf Engineer Sam Crombie
01:55:27 Share

Welcome back to the Free Code Camp podcast. I'm Quincy Larson, teacher and founder of freecodecamp.org. And today we're talking with Sam Crombie. He is a software engineer and prolific contributor to the Free Code Camp open source project. And he left his job at Microsoft, got into Y Combinator, the San Francisco startup incubator. And he is kind of currently in startup pivot hell, trying to figure out how to effectively deploy the half million dollars he's raised.

Sam, how's everything going? It's great. Thank you for having me. I was reflecting on... New York City? Yes, that's right. Okay, awesome. And I can hear like a little bit of construction outside. If anybody's wondering what that noise is, it's just a bustling, busy city, right? It's loud. Everything's happening.

Yeah. So what were you going to say before I rudely put it in? Well, I was just speaking of New York. I think it was probably a little under two years ago that we did our first podcast in Brooklyn at the Brooklyn Library. I don't know if you remember that.

Oh, yeah. I can't forget. Like, the Brooklyn Bridge is an engineering marvel to this day, like 100 years later, maybe 120 years later. I don't know when it was built, but the thing is massive, and there's this entire kind of, like, enclave underneath it. I can't remember what it's called. There's, like, some acronym, but underneath the Brooklyn Bridge, and there's, like, you know, food trucks, and it's a cool area to walk around. And the Brooklyn... Yeah. We didn't actually end up publishing that episode. We did not. Because...

You were like, I'm not sure what I'm doing in life right now. And here we are. This is a good time capsule. And now we are, you know, about a year and a half later and stuff has changed quite a bit for you. Yeah. Let's dive into that. But first, before we dive into that, I want to ask you point blank. You work in AI coding tools. Can you separate some of the hype from the actual capabilities of these tools? It's a very big question.

Some people talk about how in the next like two to five years, software engineers like as a profession is just going to be eliminated. And I read a lot from like college students who are majoring in NCS like that are very worried about them not having a job like very soon. And you hear about companies like, oh, you're not hiring new grads, you're not hiring interns, like what's going on?

I guess personally working on a startup and what I saw working at Microsoft is AI Cogen is really useful for, I guess, two things. One is prototyping. And that's if you're a designer or product manager or I guess a software engineer that wants to spin something up. If you have like some scoped idea that you want to bring to life, I think the prototyping tools like Bolt and Lovable and V0 just make it a lot easier to like bring that to fruition.

So I don't think that's going to replace any jobs, but I think it just lets people do more with the ideas they have in their head. I think the second part is...

The CodeGen I think most like software engineers use with, you know, your cursors and your windsurfs, like these IDs, where if you're comfortable, familiar with like a language, you know, like a set of like frameworks or technologies, it just makes you so much more productive. It helps you remember the things that maybe you forgot. It's like your reference manual. You don't need that. It's just like autocomplete. And whatever I guess your project is that you want to build, it just lets you go end to end a lot faster. Yeah.

And so I think, like, it's hard to say, you know, do you replace, like, the craft of software engineering and, like, engineering these, like, very complicated systems? Probably not. But I think someone who uses these tools will probably replace you if they can just work, like, 10x more productive than you can. So, you know, there's certainly a lot of hype around, like,

Can people like, is the job going to be replaced? I'm not sure, but I do know that like the productivity gains people have from these tools is incredible. And personally, it makes me feel like I can do so much more as a software engineer. It makes me a lot more confident in my skills. Yeah, 100%. Yeah.

So CodeGen is useful for rapid prototyping, which again, it probably is maybe 1% of the job. Maybe 2% of the job is like building a quick and dirty prototype so you can like kind of start a discussion and a series of meetings talking about feature sets and just see whether something is even possible. But then the second part you said is just

Tab completion, essentially. Tab completion. Yeah. And that's speeding up somebody like – I mean you said like somebody who can work 10x faster than you. Do you think it's possible with the help of AI that a seasoned engineer who knows what they're doing is already familiar with the tools and just has to frequently look things up because –

Frankly, these tools are too complicated for most people to just hold in their head unless they've just spent a tremendous amount of time. Like, I can't remember what this, you know, this function name is, or I can't remember what the name of that library that does this is, you know, and they're constantly having to tab over to Google and like search for stuff and go on GitHub and like find out the name and then like install packages or add them to their, their Indian file or something like that. What do you think in terms of,

Those types of tasks, like, is it really a 10X improvement? I think if you look at the software development lifecycle from end to end, it's like each of those pieces you're getting significant gains on, and that's where you get the 10X. Like scaffolding, like the system engineering architecting, like how am I going to build this?

Having that thought partner, comparable to a senior staff engineer, being able to collaborate with that AI, I think means whatever you're going to start building, it's going to be better. Actually scaffolding out the code, filling in those pieces, like

Like writing unit tests, like getting dummy data and then going and testing with that. Super helpful. Like no one wants to test. Like that's one of the worst parts of the job is doing that type of testing. Setting up certain like development pipelines also like kind of a major pain in the ass.

And it just makes that much easier. And so you get all the way to like, I've, I've made some, I've written some code, I've made some commit. And then you have these code reviewers that again, act as like these teammates where, you know, they'll look at your PR, they'll leave comments, they'll remind you about like style and, and different things.

And so I just think like if you anthropomorphize it as like an employee or like a coworker, you know, you think like your best coworker that gets you really excited. They know exactly what's going on with the code you're writing. Any question that you could ask, like they're going to answer it. Like how does that improve your productivity? And now imagine that like they're available all the time. I find it for each of those pieces really useful. And to your point about like understanding,

understanding and gripping open source code bases, um, looking through your existing code base super quick, like chatting over it. It just, it just seems like every piece of it, it does such a good job. Yeah. That's my experience. Yeah. That's awesome. So this is for like smaller projects that you're, you're starting up. I mean, when you talk about getting, getting various like pipelines in place and, uh,

choosing libraries but like let's say you're working with a larger legacy code base like google for example the code base is famously like two billion lines it's all in it's got its own version control system and everything like that like let's say that the the context window would literally be i don't know hundreds of billions of tokens or maybe tens of billions of tokens right um and and

there's a lot of opportunity for an AI to get turned around in there where a human engineer would have more discerning judgment probably. For sure. Obviously, that's an extreme example. But let's say a more typical gnarly legacy code base that is, I don't know, maybe like 10 million lines of code or something like that. Like when you were working at Microsoft. I don't know how much we can talk about what the code base size is, what products and stuff where you were working.

working with? So when I was at Microsoft, I was a product manager on the Microsoft edge web browser, um, chromium notoriously, very big code base, um, you know, multiple millions of lines of code. I remember working on projects where there would be a search question of like, do we have UI for this feature? Like, does this exist in the code base? And it was a non-trivial task to answer that question, like search over the chromium code base. Um,

And code bases on that side, I haven't had direct experience. How big can you build your project until AI starts to misunderstand it? I actually think relatively quickly. If you have a framework that you use, it'll help you get set up with that framework. Maybe if it's React or something, you have a few different folders, a few different components. It keeps in memory the context of, I helped you build these pieces. Here's how they interact.

The strategy I've seen is that, and this is what I think like some of the like AI software engineer and, and code agents try to do is they go through and like index your code base. And then they just keep like a, a log, like a, like a markdown document of how do these different pieces interact? Like what are the dependencies? And you try to like,

build that out and like maintain it as you get larger so the first thing that ai does is it looks at that document as opposed to looking at just your file and helps like keep it in memory better um i guess with the prototyping tools it's hard to build a big code base um so i haven't had that experience yet but it's a hard problem i think well let's let's say so just to be clear is the ai creating its own reference or are engineers creating the reference to guide the ai

I think I've seen both. I mean, like I've, I've, what I've done before is written out like a technical spec and, and then had an AI take that and go implement it. And that's very useful because it just tells it everything that you want. It also makes you think through like, how do you want to actually get this built? I think one of the ways people go wrong with the prototyping tools, and it's just like a side effect of how people go about it is they

you're kind of riffing when you're building it. It's like, oh, you could build this and you could build that. And it could be cool if you could do that. You don't go into it like knowing exactly what you want. So I've seen people, they do these large written documents. It's like 30 minutes writing out a prompt. And then it's like, I will manage the AI, like an agent, like agent and cursor or windsurf to go do it.

But then conversely, I've, you know, used like some command line, like Claude code, code buff tools like that, where you will ask it to like, Hey, can you summarize like what you've been building or it'll do it itself. And it keeps a document that's like, here are the features we've implemented here, the dependencies. And it seems to get rid of the hallucination for a time. Interesting. So it's kind of like humans. Like we just have like a captain's log, so to speak, like some, some sort of document that like,

just tracks what has been happening and what the priorities are and, and,

Then they might go back and reference that. Okay. Like, like why did we end up implementing this feature? And you know, humans, it's probably in this giant pile, like free code camp has more than 50,000, you know, pull requests issues on getting right. That's a lot of, that's a lot, you know, previous decisions and deliberations to wade through. If you're trying to understand, okay, why did we like the other day, it was like, we used to have this button that would just take you right back to where you were in the free code curriculum. I was like,

wait, when did that go away and why? And then we had a discussion and it was in the tacit knowledge. Somebody knew because they'd done it. Oh, we had like an API that was getting like hit way too many times and it was overwhelming the API. So we temporarily disabled the feature and we just forgot to add it back in. So I think that that should be back in if anybody's been missing that button.

Sorry, that was meant to be re-implemented after we've fixed the AI. But that is the kind of thing that having the tacit knowledge of people in the organization who have been there for a long time, you kind of pick that up. And an AI doesn't necessarily know the context behind that, especially if that wasn't debated in a GitHub issue and it was debated in Google Chat, for example, where we have a lot of our deliberation. We don't use Slack. We use Google. Their entire suite is free for charities. Shout out to Google for doing that. Yeah.

But yeah, like I can easily see where having that sort of reference would be huge for a human that's trying to manage a product.

And similarly for an AI, if you presume that AI is endowed with human-like reasoning, even though it's really just a token predictor, you can get pretty powerful emergent properties from something as simple as that, right? For sure. And I've seen – I think we were talking about this or talking to someone about –

everything that AI code gen is good at, like what are those next primitives that you need? And one of the things is like bigger code bases. How do you understand those code bases? But then another one is, I think we were talking about this. If you like have AI generate code and then the AI looks at it a week later, or you look at it a week later and,

Having a trace of, like, why did I do this? Not just, like, what is this code base and, like, the objective description of it, but, like, why did we implement it this way? I think that's actually really missing from a lot of these code tools. And it's important, yeah. It's, like, the implementation decision. And it's that tacit knowledge that you can, ideally, with any new code you generate, I think retroactively it's probably really hard to, like,

I don't know, maybe you give the senior engineers at your company like 30 minutes with an AI and it just like looks at the code base and then it like interrogates you and then you can collect that. But at least for any like new code generated, I think it's going to be super important to have that. And you can just imagine like once you have these super large code bases where you just know every piece in and out, it's a low latency. Any change you want to make, you have the full context there.

I don't know if you think like you're productive now with AI, just imagine in a few years when it's going to be like, I think that's crazy. I mean, and especially if people are like not necessarily building giant monoliths, but they do have like these more, I guess, intelligible microservices. Like I do think there's going to be a boom in microservice architecture just because it's going to be way easier to leverage microservices.

AI coding tools if you have these small kind of like modular code bases that the AI can easily rock and then go through and build out features for and stuff like that. For sure. Do you think that's going to happen?

I mean, I feel like that makes sense. It's like not, I don't know, some people talk about context window size, and it's like you can now fit this code base of this size into this context window. But I don't think just because we can fit more and it can remember more means that we abandon organizing things and decomposing them into more understandable units. So if anything, that's more important.

And just to make sure, like when you increase the context window size, you're saying that the AI can bring more in, but it doesn't necessarily mean that there's not diminishing returns to dumping more and more data into the context window, right? Like I would imagine the more it's working with, the more room it has to get confused and start hallucinating.

For sure. And I think, I mean, speed, cost, you know, that's the trade-off. So, like, there's the needle in the haystack benchmark. Maybe it's not, like, a benchmark anymore because it's solved. But I remember when Google came out with, I think, like, the first Gemini that had, like, some million context or mid-million token context window. Yeah.

there was like an evaluation of this small detail and this huge corpus and that you put into the context window. Like, can I remember that? And it's like getting better at yes. Like it can find that very tiny detail, but at what cost, like all of this compute and then like taking 20 seconds, if it's just right there, it's like 50 tokens, it's better search. And yeah, that one, that one ever go away.

So, so essentially just as the, the potential space in which the needle could exist, you said, you said, do you think it's solved? Like, do you have any insight into like how it was solved? I don't honestly. Um,

No idea, but they seem to have figured it out. Yeah, I've heard talk of 10 million token context window. There was a paper on infinite context. Theoretically, it could just go on forever. I don't know. It's like magic.

I just benefit from it. I don't do it. Well, a lot of the hard engineering is done by scientists and researchers, and then it's just picked up by engineers like us, and then we just kind of like –

Bunch on it and figure out ways to apply it. And a lot of engineering is really just taking things that have been figured out, discovered. Like I consider LLMs and the emergent properties of them and their efficacy to be more of a discovery than an invention. It was discovered that there was these emerging phenomenons. Agree or disagree? Yeah.

Agree. Well said. Okay, cool. So we're going to dig a lot deeper into Cogen, how you can applicably use it. We're going to hear lots of tips from that. We're also going to learn a lot about your developer origin story, Sam, because it's a really cool one. First, I want to say support for the Free Code Camp podcast comes from the 11,384 kind folks who support our charity through a monthly donation. You can join these chill human beings today.

and aid us in our mission by going to donate.freeco camp.org. Support also comes from a grant from WIC studio. WIC studio provides developers tools to rapidly build websites with everything out of the box, then extend, replace, and break boundaries with code. Learn more at wicstudio.com. Sam, you grew up as a public school kid in Florida. Take us through your early years. Yeah. Um,

So I guess like, I don't know, I grew up doing a lot of sports. My parents always kept me busy with like clubs, activities. I was usually at school early and leaving school late. You know, played a lot of video games, spent a lot of time either outside or then like inside. Very political growing up, like,

grew up in a political household, um, got very accustomed to like politics and, you know, having a new station on. Um, so I don't know. I feel like my first, like pre high school, like 14, 15 years, I just did a lot of different things and it was just like a way to keep myself like busy and engaged with, you know, off the streets, I guess. Um,

So I feel like my first – and over those years, like my parents like got us to use computers from a very young age. I was actually just reflecting with someone the other day. Like my school district had us doing typing lessons very early on. They had us doing stuff like Scratch pretty early on. Yeah. And even though like –

That's right. Which, on reflection, I don't know. I don't know if that was the best intro. After using Scratch, I was like, I don't know if I like this programming stuff, but that's another story we can talk about. But yeah, just different ways I interacted with computers and through video games, I always was using a computer. But I never... I knew about programming and coding, but I never...

attempted to do it. I never like found anything. I just like was aware of its existence. I had one friend who had a dad who was a software engineer, but otherwise it was like this very like unknown thing. Like I don't, I didn't know anyone else. No one around me was in tech. Um, so I was, I was aware of it growing up and then it wasn't until like starting in high school and then obviously in college where I really dove into it. Um, but up until, you know, the first day of high school, it was like,

I don't know, just a normal suburban upbringing, you know, nothing too interesting. Yeah. Yeah. And you got into Dartmouth, which is one of the best schools in the world. It's one of the Ivy League schools. And you studied CS there? I studied, yes, that's right. Yeah, computer science.

How do you go about getting into such a prestigious school? And you're not like my understanding is your family didn't have just tons of money. It wasn't like you were like, you know, related to somebody up in the admissions board or that you had like donated a ton of money to the school or anything. You had to you had to go in through the front door essentially by getting like really good standardized test scores and stuff like that.

Yeah. It's, it's interesting. Like I'm trying to just remember all of the advice you hear when you're going through the college application process. Like you want to be well-rounded, but also you want to have a spike. You want to do SATs, but then also do ACTs. There's a lot of like tactical. I,

I did just SAT. I did not do ACT. Although I had friends who like would do SAT for years. And then like the month before the deadline, they would do ACT and they just get like a perfect score. Like they ended up doing much better on it. So there's like, there's like, you could spend hours talking about like your academic advice and the APs and then, you know, your, your clubs that you do and your extracurriculars and the way you frame it and your personal statement. Like there's a whole bunch of stuff.

I think for me, my thing in high school was government and politics. I did speech and debate. We had a program that kind of competed locally, but we got a new teacher, a new coach, my freshman, like second semester, end of my freshman year. And a few of us were like, well, we want to go compete at the state level. So he took us to the state level.

And then we're like, well, why don't we compete nationally? Let's go to Harvard and Yale and these different tournaments. And he just would take us. That's great. So he was just like a huge advocate for your ambition. Whatever you wanted to do, he's like, all right, I'm in your corner. I'm going to figure out a way to make it happen. And debate is like one of those really wonky things. It's not necessarily about actual political opinions or positions or anything like that for anybody who's unfamiliar with this. It's just about argument, essentially. That's right. Logic, reasoning, logic.

Uh, rhetoric, persuasion, all, all these things that people studied back before they had computers. Totally like speech and debate and they, and the United States, um, like before I did debate, I was very ignorant and dogmatic and I had my opinions, but I didn't know much about them. Um, and the event I did, which was like congressional debate, cause there's like different events and there's different ways you can do it.

Um, you're the way that you win is you, it's not just like taking a certain side and being like polarizing about it. It's you want to get in as much speaking time and ask really good questions. And so as a kind of function of wanting to, or a function of like doing well and like scoring well, you have to take sides. You wouldn't, or speak on bills that maybe you don't care about for topics. So, you know, when you're forced to like, um,

pathologically take an argument and advocate for it and do so confidently, it really gives you more perspective on things. And it could be an argument you...

completely abhor, but your job is to make the best case argument for it and to be convincing, essentially. And kind of like my experience with debate is very limited. I watched some debates and I reported on it when I was writing for a newspaper. And essentially it's kind of like landing blows. It's like almost.

Like you make this point in this point in this point and you talk really quickly and your opponent doesn't have time to refute all those or they're unprepared to be able to refute them and things like that. And then the judges are like, okay, this person was more convincing. They win. Yeah. I mean, the convincing part, like you said, is like very core to it. Um,

And so, yeah, like, how do you be convincing? Well, you want to be well-researched. You want to, you know, not just speak on things, yeah, that you agree with. So that was, like, super formative. And what you said, like, our coach was such a champion. Like, he was fresh out of college. He came back because he did debate when he was in school, at our high school. We all knew him. It was like a family friend's older brother. Yeah.

And he just like was such a huge advocate for us and would take us wherever we wanted to go. And that was so transformative, not just from like a, you know, doing an activity and like getting good at it and like building confidence in yourself and all like the lessons you learn from doing debate. But I think also meeting people outside of our bubble, like where we grew up at these different schools, like competing at these different tournaments, it was like a very life-changing experience. So I did debate,

Sorry. Keep going. Yeah. So debate was like your extracurricular. So a lot of these big prestigious universities, they want to see perfect SAT scores, of course. But a lot of people with perfect SAT scores may not get into Dartmouth. But they also want to see tons of extracurricular activities that are like, oh, I did fencing or I did debate or I did chess club or I built a startup or something like that, right? Yeah.

Yeah. Back when I was applying, it was like, can you stack up as many extracurriculars as possible, but stack them up in some niche? So it's not just a breadth thing. It's a depth thing. But also, do one or two so you're a little well-rounded, but have your spike. That unintentionally ended up being student government politics. I don't know if I ever told you this, but this was kind of part of my application, but

my freshman year, that was the 2016 election cycle in the U S. Um, and I was getting into like videography and like interviewing people. And so I ended up going to a few political rallies, trying to interview candidates. And in March, 2016, I actually ended up, I was like 15. I ended up going to a Donald Trump rally in Boca Raton, Florida. Cause I, you know, I was interviewing everyone. It was, you know, um,

nonpartisan. And I ended up getting, there's, there's my video still on my YouTube channel somewhere. I think it's unlisted. Um, there's a video of me like interviewing Donald Trump at the end of this, um, at the end of this like super long event that he had. And that was just kind of like part of the pie of like, yeah. And,

And politics and this is an apolitical podcast. We do not have any political affiliation. We are no political charity and we are not advocating for anything going on. We're just but it is substantial that you actually got to talk with the current president. Yes. And other politicians in the area. Yeah.

So that all kind of levered up to like what I applied to Dartmouth with. Although I've heard nowadays, it's like, that's not enough. Like it can't be doing well or like whatever, like having an interest in a school regulated activity. You have to like go out and do it on your own. Like that's the super interesting thing. They're looking for people that are not merely like conforming to like the checklist approach and reading the books about how to get into Dartmouth. They're actually going out and kind of like charting their own course. I think so.

Yeah. So what advice would you give to somebody who's in high school right now or parents of kids who are in high school who would like to attend some of these prestigious institutions for how to get in? If you had to distill it to like just a good question, bullet list, bullet point list based on what you experienced and your peers whom you talked with. I think one, go visit the colleges if you can or spend as much time as you can with

people who went there, current students, stuff changes so quickly and like make, like it can make or break your four years. Like it doesn't matter if it's a good school, however you rank that, if like you're going to be miserable. So I think one, like get to know your college. I think I lucked out with Dartmouth. I ended up really liking the people, but I didn't have like that same level of exposure. I mean, two, I feel like this is kind of cliche, but like do what you're interested in and like,

you know, like if there's something that you really enjoy doing, then just do that like a thousand percent, like forget fractionalizing it and like trying to get your, you know, your well-roundedness. Like if you really are into computers or biology or I don't know what, like a sport, just like be world-class at that one thing. I think I also didn't do that well in high school. It was like

I enjoy debate a lot, and I wish I had gone a lot deeper into it instead of spending time on the BS college application games. And then also, three, it doesn't matter. Wherever you go, I guess it's easy to say, but...

It's one part of your life. Yeah. Yeah. And what I understand is not necessarily the knowledge that you gain from it. It's the caliber of the peers that you're next to in class that you're potentially staying in a dormitory with, that you're building these relationships with and access to that network of like ambitious people who also worked hard enough to clear the hurdle to get into that prestigious institution.

Right. So that's accurate. Like, okay. And, and the education, of course, a lot of courses, a lot of universities famously put their coursework up on the internet. Like you can watch Harvard CS 50, you know,

Definitely the best production of a computer science course ever. And David Mellon is great, and I've interviewed him on the podcast as well. And you can watch the entirety of Harvard CS50 on Free Code Camp's YouTube channel or on the CS50 channel. So you can kind of get a lot of the actual knowledge, but to actually attend a prestigious institution, you're getting that peer group.

For sure. It's yeah. Like you said, it's you're surrounded by people ideally that you like to be around and that you enjoy working with. Like they give you energy. You enjoy their presence. They push you along certain dimensions and you push them along certain dimensions and

They have some like diversity of background that you wouldn't otherwise be exposed to. It's just like a group of very interesting people that will make your life better and you can make their lives better. Yeah. I feel like that's my biggest takeaway. Were you able to get any financial support? Yes. Yeah. I mean, Dartmouth is like notoriously, I think pretty generous. They've always kind of like led the pack and like,

If you make under a certain income, then like it's, it's a full ride for international students to actually there. They were need blind before many other schools. So fortunately. Okay. Awesome. Congratulations. So you're at Dartmouth, you're studying CS. You're getting to live that you lived on campus.

I live on campus. Yeah, so you get the full college experience and come time to start applying for internships, which is very important. Young life is just clearing hurdles, right? You've got to do well in high school so you can get into college. Then you've got to do the internship game, and then you've got to do the job application game. And finally, once you have some experience, then it's a little easier to get a job. But those first few years are just run, run, run, especially if you're in a place like China or India where it's super-duper competitive. It makes America look like

like child's play in terms of like what you have to achieve to get into an elite institution uh but i'm not trying to denigrate the accomplishment of people who get into elite institutions here but just wait like read about the gaokao in china read about the insane pressure people studying six to seven days a week really for three years for a standardized exam uh that determines whether they go to college or whether they work as a subsistence farmer so um

What did you learn during that period that stood out to you most and prepared you best for the internship search? So I had an interesting few years leading up to my first tech, like big tech internship that turned into Microsoft. Because I started as econ, actually. Like I applied to be an economics major. And I did...

my like finance internships. There's like a certain way if you want to do investment banking or private equity or like consulting, there's certain internships you do. And I was doing those also while COVID happened.

So it was like my spring freshman year, my spring semester freshman year. And I was just like doing a bunch of random stuff. A little bit after that, I actually did like a law internship, which I had done also in high school. And so I was kind of all over the place. A law internship? Were you considering going to law school? I wasn't. It was like, it was kind of like a tech law, like bolt on. Yeah, I don't, maybe at the time I was like more open minded to it, but yeah.

It was just like some, it was actually my, my interviewer for Dartmouth, like my alumni interviewer. She was very nice and very smart. And she gave me like a mini internship when I was still in college or when I was still in high school before I found out about Dartmouth. And so it was like the middle of COVID and we were talking about something we had spoken about when I was in high school. And so she runs this law firm, her law firm. And so I went back and interned with her.

So I was kind of all over the place. Maybe I was thinking about law school. But that is kind of good, though. I mean, people should have varying interests and there should be some deliberation. I meet people all the time. They're like, my parents pushed me to go into medicine. So here I am. I'm in medicine. And I hate it. I hate being a doctor. I talk to people like that all the time. And they're like already like 28, 30. And that's all they've done is prepare to be a doctor. And they don't like it.

Yeah.

that can support like basically an entire village back in India. But they, they feel like themselves, they're like suffering to try to help their, their family and stuff like that. And they're, they're going to have their kids go and be software engineers or something that is, I guess, more fun and creative and not just like the textbook, like who makes a lot of money, who is, you know, valued in society. I mean, doctors are like basically the best thing you could do if you just want to be respected. Yeah.

Like doctors get pulled over by cops and they're like, oh, you're a doctor? Don't worry about it. I hear the stories from the doctors all the time. They don't get tickets because police officers don't want to be on the table and the doctor is like, oh, man, this guy gave me a ticket the other day. Yeah.

Anyway, I don't know how that is, but I've heard a lot of stories. Yeah. So if you want to just be respected, go into medicine. If you want to help people, go into medicine. But if you want to have an actual fun, fulfilling career, consider software engineering is what I always tell people. Yeah. You're doing all these internships, and then at what point do you actually look at software companies, start applying to those, and what's that process look like?

So probably between sophomore and junior year, I had worked for a few startups, like in between like breaks and stuff like informally. And I'd always been interested in startups. I worked on a bunch of different projects on the side. I had started to take CS classes and started to like program on my own. And at some point during college,

I think it would have been like my junior fall. Recruiting timelines are weird, but around like my junior falls when I applied to a bunch of different internships at startups, kind of mid-sized tech companies, and then like big tech. And I ended up getting an internship at Microsoft as a product manager on the web browser. And so that was just like...

similar rat race. It's like, you just send out an application, like a hundred different places. Some places you hear back from some, you don't. And then you just like do the interviews and then you find out where you're going that summer. Um, it just happened. Yeah. Well, I mean like, this is a good place to be like being a PM, uh, intern at Microsoft, like arguably the biggest corporation in the world, depending on the week by market cap. Uh, so, uh,

Working on a browser, that's very exciting because there's a lot of movement in the browser space right now. So you got to like – and I should point out Edge is built off of Chromium, which is kind of the open source Google browser toolkit, whatever. Like the collection of V8 and like all these other things like jammed in there. And like practically every browser, I think other than like Firefox and Lady Bird, which is a new browser, uses that. Like Brave uses that. Yeah.

I don't even know the names of the other browsers. Chrome uses it and edge uses it. Uh, I don't think Safari uses it. Does it? Safari has Safari's web kit. Safari's web kit. Okay. And so there's like, yeah, that's right. So there's like three archetypes essentially, or there's like two arcs. Yeah. Three Firefox, Safari, Chromium and Chromium based browsers. And then there are like the rogue builds. Yes. The, the, the indie engines, um,

Although, it's interesting. A lot of your smaller browsers are either forks of... I think Gecko is what Firefox is built on, and then WebKit, and then Chromium. So stuff like Lady Bird, it's trying to separate itself from those different ecosystems and be truly independent, which is super cool. But I was working on Chromium, yeah. Yeah. So what was it like being an intern at Microsoft? Is it hard work, or is it like...

Party, socializing, you know. Is it one of those things where it's really hard to get in, but once you're in there, it's not that hard? Yeah. It's like, hmm. This was my first big company internship or job or anything. I think it's probably similar to like college. Like you get in and then it's like not the same level of work to get in or like what you assumed it would have been to like get there.

Um, so it was like pretty laid back. I think like interns, it's particularly laid back and it was still kind of COVID restrictions. Like people were traveling. So my team was off on a couple of different places. So it was a pretty like relaxed summer. Um, but yeah, it's like,

Microsoft's a massive company and all the big tech companies are like very unique. Our team in particular, I think was like pretty work-life balance. You know, like you're nine to five. Yeah. Yeah.

I want to kind of like jump ahead a little bit because this internship turned into a job, which is the dream. I presume when you get a good internship, you want to convert it. And I think the employers want to convert it too because they've already done all the work and stuff. And like, hey, if we can add somebody to our team whom we've already kind of like tested, they've already invested a lot in you. So you did accept an offer from Microsoft. Maybe you could talk about that process and the internship to job offer pipeline and what that –

process felt like. Yeah. So I think I'm trying to think of like the best way to say these things. Cause maybe it's a little rough in my head with Microsoft. I think like, I've heard that if you get an internship, it's really hard to not get a return offer. Like you have to really mess something up or like not do any work. So it's like,

we accepted you for X, Y, Z reasons. Don't prove that we were wrong and for accepting you for those reasons. Um, so like do, you know, what was expected of you when, when you applied, um,

And so, I mean, like that summer we were all given projects, you know, doing your best to like see that project forward as far as you can. I don't think there's ever an expectation that like that's actually going to turn into anything or, you know, like you're going to work on that when you come back. It's more so like, can you be given some scope in the way you've been hired and then like do your job effectively? Like Bill Belichick says, do your job.

And that's kind of like what the experience was. I think it was helpful having teammates to talk to about ideas and like, oh, when I do come back, I'd love to work on these things on the browser. And here are different features and here's what I was given as a project, but we could take it in different directions. I think that's always super useful. It's not just like...

here's work, go do it. Um, you put your own like creative spin on it. You show that you're, you know, interested in, in making it as good as you can. Um, and just, yeah, being like diligent, being, being fun and like a good coworker to other people, I think is underrated. Like the other interns, how you work with them, being sociable, um,

So not being just like a total recluse and not just being focused on the problem at hand, but actually building these relationships proactively. Just for context, Bill Belichick, I believe, coach of New England Patriots? He had that quote, like – yeah. Yeah.

Do your job. New England Patriots is an American football team for everybody listening outside the US if you haven't heard of the Patriots. And New England is... What city are they based in? Philadelphia? Boston. Boston, okay. Foxborough, Massachusetts. Yeah, it shows my relative ignorance of sports. But the important thing is do your job.

It's part of the team. Like if you don't do your job, it doesn't matter how like buddy, buddy you are with everybody. People are not going to be happy with you. But if you do your job, then the second job to do is to build those relationships kind of for sure. An accurate take.

Yeah. It's like, you have to show up and do what's the minimum expected of you. And then once you, it's like, once you know how to play like chords or like your scales, then you can like riff or you can freestyle. But if you don't know like the fundamentals, then it's hard to do that. And you can't do that in place of your core work.

I've certainly in my life had a temptation. I don't know about you to like sometimes ignore the core work and be like, this isn't important, but I think this is important. And depending on like your employment situation, you can't really do that. It's like you have your core work to do. And then maybe on top of that, you can do other stuff.

Yeah, I can relate this to learning to code. A lot of people, and myself included, are tempted to skip ahead and just work with powerful frameworks and not put in the time to actually understand the underlying scripting language, let alone the C code that runs underneath it and stuff like that. So putting in the time to actually build up the fundamentals, very important. And don't neglect those chords and those scales, as you said. For sure. Eat your vegetables, and then you can have your dessert. Yeah.

So talk about the offer. And like you said, it's kind of like a foregone conclusion unless you really screw up that you're going to be able to return, which is basically convert your internship into full-time offer. Can you talk about that a little bit? Like compensation, like the negotiation process, anything you're willing to talk about. Like just give context.

For sure, yeah. It's funny. It happened so fast, and maybe other people were thinking about it more than I was. But I remember in our intern group, there were spreadsheets going around of people's offers. I don't know if I'm allowed to... I mean, it wasn't like a Microsoft-endorsed thing, but I think...

The interns all talk and it's like, who's getting what offers and what jobs and what level and what's like their base and what's their bonus. And there's like historical data. And, you know, the people he went to college with are like, oh, you know, this is how you negotiate. And so like I was told, you know,

it doesn't hurt to ask for more and make the case for yourself. And so like advocate for yourself and ask for more. And so I did like Microsoft when you start, it's like level 59, 60. And so I think I originally was supposed to start as a 59. And then I said like, Hey, make me 60 and also give me like $50,000 as a bonus. And they didn't give me the $50,000, but they were like, yeah, we'll make you 60. And it's just like a one-time quick call.

And 60 is like a higher pay grade essentially, or is that... That's right. It's a higher pay grade and then it's like more equity, higher base, and then it puts you kind of further ahead to like get promoted. Yeah. And if you were to like...

I guess like the total compensation package, like ballpark it for people listening who are wondering if I get work in big tech, what can I expect as an intern who's converting returning? I think is the term you use. Who's coming from like an Ivy league CS program. This is like the, the bull case. Like this is like, you've done everything right in your life to end up graduating with a CS degree from Dartmouth, lining up the internship and having this opportunity. Like what is that worth in terms of annual comp? I mean, I,

Microsoft, it's probably like $150 to $200, but also Microsoft doesn't... US dollars. But Microsoft doesn't pay nearly as much as Google or Amazon or many other companies. So we're kind of on the lower side, but I think that's kind of the starting bound. And then people who make $250, $300 all in...

Yeah, they're competing philosophies on this. Like, I've talked with people that, like, made, you know, 300, 400K on their first developer job. That is not common. That is an outlier. Statistical outlier. I tell people, you're lucky if you're getting, like, 80K working at, like, a hospital or a –

a bank or something like that. Like if you're working in big tech, yes, the compensation's better, but usually those are like second, third, fourth, fifth jobs that you work up to. You were just able to jump ahead because they have this pipeline and universities, frankly, are, are, uh,

big companies are frankly biased toward just hiring low risk, you know, CS grads from Ivy league schools. Uh, they have like, did they have like career fairs and like where people are proactively reaching out to you and stuff? I mean, we didn't have a Microsoft like tent on campus, but to your point, like the university recruiting pipeline was well established, um, well before me. And I think that's where most of at least Microsoft's intern interns come from. Um,

So, you know, you had like your Berkeley's, you know, you have like your big state schools, like a few people from each. Um, you had your Carnegie Mellon's MIT's like George's Hex, a lot of George's Hex. Like the best CS programs in the country, essentially, which is largely the best CS programs in the world. Uh, cause like CS, like the U S really excels at teaching that well. Um,

And we don't have a lot of CS grads, but they generally do pretty well for themselves here in this country. Whereas in China and India, they have like a million CS grads a year. But like that doesn't mean as much and it's still much harder to get a good job even if you have like

the CS degree is still a scrap because there just aren't as many roles. And so you'll find people having to work outside of software development. But generally, if you're one of the, I don't know, 100,000 people that graduate with a CS degree every year in the U.S., you are going to be able to find a good job somewhere. But maybe you could talk about the current, I guess, miasma around like people getting into the field. And like, do you have very many colleagues, friends that are struggling right now trying to find developer roles? Yeah.

Not a helpful answer. I feel like the people I know are locked into some job or the types of companies they would work for. It's maybe easier for them to recruit. I guess like, yeah, what I've found is like for people who found a first job,

And nowadays they, if they ever talked about like going and doing something else or like going and getting their masters, they're just kind of locked into that because they're maybe a bit worried about like leaving and then not being able to come back.

So it seems like there's a bit like, it's like if you're in, you're in and try your best to like stay in. But then like, if you're out, how do you get in? I'm not sure. Yeah. Yeah. So I want to talk about the process of deciding to leave Microsoft because that's a big deal. You're making like, you know, 150 K a year. Your parents are probably very proud of you. Like the kids from your school, your high school, like, Oh damn, look what Sam's doing.

I mean, they're like, a lot of them are probably working, you know, in the service sector or something like that. Some of them who went to good schools maybe are working in tech as well. But like, I would like of your graduating class, how many, like what percentage of them do you think were making more than 150K when you were making 150K at Microsoft? A college graduating class? Your high school graduating class. High school. Yeah.

That's a good question. I don't know. I'm just trying to underscore how big of an accomplishment it is to get to this position as quick as you did. How old are you, 25? 23. 23? Okay. I think of the 500 I graduated with in high school, I would say, I don't know.

Top few percent? I'm not sure. Maybe I need to catch up more. Some of those people are lawyers and doctors, but they won't be those for a few years. That's right. They're in law school. If you go to a good law school, you can make millions of dollars a year.

If you go to a good medical school, maybe you can make like 500K or something like that, right? Maybe a million. But like software engineering is like right at the tippy top of not being like an actual entrepreneur and like benefiting from like scale and stuff. Like the way you get really rich is you essentially benefit from scale and you do things at great scale. Like Wikipedia has –

Like, I don't know, billions of users. They have probably hundreds of millions of page views a month, maybe billions of page views a month. Free Code Camp, we're operating at scale, but we're a charity. Like Wikipedia, there's no money in being a charity and essentially having everything free. But we're operating at scale, so you can imagine if Free Code Camp weren't a charity and if I actually had equity.

I would probably be out competing a lot of my high school classmates and stuff like that. But the important thing to note here is that you were able to, without necessarily creating a successful entrepreneurial project, get into a position where you had stability. Do you think that your job was in danger or anything like that? Do you think you would be able to continue to work at Microsoft and continue to ascend? Of your peers, are most of them still working there or were they affected by the layoffs?

Microsoft is unique for a few different reasons. The promotion schedule, the compensation schedule is notoriously bad compared to other places. Um,

part of the reason I decided to leave was I felt like, I mean, we can get into the details, but one of them was like, I felt like whatever work I would put in, I wouldn't be promoted commensurate to like what I was putting in. Whereas at like Google or like, I don't know, I've heard like Amazon and Netflix, that like career ladder is a lot steeper and you can climb a lot faster as opposed to a Microsoft. Okay. And do you think, but like, do you think that,

You would have been affected by the layoffs. Like what, what probability would you assign to that? Maybe, um, non zero, like, I don't know, 10, 20%. When I was an intern and I got my offer that fall, half the team I worked with got laid off. Um, so we were coming back from being like interns, getting full-time offers to a team that was like, frankly, most of the people I worked with that summer were no longer at the company. Um, and then, um,

Even when I was there, the short tenure I had, there was a lot of reorganization going on. Mustafa Suleiman came in for Microsoft AI. Before, there were a few reorgs, and then that was a reorg. Then after I left, there was a reorg. Reorg is like a reorganization. Essentially, what they're doing is they're merging different departments. They're moving people around. They're figuring out who is redundant, whom they can let go. That's right. Bosses change and skips and...

And teams and skip level manager would be the manager above your manager. That's right. I'm just trying to define all these. Yeah. I actually learned a lot of these recently. I've never worked in a big tech company. Very big tech lingo. I've worked in a big company. Like I worked in a giant Japanese conglomerate, but it's, it's not a tech company and, and tech companies use like their own terminology and it's a little more sophisticated and systematized than like companies. Yeah.

They make terms to make the job seem more important, but yeah, it's like just your manager's manager. That's Skip. So yeah, I guess reasons why... I think a few things. So one, my rate of learning when I was at Microsoft, I think, was lower than I was hoping. And so my 40 hours a week or whatever I was putting in, whatever I was learning during those 40 hours, it didn't feel like...

It was enough and like the best environment for me to be learning. And I think that really bothered me. Like I would come home and I'd be like, what did I do today? Like, I don't feel like I really did that much. Um, and 40 hours of like really locked in work, you can get so much time and you can learn so much. I think, you know, when you're really making progress and I didn't feel that way most, most days. Um, so that was a big reason. And then to the conversation we're having now, the uncertainty of, um,

Like organization's changing and like your scope and ownership. And can I get promoted if I put in more work than other people, you know, will that be rewarded? Can I like build the things I want to build? Like, I have an idea. I want to take this to the goal line. Can I do that? Like none of those things really existed for me. So it was like, I felt like I should have been paid more.

I felt like I wasn't learning enough and I was worried that my job was going to be, you know, redundant in a year. Okay. So, so there was a combination of fear and also disappointment. Like, like again, we're not trying to throw any shade at Microsoft. Uh, we like Microsoft, we've worked closely with them as a charity. Like we're doing, we're working on some grants with them right now, which we're very excited about. Uh, but this is just like big tech in general. Like this is, this is cinematic of like ambitious people like you, Sam, who are 23 years old,

For whom the world is your oyster, you've got these skills, these coveted skills that like tons of different governments, tons of different companies, tons of different entrepreneurs would love to have a technical co-founder who understands all this stuff and who has actually worked. And so you have this kind of critical mass of opportunity.

And you're looking at what you have in front of you and you're saying, is the risk of me leaving this behind – is there a greater risk of me staying here and not realizing the potential than there is of me just going out and trying it and potentially failing? And most entrepreneurs –

entrepreneurial efforts do fail a vast majority, especially doing something ambitious like consumer tech, right? Or even enterprise tech, which I guess you could describe like a lot of coding tools. AI coding tools would be in that category, right? Business to developer is the category that I've heard marketers call it. Business to developer. Yeah. There's B2C, B2D is business to developer, right?

Customer, business to business is like enterprise and B2B is the acronym they came up with like, I don't know, 15 years ago when I was like learning stuff. Because I was first reading all the entrepreneurship books and like, okay, what is this stuff? Because I did actually go to business school and I learned like how you basically be a middle manager at a giant Japanese corporation, right? Yeah.

And then I actually learned the San Francisco way. And there's somebody like the Lean Startup and Stephen –

I can't remember his name, but he's like the Stanford professor who talks about like, you know, like just get a web, get a squeeze page up. We call them squeeze pages, get the credit card forms up, you know, get people to put in their credit card. And then, you know, you've got something people actually want and then actually build it, you know, stuff like that. Like, like basically how you can iterate as shamelessly and fast as possible to get something, you know, white commenters model is literally build something people want. And most intuitive business ideas that like,

you know, some, uh, some engineer who's just been working at Microsoft for 20 years, they may not necessarily have the best sense of what people actually need. They may have this pet project idea. And that is somebody who actually works as a software engineer. You start to get into people who don't know anything about software engineering and they start coming up with wow concepts. And those are the kinds of people you talk to on the airplane. Sometimes it's just like somebody regales you with their crazy business idea. And you're, you're like, if you know anything about this, you're like, yeah, okay. How are you going to like,

Build this. How are you going to get this funded? How are you going to get customers? Like all these different questions because they watch too many episodes of Shark Tank or something like that. Shark Tank is not business. I think it's entertainment, by the way. And I think it hurts you more than helps you if you're interested in actually running like an organization, whether that's a charity or a startup project or just like a consultancy or something like that. But anyway, I digress.

You are going in to start a plan. You've made that decision. You're leaving Microsoft. Talk about that process. Yeah. And so like, I will say like my team was great. My, my managers were great. Like it was great. Um, I think I'm impatient and there's like specific skills you learn as like a PM or software engineer or whatever, like your role is, um, in bigger organizations. And I can understand if you are like really into like

making Chromium contributions and working with like massive code bases, then I think being a software engineer at a big company that's like contributing to Chromium is probably one of the best jobs you can have. But I guess like maybe an anecdote to summarize what we're talking about. There was a hackathon in February. So I started December, 2023. There was a hackathon internally a few months later and our team won the hackathon and

And it was so exciting because like the, a few levels above me, like someone really important was like, Oh, this is a great idea. We'll give you resources to work on this. And I was like, Oh, this is great. This is like entrepreneurship. This is what I wanted. And by the time that I left about four, five months later, we had made like actually zero progress on that. And I would check in with people who had knowledge of the progress and

It just ended up dying out like a few months later. And I think that was like, by the time I was thinking about leaving, I had seen what the last few months had been where I felt like I had ownership. Like it wasn't something assigned to me. This was my project and I could see the writing on the wall. Like this wasn't the place I wanted to be. Yeah. Yeah. So,

What do you do? Like, did you call your, your parents and be like, Hey, I'm leaving my very cushy job. Like pays me better than like probably 99% of my high school class and like lets me live in like a big city. Where were you? Were you living in Seattle? Seattle. Yeah. Yeah. So you were like on premise, like at the, at the, in Microsoft HQ. Yeah. So, yeah. So like walk us through that, that.

that transition, that decision, how you communicated that decision to your friends and family? The call was definitely made. I think like from the time I started to that June, I had been working on different ideas and projects with, with friends. And I think like, as the months went on, I spent more time like after work tinkering on different things, like just working on like random projects, working with like LLMs and stuff.

And by the time that June rolled around, that's when we were like thinking of applying, my co-founder and I. And we actually had been working on like,

a version of the hackathon thing, but like a different version, just the two of us. And we were getting all excited about like, oh, we can do this in Microsoft and like it's best done at Microsoft. Let's try to get something done. And then we started seeing the same things that I had seen for the hackathon project where you just can't get stuff done.

And so that's when we were like, okay, let's apply to YC. So ultimately that call, yeah, made to the parents. They were like, why? Why are you doing this? It doesn't make sense. But they were supportive. Has you gotten into YC before you left? Yes. So that was like... YC is Y Combinator. It is basically the best, most prestigious...

accelerator you can go to where they give you a bunch of money. They give you a bunch of tutelage, essentially show you how it goes off. And then they're like, all right, go make us proud and make us lots of money essentially. And it's a bunch of like, you know, people who've been there and done that, like Sam Altman used to run it and now he's running opening eye. You know, a lot of people who are like multi founder, multi time founders, or you know, just have lots of connections to like Stanford, Berkeley, and like all the kind of like talent.

Out in the Bay Area that the Bay Area kind of like feeds off of essentially. Like getting this engineering talent, getting this kind of like capture of finding these researchers and pulling them out into industry. Google itself famously founded by two PhDs who had this – like done a bunch of academic work and essentially productized it.

So that is the kind of model. So just give me some context into what Y Combinator is when you say YC. And the Y Combinator itself is like some sort of mathematical concept. I can't remember how – I'm not going to teach it right now. But you did get into this program. It was pretty prestigious. It's hard to get in. I know lots of people have gotten in, but usually they apply like three or four times. And a lot of times they're a little bit later in the stage of like proving out their business model than you were just like applying.

And I had applied, I think, four times in college and never gotten an interview. So it's weird because I had applied in college with... I think ed tech is notoriously difficult. Educated technology, like free course. Yes. It's a hard market. And so I applied with different friends over those years, and we never got an interview. And then once we graduated and went off to work at a job, we applied with...

effectively an idea, um, like that we've kind of whipped up over a weekend and we got in, um, we got an interview, we got in and it just like happened very quickly. It was very strange. Like it was like one weekend. We're just talking about this thing we want to do at Microsoft and

A week later, we're like, this isn't going to happen at Microsoft. Let's do this somewhere else. I don't know if I could be safe, but it doesn't matter anymore. Microsoft, if you have a problem with any of this stuff, please don't sue us. I will take your part in the video. I don't think we had NDAs. So, no, not that serious. And then, yeah, we just applied. And we...

Filled out, it was like 30 minutes. We, you know, had this idea at the time of like building an index of all actions you can take on the web for agents. And the idea was like, if you're going to automate web browsers or like have web agents before people call them agents, then they need to know what actions to take.

And like the open API spec that open AI was using for like chat GPT for like function calling didn't really work that well. So we're like, what if we index the web as like a agent search engine, like whatever that meant? That's what we applied with. And our group partner and the YC people were like, sounds good to us. Yeah, go work on it.

Okay. So that's crazy that like you had a relatively, I mean, what did you have in terms of prototypes and actual working proof of concept? Um,

We had started working on like an MVP, but when we had our interview, Minimum Viable Product. Minimum Viable Product. But we had nothing during the interview. Like we didn't show them anything. We had no revenue, no users, no LOIs, letters of intent from like a business. It was just idea. Yeah.

Wow. Well, congratulations on getting in with just that, like somebody getting in with so little. Um, but that, that's a testament to like, I guess how good you were at applying and, and how, uh, much potential it's on you. So you hit the ground running, walk us through the YC process. Like what did you learn while you were there? You moved to San Francisco or was it remote? We moved to San Francisco. Um,

Put our stuff in the U-Haul, got our asses down to San Francisco, and we started. Am I allowed to swear? I apologize. Yeah, we don't care. Okay. I don't, but you're welcome to. And yeah, it was a very quick, like 10 weeks, 10, 12 weeks. I think like there's different people and different partners and different techniques and different advice and some of it's public and some of it's private and like whatever. But I think like the overarching thing is kind of similar to college. It's like we've chosen you.

Um, first, whatever set of reasons. And now we're going to give you everything you need to succeed, but we're not going to tell you how to do it. We're just going to give you advice and let you do it yourself. And I think that's like kind of the model is, um,

there's some things that can be taught. There's some things that are like, this is true. You should do it this way. And then there's a lot of like contextual, what you're working on, how you're doing things like trying to work through it. Um, and I think it's just like, they give you an environment to learn as quickly as possible. Speed is like that one factor that's most important to them is how quickly can you make mistakes and learn from them? Um, you know, do user research, build a prototype and

And then find out that that's not a good problem to solve or sell to people and get like 10 contracts, but then realize like, this isn't going to be big enough market. Like how quickly can you learn and then use that to like build something better. So it's really like a pressure cooker for the top of the batch. Um,

And you're just like running the entire time and your friends are doing stuff. And some people already have customers. Some people are pivoting like the week before you're supposed to pitch to investors. It's a whole very different people along like different paths. So it was just kind of like a blur, honestly. Like we got to San Francisco and it just happened. Okay. So walk us through that process of,

What happened after you finished the actual incubator process or accelerators, they prefer to call it. It used to be called incubator and that term fell out of favor. And like, oh, accelerator, that sounds more grown up than like an egg. But the incubator analogy is much more apt in my opinion. So what is this program and like how much did you have and like what was the actual iteration process on your product like? Yeah, so...

We started with the agent index idea, and I think we talked about that for a few days. And then we were like, what's actually something we can build? And God, I'm trying to remember. We worked on a bunch of different things. I think first we started trying to sell to people.

We were like, okay, why is an agent index not useful? Well, because automations are unreliable. Web automations don't work. They break. Why do web automations break? Because of X, Y, Z reasons. Okay.

Okay. Well, you want to build the picks and shovels. You don't want to – like whatever. I forget the –

through which people you know uh people had to buy durable jeans to go down there and in in mine uh and uh yeah so so in a gold rush you make money selling the picks and shovels you don't necessarily right well it's a lot safer to make the picks and shovels and yeah maybe somebody comes out and they got this giant nugget of gold and they're like yeah i'm set for life uh but like

Practically a lot of people didn't necessarily have positive expected value from going ahead and mining gold. But Levi Strauss made an entire empire, and I wouldn't be surprised if some of the people listening to this podcast are wearing Levi Strauss jeans to this day. It's an enduring business built around the picks and shovels thing. Anyway, I digress. I didn't know that was Levi Strauss. But that's what you're referring to when you say picks and shovels.

Yeah, so we thought, like, if we build the infrastructure of, like, web automations, that could be something. So we worked on, like, different versions of that, worked with, like, a bunch of companies in our batch, like some healthcare companies that wanted to automate, like, EHR, like, online medical record, patient portal systems using automations, etc.

we worked with one that like wanted to, um, automate like checking websites for price changes for different like sizes and like skews of, um, like goods. SKU is like basically the barcode on different, like it means different products. Different products. Yeah. Sorry. I'm going to continue to, um, you know, I don't even know why I said screw. Like that's like, so yeah,

Because you're thinking about this and these terms, they may sound ridiculous. Like, oh, they make everything sound so important through this alphabet soup. But yeah, this is just how people talk, right? In the military, it's like all freaking –

you know, acronyms and like, we don't understand anything that people in the military are saying because they're using so many acronyms. But Michael, in interrupting you frequently to define these things is just to fill people in. Because you're going to hear these terms all the time when you're doing product development. Right. So yeah, like we, so we started with the agent index stuff. We worked on like web automation infrastructure. Like, can you make them more reliable? And that kind of turned into like,

Um, what if you make authentication better? Because like web automation, the scripts break, the access to a website breaks if you're trying to use someone's cookies. So maybe like you can make that piece better. Um, stuff about like updating, like the scripts or like the selectors we worked on. And then like halfway during the batch, we went to, um, Paul Graham, the founder of, uh, YCs, um,

I don't know if I'm allowed to say this. I don't think it matters, but like we got to meet with Paul Graham and we ended up going out to dinner with him because we walked around his neighborhood and he was like, guys, like you don't really know what you're doing. Like, or it doesn't sound like you know what you're doing. Yeah. I think you need to pivot. And by the end of our walk, we had not figured out an idea. So he was like, well, we got to go to dinner now and talk about this idea that we need to pivot to.

Um, so we like talked about different ideas. Essentially the entire batch was like thinking of something, trying to build like a version of it, not really like following through all the way or like launching it, which was a big mistake. We can talk about that more. Yeah. Um, and then like thinking, yes, the squeeze page up with the, with the credit card form. The important part of that is you need to get the page up, right? Like you can do the user research and, um,

the problem finding the need finding you can talk to people like you can get people to make commitments to pay money but like you do need to put it out there and I think the first few false starts we had we kept not putting it out there and that was a problem

So enough theory crafting. Ship it, squirrel. Just ship it. Exactly. Just get it out there. So it was just like a combination. I'm answering this so terribly, but it's like different web automation, web agent combinations.

tools. At some point we were like, what if we take the rabbit are one, like the handheld and like, we boot up our own software onto it and then like sell it to people. And it's like, we can like punk on rabbit because like we could do automations better than them. So we were just like the, uh, ill-conceived tool that was supposed to be like an LLM hardware essentially. Yes. Like you're failed catastrophically and very comedically. If you watch like product reviews, uh,

Yes, it wasn't a good product. It did not work. The hardware was nice, like the form factor. It was cool, but the software was very bad. Like you would ask it to tell you the time, and it couldn't tell you the time. I mean, why make custom hardware? Why not just tap into the smartphones that everybody's carrying around, right? This is my banana phone here. My kids bought the banana phone because they didn't think I'd actually get it in yellow, but I did.

It's an iPhone? It's yellow? Yeah, it's like the cheap model. So I always get the two-year-old models. They make them in like – I think they're made in India. And a lot of people in India get these phones because they're way cheaper. I think I paid like $400, but it's like a full iPhone. It was just like two years ago's model or something like that.

So you can have the security of iPhone without the price of paying like $1,500 or whatever the hell a new iPhone costs. I have no idea because I've never bought a new current-gen iPhone. I always buy the SE, I think is what it's called. The SE. Yeah. I had the SE in high school. It was great. Also, it's like the new ones are too big. I feel like having like smaller – I don't know if it's like the same dimension. I remember mine was –

It's convenient. It's nice. It's like, reminds me of the old days. Yeah. Yeah. It doesn't, it doesn't feel like a phablet or whatever the term they use for it. Like a tablet. I thought that was a really ridiculous term, but I've never actually gotten to say that word on the podcast. And I'm at that opportunity. Phablet, phablet, phablet, right? Cause the new phones are now the same size as like the phablets were. And we, we clowned on the phablets, but now everyone's carrying around a phablet.

it. You're going to spend all your time looking at a rectangle. Why not make it a bigger rectangle? But I spent all my time on my computer. I spent like basically zero time looking at my phone unless I'm doing, you know, Chinese Japanese like vocabulary drills and stuff. Like I try not to even look at my phone throughout the day. That's just like life advice. Don't let your phone enslave you. Use a full form laptop.

laptops are for creating. They're for interacting. Phones are for consuming. And yeah, don't be a mindless consumer. Go out and create. That's right. Anyway, thank you for attending my TED Talk. Let's jump in. So you –

I want to talk a little bit about the strain on the relationship between you and your co-founder. I know it's sensitive, but like every organization has like founder issues. You know, like Steve Jobs, Steve Wozniak, Steve Jobs screwing Steve Wozniak at a comp early on on different projects when they were working at Atari together. And then ultimately like trying to like basically wrest control of the organization from Steve Wozniak, who was just –

Yeah.

And you – like Y Combinator advocates having a co-founder. Like most of the projects that get funded are two people, maybe even three. But they're usually at least one person who's technical, ideally both technical. When I say technical, I mean people that understand software engineering and can build stuff and not just make –

figmas of stuff, which is useful too. I don't want to talk down on it, but it's great to know and actually be able to build a Next.js app or a Flutter app or something. So if you can talk a little bit about your relationship with your co-founder, and I know it's not the most pleasant thing to talk about, but just contextualize this. Sure.

Yeah. So I had known my co-founder from college and I had known him at Microsoft. That's where, you know, we were both like living in Seattle. And yeah, I think like my takeaway from the relationship and working on the company together is that working in, working on a startup and working like in a startup esque environment is very different than other types of work.

Like it's for specific reasons. You don't have anything. There's no foundation. So you're ideating, you're prototyping together. You're taking like bets and like taking swings, making bets and taking swings on like brand new, like there's no like foundation to like live off over like,

I don't like that. There's, there's no cover fire. Like it's just you, someone else, your pure like ability and you're trying to like chase after something. And so I think like you need to work in those environments with someone. And then when you work in those environments, you'll see like, do we work together in this specific vibe? Like, you know, knowing someone for a long time and like being friendly with them, I think that's like a good, like,

First step or working with them at like a big tech job. Like that's the second step, but it's just so unique to like do your own thing. As you know that it's a set of like pressures and like conflict resolution skills and preferences for work style and like collaboration that you don't get pressure tested until you actually do it.

And so I think like our failure was we had never worked together on like a startup or like even a weekend project that was like serious. Like here's an idea. We want to ship it. We ship it. And then you go through that cycle. And so I think what we learned is like that process we weren't good at doing together.

And so your relationship was one of knowing one another, but not having, having forged in like a hackathon crucible where you're like under a high pressure situation, having to work together. So at what point did like the cracks start to appear? Like how early? I think immediately, honestly. Yeah. Like I think it was clear pretty soon, even like renting a U-Haul and moving out of your apartments and finding an apartment to live in, you learn things about the other person, right?

And, you know, you reflect and you're like, I wouldn't have done it that way. And I guess like if one out of every 10 things there's a disagreement on,

Even if it's like a big disagreement, it's like, that's manageable. That's conflict resolution. But I think once you see a higher frequency of like conflicts, then you realize I probably don't work well with this person. And it's not like a judgment on that individual. It's just, you don't work well together. So I think like pretty soon we saw that, but then like with the pressure of startups and, um,

other people in the batch and like no one knows what they're doing so like that's okay that we don't know what we're doing i think that kind of dragged on the relationship um but i think by the end of the program we had come to the realization that it was not going to work out like a long term yeah you know they say like growth can fix things or there's a better problem

It can mask, yes, the problems. Growth can mask all problems. All problems. When you're on that rocket – I mean look at Uber, right? Look at all the problems they had and yet they were going like crazy. So everybody just kind of like turned a blind eye. Like you should watch Super Pumped with Joseph Gordon-Levitt. I loved that. Oh, it's great. Yeah. Yeah, it is great and you can read the original book too if you want. But basically like just all the structural problems were masked and everybody is like, it's cool. You're growing. Whatever. Whatever.

You know, in Silicon Valley, that's like if you watch the Silicon Valley TV show.

required viewing for software engineers in my humble opinion. That is one of the other themes. It's like the moment something starts growing, nobody wants to touch it. Nobody wants to mess with the toxic mix of mad audiences that are getting this growth to happen. They'll deal with it later after the growth flows. That's when the tide recedes and you see who's swimming naked, so to speak, and swimming without their swimsuits on. That's like the Warren Buffett thing, who is not covered.

When the market is booming, nobody cares, whatever. And then when it recedes, that's when people actually start looking around and figuring out, oh, damn, we were doing this suboptimally and all this stuff. And so it sounds like that is kind of what happened. And I'm just going to say one little anecdote. Just give me a moment because I think this is so cool. So I've always heard

you know who your real friends are when it's time to move. And you, when you were talking about U-Haul and all that stuff like that triggered in my memory, you do learn a lot about people when it's something as menial as like, Oh, uh, people coming up with excuses to not show up and help you like move couches and stuff into the U-Haul. Uh, because like, I've never used a mover. Uh, I don't know. Most people who don't use movers, just move it yourself. If you're like able-bodied, uh, and, uh,

I had a dear friend show up in San Francisco and help move, and it really changed. That friendship has been rock solid. Even though we have a lot of differences and we don't always agree on everything, we're both hardcore trickies. That person showed up in my time of need and helped me move these gargantuan couches, and that was a big deal.

Moving couches downstairs. You really have to trust the other person. Yeah. That's not for the faint of heart. I mean, if someone moves a couch down a set of stairs with you, I mean, you're effectively friends for life at that point. Yeah. Like that's, that's big. So there was this prof, this professor at a local city college and she's moving. And, uh,

I showed up and I'll never forget. There were like 20 people who showed up to help her move. And I was like, man, this woman has some insane friends. Like, this is awesome. And we moved and like, we just grabbed everything, got it out. It took like 20 minutes. Many hands make light work. And that was just really profound seeing like the impact that she was having. And even though she was probably like an underpaid mid-level prof who was probably like,

I think she was like an adjunct. I don't even know if she was like a full prof. But yet people showed up to help her and I was like, man, that's where I want to be when I'm like 50 years old trying to move out from one cheap apartment to another apartment because I'm underpaid, underappreciated, like academic. That really stuck with me. So you learn a lot about people when it comes time to move or get an apartment or like are they –

chill or are they not chill essentially and so so you said like this the crack started and we're not going to name this person i don't want anybody to like research this or anything like this uh i that's not how we roll at frequency we're just using this as a teachable moment your personal experience that's right yeah the crack started to show and i think i guess in some ways it's it's a good thing because we had been living together and we were working out of our apartment

So like personal takeaways is I probably want my own space. So living different places, probably better commuting to an office. I like offices. I like being around other people, especially if it's coworking, we weren't doing that. So like you learn stuff about yourself. Um, but I think when you live with someone and you work out at the same place, you re it's like, it's, you learn those lessons like hyperdrive. There's no like dragging out the relationship. Like you just know very quickly. Um,

Yeah. And so, you know, it's like, and I think because like there is so much to do and like so much to work on, there is no point in like, you know, dragging it out. You just kind of rip off the bandaid and get it over with. Yeah. And so we did. And how far did you get into the process for, before the relationship like totally broke down and you decided like, we're not co-founders anymore. We were at the end of the patch. So probably the last two weeks is when the split happened.

Yeah. And they just went back to industry or did they create their own project? Whatever became of them? He's working actually at a startup in San Francisco that's doing very well. So he's doing very well. Yeah. Great. So you're there. You've got the funds essentially. Like when you get into YC, they give you 500K? Yeah.

That's right. Yeah. And so you've got the funds, you've got like a runway, you can essentially, you know, rent a WeWork and just like live out of SF, which that's not trivial, by the way. SF is maybe the most expensive city in the U S like more expensive than New York city probably. But it's like, why, why is this so expensive? Because they don't build enough housing. Yeah.

Anyway. Not political. We're not political. So you're there and I've lived at SF. I don't want you to think I'm like an SF hater. I lived out for like five years. Four or five years. But you're there and you're just kind of like pivoting repeatedly, trying different things. And now you're a lone wolf, Dev. Yes.

Yes, and that is, I guess, the title, perhaps, of this episode, Pivot Hell. Lone Wolf Dev. Sounds cool. It's crazy, yes. So yeah, that was like the end of the batch, and that's been the last nine months, eight, nine months, is just being solo, working on different ideas. And yeah, when they say Pivot Hell, I think...

Whoever coined that term felt like they actually went through hell. And I can now relate to that. Um, having been solo, like trying to work on different ideas and yeah, that's what it's been. We can talk about why it's pivot hell. I'm like, Oh yeah, let's talk about it. I want to hear the ideas that you've picked up, why you picked them up and why you discarded them. And then like gradually like to what you got to now. Yeah. So I guess a few things. Um,

I worked on, I mean, we can just like fire through them. Yeah. I worked on like, they, they say you can arrive at startup ideas from different ways. So one is like you do user research and people tell you a problem and you solve it. Um, I've never had success with that. It always felt like very impersonal to like the industry or like the tool. I've never, I've never been good at that. Um, you can have like projects you worked on in college or like ideas you have in your free time. You can have like problems you would solve for yourself, right?

Um, and so going through a couple of different flavors of that, you know, I worked on like a personal finance coach, a health and wellness coach. That was like, you texted on your phone, um, like a chat GPT, but it's like through your phone and it's like specific to like health and wellness, um, tracking. Um, I worked on that for myself, but like, I never thought like, I'll go launch this. That was the mistake, never launching it. So that was one. Um, I worked on, um,

a web browser that was probably three or four months where I, when I was at Microsoft wanted to build all these AI features and I never got to build them. So I built kind of my own browser on top of electron, which bundles chromium. Um, and, um,

Got burned out on that because I wasn't getting users. Kind of like a shipping, like, you don't get... You don't put it out there so people don't see it. Like, kind of a problem there. But I think also, like, if you don't have the longevity... Like, if you're not, like, super excited about something, then it's hard to, you know, stay motivated. So...

That was that. I guess we can dive into any of these. Yeah. And I'm curious to hear, when you say ship something, what would that look like in 2025? We were using Producton for a while. Twitter, obviously, not getting as much usage as it used to get.

Social media, I like to think of social media as largely dead. SEO is largely dead. Like most of the free avenues to getting the word out about your thing that Free Code Camp was able to leverage in 2014, those have dried up for the most part. And now it's like mainly paid customer acquisition or, I don't know, getting lucky, just getting like mainstream coverage or somehow going viral on like TikTok. Like you might be able to engineer something, but –

It's definitely harder today, I would imagine, than it used to be. How do you go to market with your ideas and get user feedback?

Yeah. So it says like an aside, it's kind of like the canary in the coal mine. If you go to launch something and you don't know where to start, you're like, Oh, maybe I shouldn't have built this. Or like, I don't, I don't even know who I would even reach out to. Like forget doing it like at scale or like good customer acquisition costs. Just like, where do I find them? I've learned that lesson before building things. And then I'm like, not only do I not care about what I built, but I don't even care enough to find out where people would, who would use this are.

What I would find is like, um, hacker news is, is really good. Um, that's kind of like news aggregator, if you will. It's like Reddit or developers. Yeah.

And free cocaine has been at the top of hacker news several times over the years, but it's been many years since I think we've just like picked up anything there. Mostly because we've just been quietly working on improving what we have. A lot of people are like, what happened to free coke? The curriculum is way better than it's ever been, but you know, we're heads down like refining it and continuing to improve it. We've got like the new English curriculum. We've got a lot, we're making a lot of other moves.

But we're not necessarily making a huge effort to publicize them like we did because now we've got a newsletter that goes out to more than a million people and all this other shit. But those are kind of like developed organization problems. When you're early on. We're not trying to get any semblance of –

And you're trying to break through that universal apathy. The universe is indifferent, and that is the default kind of like aether you have to like propel yourself through if you want to build up any sort of momentum. Yeah, something I learned during the batch was how to write like a very compelling, warm email, or I guess maybe somewhat cold email. You know, if you spend...

10, 15 minutes researching like a specific company that you think could use your product. If it's a developer tool or if it's like someone you think could use it, if it's like a consumer app or something like that, you just have to like put in like the raw hours into reaching out to people. And that could be a Twitter DM. It could be, I guess if you have to go in person, like, you know, go canvas, go put up a table somewhere. If you can't do that, then like you have to do it through the internet.

reach out to companies. Don't discount IRL. Don't discount like hitting a local Linux users group and showing them what you're building or something like that. Totally. Do things that don't scale. That's one of the YC maxims, right? And a lot of reaching people doesn't scale anymore. I think people, they discount those techniques because they're like, if it does, you know, I guess that's why it is a maximum like,

Oh, I shouldn't do this because like, it's not going to scale, but you kind of need to grind in that way to get that initial feedback. I remember one of the startups I interned for in college, we would host meetups on meetup.com and we had like many meetups that we would manage maybe as like shadow meetups. And that's how we found like the initial base of customers ended up being a good channel.

So yeah, like where are they? How do I get to them? And then whatever number of hours you need to get them to respond to give you feedback. And, you know, they say if you get no feedback, like that is feedback. And sometimes like even lukewarm, like temperate, like, oh, this is cool. You know, you get on a call with someone or a company. They're like, yeah, maybe we'll reach out to you about this. Like that's probably the best feedback you can get because no one really cares. So yeah.

That's how you know to cut your losses. Yeah. So you're actually looking for signs that you should cut your losses. Yes. And of course, people are like, oh, the next person will care. And I know people who've been building a project for literally years, and they're still kind of looking for that person who cares. And they're so convinced that they're blind to – the negative feedback is much more useful. Like if you're playing a game and you keep losing, that's actually much more useful than if you keep winning. What do you learn from victory? Nothing really. Yeah.

You had what it took to accomplish, but that's not actionable tips. But if you keep losing, oh, I need to parry this attack and I need to like roll here and stuff like that. Like, okay, now I'm starting to see why I keep losing. And it's the same thing. Like if you're getting like ambiguous noncommittal responses or no responses, that is a response. That is a response. That's something else that I think YC imparted was you're trying to seek –

you're seeking truth as fast as possible. So it's like truth seeking at a very high rate speed. And it's not just positive. Like you want to be objective and self-critical. Like,

We're not getting anything. And it really makes you evaluate your assumptions. Like, why did I think this would be good? Because I thought it was a cool idea because I know how to build it. And I want it to waste a week because I have that problem. Okay. I have that problem. So why do other people not have this problem? How am I different from those people? Did I reach out to the wrong people? Um, it really forces you to like reflect on your assumptions. And that's always been super useful over these last eight or nine months. And before that, um,

So, yeah, test your assumptions and you'll get to the truth immediately. And I think it's also made me think a lot about, like, why do you choose an idea? I guess I was signposting a little bit before, like,

If you think it's fun to build something, that's not really, you know, it's hard because some really cool businesses have come out of building for yourself or because you're taking a new technology and you're kind of hacking around with it. And I think a lot of people maybe get it wrong by saying, I'm going to find a problem to solve and then trying to find it. And they don't have the motivation to go deep enough to find something worth solving.

So lots of founders like that, like they're like, okay, I built a tool that does this, like this email, like client thing, like does this very super specific thing. It solves this problem very well, but people use it. They'll give me money, but nobody's super passionate about it. And I don't really feel passionate about it. And I've got to maintain this indefinitely because this is my livelihood now. Yeah.

Yeah, that comes up a lot in a lot of like lifestyle businesses. Like I said, the person who's making – who's working as a physician but hates being a physician but their family is proud of them and like they have all this thing. And now they just kind of like feel boxed in and essentially beholden to their own accomplishment. And it can be very hard to walk away from something that is working and that's what a lot of burnout is. People burn out because they're doing something they don't feel passionate about but it is –

successful. So it's, you're, they're justified in continuing doing it. They just wish they didn't have to keep doing it. Right. Yeah. I, you know, it's like another maximum of you want, you know, 10 people or a hundred people to love you compared to like thousands or 10 of thousands who just like you, you know, you want the painkiller, not the vitamin. You want the deepest problem you can solve, right?

And if you're not – and another maxim, like use your own product. If you're so disinterested in what you're building that you can't even like use it yourself, that's also probably like a negative sign. Yeah. And some of the things I've been working on, I try to like build them because I know I would use them. Yeah. And it just increases that feedback loop. Yeah. And at the risk of sounding like a lucky bastard basically, like I use FreeCodeCamp all the time.

I Google stuff and I'm like, oh, free CodeCamp article. Awesome. I'm going to read this. I read a lot of books and articles and tutorials that people have written when I need to get up to speed, like researching a podcast interview or trying to make decisions around, okay, what tools should we use for XYZ? We often bandy about free CodeCamp articles and often I'll watch full video courses.

to learn different concepts and stuff like that. And it's very exciting for me to be able to eat my own proverbial dog food all the time and continue learning. I wonder... Go ahead. Do you think that's luck or do you think the reason why FreeCodeCamp has been so successful is because you found something that you can use every day, like the articles and the courses? I think it's luck. And I mean...

I'm a simple person pretty much. I'm not going to lie and act like I'm the most ambitious person in the world or anything like that. I would have been very satisfied if free cooking was even a fraction of what it is today. Today we get nearly a million visits a day, which is incredible for a tiny charity that is... We have 30 people on our staff. We have...

maybe like $2 million a year in budget. Uh, just because of the kind folks who donate the free code camp. We have like only, we have like a hundred servers around the world. So everybody has like high up. We get 99.99% uptime. Right. And we're a tiny little charity. Uh, and like,

Frankly, we're like a candle to the sun that is Wikipedia. Wikipedia will continue to be one of my hugest inspirations. But I would probably be totally content if there were just like a thousand people hanging out in the chat room like, oh yeah, this is really cool. I like what you did here. And if it was like a much more niche interest. And that's how I can say with confidence that even if vibe coding solves software development, I don't think anybody believes that unless they're paid to believe that. You can convince

a person to believe something very deeply if their income is derived from that. And I think all the people who are heavily invested in NVIDIA really do want AI to be the be-all, end-all, multi-trillion dollar industry that they've essentially propped up the market cap to represent. I'm not convinced. I'm a skeptic. But I do use AI all the time. So I like to see... I'm sober and practical in these things. But let's say hypothetically...

does recede from being 30 million developers around the world who are extremely powerful, extremely well compensated, who are running the biggest tech companies in the world and using software engineering as like a stepping stone to like being like a captain of industry. Let's say that like receded and it was just like a bunch of hobbyists

develop stuff like... I'd still be thrilled to just send out the free Coke and Newsletter and hang out in the chat periodically and answer questions that come to me in email and stuff like that. I'd still be thrilled to be a part of this community because I believe in open source. Fundamentally, it's interesting to me. It is...

An interesting problem. How do you get people to collaborate and get things done? How do you work with extremely meager resources and serve people at scale? There are people who like the average person on earth makes less than $10 a day, right? Like, like there, there are literally billions of people for whom if you charge anything, they would not even have a credit card to be able to pay. Right. And like free code camp can continue to serve those people. And, and so I get extremely fired up.

uh, every day helping out with this endeavor. And I feel incredibly blessed to have this opportunity to make this impact. And, uh, for everybody else who's in the community, who's contributing to the open source project, who's hanging out on the forum, answering questions, uh, who's publishing articles, video courses, uh, things like that. Like,

I feel extremely blessed to be able to work alongside those kind, thoughtful people. So yeah, like, like everything could recede, like the organization could become a fraction of itself and I'd still be doing it. And yeah, maybe, maybe some of free cocaine success is bound up in my own individual enthusiasm for tech, but like, frankly, it's like probably 99% open source contributors. And the fact that we were in the right place at the right time and that software development is such a deep skill and,

That you can never learn. It's like learning a world language. It's like learning music. It's like learning chess. There is no final level. And so I feel confident that there will still be many things to teach in

decades from now when I'm like hopefully dying at like age 120 or something I don't know I'd love to live that long we'll see but practically like for the next 40 years of my career I'm like 44 I'm like nearly twice your age and yeah I'm hoping to still be kicking it at 84 like working on free code game

So that was a very long, I'm supposed to be interviewing you and yet. Sorry. Yeah. No, but I attribute that to like, I ascribe maybe if you won't and you know, you,

The open source contributor is very important and like luck and like timing and all of that. But like being passionate about what you do every single day, using your own product, seeing just like having so many ideas for like what you can do with it. There's always like growth and expansion channels. I feel like those are some of the primitives of like what makes a good idea to work on. And I think the last eight or nine months have been idea picking and it really, it makes you,

evaluate yourself very critically, evaluate the ideas critically, evaluate the progress that you're making or you're not making. And so in that regard, it's been like super productive, but in terms of the idea of like finding something to turn into a free goat camp, it's been, it's been tough. Yeah. Yeah. Well, I encourage you to continue to explore. You've got runway. How many years of runway do you have with your current burn rate?

A few years. Yeah, so you can continue to iterate. And I'm optimistic. The thing that strikes me about you and what's so inspiring about you, other than the fact that you...

Just put yourself out there and you keep creating these helpful resources for the community. You and Seth, whom I've also had on the Free Code Camp podcast. You two have developed some courses and stuff together. And you just seem genuinely intellectually interested. And you're not just somebody who's just interested in software engineering. You're somebody who studied debate and so is interested in the trivium and the quadrivium and that kind of liberal arts ideal. So you're like...

At the risk of sounding like old-worldy, you're like a well-rounded individual who has a diverse skill set and a general curiosity about the world. And you continue to explore that. I think it's just a matter of time. All you need is time.

Right. You can bleed out, but you can't blow up when you're a solo founder. That's a very important lesson that I learned along the way. If you keep your costs super low and it's just you essentially, like I was a solo founder of free code camp and I tried a tons of different projects that nobody cared about. I was on hacker news with lots of silly ideas that I thought, Oh, this is cool. But like, ultimately like, is this like an actual business? No, it's a cool idea that like did well on hacker news. Don't confuse the two. And then gradually like just discovering, Oh, okay. Like,

This is actually something people want. People actually care about learning JavaScript and getting good at it. And maybe they're intellectually curious. Maybe they'd like to have a career as a software engineer, whatever the cause, there's plenty to explore here and do here. There's plenty of work to be done and people actually care about it. So, um,

I think you're going to similarly iterate your way to success, and I think you've been very smart in figuring – like right-sizing your burn rate. This is not like a startup advice podcast or anything, but keep in mind, FreeCokeCamp is a charity. I own zero equity. It's like a parallel universe. It's like the bearded Spock is like the entrepreneurial world, and the Spock without the beard, the nice Spock –

If anybody watched the original series of Star Trek, again, I always use that for analogies because I feel like enough people are familiar with Star Trek in software engineering that they get the references and it's just a... That's your audience. Yeah, yeah. So...

The beardless Spock, the good, kind, logical Spock is kind of like the charity. And I think in the 24th century, post-scarcity world, a lot more will function like a charity and less will function like a cutthroat capitalist endeavor, which usually start a plane. You don't actually have to encounter that. It's later that it becomes more cutthroat because initially growth hides a lot of those interpersonal conflicts and people are not necessarily jockeying when the ship is taking off. It's once they're up in space that they start fighting over the rations and stuff.

Right. So, um, I'm optimistic you'll get there because you keep your burn rate low. And that is my, like, I guess number one advice to everybody out there is keep your burn rate low, be fiscally responsive, responsible,

And just keep all your cash, hoard that chest like a dragon would, and do not just frivolously spend. And that is time. Money represents time. People say time is money, but money is also time to an extent. Obviously, you can't buy your longevity if you watch the movie Contact, another excellent science piece of work. The guy – this is not a major spoiler alert, a spoiler in the movie for like a 30-year-old movie, Carl Sagan, awesome dude, science communicator.

The guy is extremely wealthy, but he's got this form of cancer and he literally goes up into orbit and he's living. First, he's flying around in planes because being up high in the sky reduces the rate of cancer growth in his body. But ultimately, he's up on the ISS, the International Space Station, because he's paid money to go up there because his cancer develops even slower up there. So he's literally so rich that he can buy time.

I'm not saying it in that literal way, but I am saying that if you have cash on hand, that's the time that you don't need to go back to work at a tech company to store more cash. And I say this like if you have any sort of kind of like – obviously if you're living a hand-to-mouth existence, you don't have any capital that you can save really. I'm not going to tell somebody who's struggling to put food on the table for their kids, oh, you should save money.

save money for the future. No, but you should, as soon as you start to make enough money, that's like slightly in excess of your lifestyle, sock that away. Keep that lean lifestyle. Keep, keep the, uh, the cinder block bookshelf. Don't go out and buy some, some fancy one off of Wayfair or whatever. Like I paid this like 200 bucks for this. I kind of wish I just had like a, a cinder block bookshelf. Well,

But it is prominent in it's a prominent prop for those of you watching the video edition I put on like some arcade game and I've got like, you know, my my books and stuff on there But I got my this is a hackathon prize that I did not earn For the the icy hacks a big European hackathon It says winner, but I did not participate. It's a nice 3d printed thing Where do you get where did you acquire that?

like at this London University where I was a hackathon judge, Imperial College. And it's one of the more prestigious. And then this is from San Francisco State University. They had like a Grammy, but it's got like an alligator. I believe it's alligator, not crocodile head, but it looks like the Grammy. It's 3D printed. Very nice. So I got a couple of those, but I did not win the hackathon. I merely judged and gave like the closing. To be clear, you were a judge. You did not...

Oh, you still got a trophy. This is not stolen valor. Okay. But like that bookshelf, right? That was an expenditure that I'm like, okay, I'll do it. Right? But like keeping costs super duper low is the name of the game if you want to buy time. And I applaud you for kind of mapping out and saying, I'm not going to hire people. I'm not going to, you know...

get the services of some like business consultancy or anything. I'm just going to sit here. I'm going to think I've got money to sit in a room and think. And I think so much, uh, like spend that money on flights to New York to meet with potential clients. Right. Uh, spend that money on, um, you know, potentially attending some industry event or something like where you think you can get sales, right. Don't spend that money on, you know, Oh, I really wish I had like a Tempur-Pedic mattress, you know,

Yes. I think people spend money on a lot of things they don't necessarily need to, and they could just save that money. And like, I would take that money like this and not an investment advice, but just put it in the aggregate market. Like that's what free co-camps entire endowment is in. We just buy the entire market and we just chill. And you know, it's, it's,

As Warren Buffett says, I'm not trying to make him out of some saint or anything for anybody who would think that I'm a bloodthirsty capitalist or anything like that. But like Warren Buffett says, it's not timing the market. It's time in the market. And if you look at most of the money he made, he just retired. So this is topical. He formally retired at like age 90 or something, right? After Charlie Bunger died at like age 100 or whatever. Okay.

You can basically just most of the money that he's made, he made in the last 10 years because of the nature of compounding interest and just taking that cash, putting it in some sort of like index fund or something and just leaving it there. It will start to gradually compound and the market goes like this. But if you bought a bunch of, you know, S&P 500, I would encourage you to go bigger than the S&P 500. But if you just bought like the S&P 500 index fund,

With a low management fee, don't pay more than like 0.04% is what I pay. If you bought that on the worst day of the great recession, like the housing crisis, if you bought that, it would seem like that was the absolute worst time to buy it. But yes, it was the worst time to buy it. But you know what's happened since then? Over the last 20 years, it's gone up like 200% or something, right? So like...

Even the worst day or maybe like buying it right before the meltdown. But the market will eventually go back up, right? Like you have to have faith that like historically over the past 150 years or something like that, the market has gradually gone up even though it went down for a long time. If you just put the money there and forgot about it, right? So if you can do that, if you have the luxury of just reaping the benefits of compound interest while you're sitting in a room and thinking, that is a strategy. For sure.

Creating luck. Yeah. I'm sorry. I think like creating luck for yourself. There's like, there are moments, you know, you can't like go like brute force luck, but I think there are generally like moments where you can put yourself in good positions and good places and be good people. And if you are spending your money, you know, on that and the most important things. Yeah. I think that's probably a better formula than,

Tempur-Pedic mattresses. Yeah. That's once, you know, you have the, that's, that's an optimization thing. That's not the core piece. Yeah. So again, I just want to emphasize that you're buying time and during that time, you're going to have all kinds of experiences and it's time that you're not at some big corporation focused on like their specific needs. Cause when you work for somebody justifiably, they want you to be working on their problems. Uh,

But you had the luxury of being able to just spend all day thinking about the problems that you want to solve and trying to find something that will work. And that is what Pivot Hell is to me as somebody who maybe was there for a while, but I got very lucky very quick. I don't expect most people to, within two years of iterating on ideas, to hit something like Free Code Camp. So I just do an emphasize. I am an outlier. Do not look to me as like, oh, I'm going to do exactly what Quincy – like I got so lucky, so freaking lucky. And luck can absolutely –

you know, hitting, getting a critical hit can, can make a big difference with the skill Delta. Like even if you're like way down here, you can beat somebody who has much higher skill than you if you get lucky enough. Right. And, and that is kind of how like the world works and just getting a whole lot of at bats, right? Like you're getting a lot of attempts. That's right. At getting lucky. That's having a lot of time. More at bats. Money gives you more at bats and then more at bats made you can get more lucky.

Yeah, absolutely. So I've taken a lot of your time and gotten a lot of this time that is precious to you because you're trying to figure out what to do. But this has been a fascinating conversation. I want to close with a question that I like to ask everybody. If you could send some advice back to yourself when and let's pick a specific point. If you could send some advice back to yourself when you were a CS player.

undergrad. Like, like we'll, we'll go ahead and skip ahead. Like you got into, we'll assume that like everything you did up to that point was great because you got a scholarship to go to one of the greatest universities on earth. Right. Uh, and you're studying CS there. You got into their prestigious CS program. What advice would you send back to yourself when you were like, maybe like 19 years old that might help you dodge or, or might help you be in an even better position than you are today? Uh, focus, find something and focus on it.

Don't live life like working on things between attentions, like a phone in one hand on social media and like doing something in the other. Like if you're going to do something, just go all in on it and just focus. Yeah, I think that would have saved me a lot of grief over the years. And I think still to this day, I'm terrible with focus and patience. So focus. If you're going to do something, just do it. Right on.

Thank you so much for everything you're doing for the global developer community through your contributions to Free Code Camp. Again, I'm going to put some links to some courses that you've worked on in the show notes. Everybody check those out. Sam, it's been a pleasure continuing to talk with you and follow your journeys. You're a big inspiration for me. And I'm excited to see what you do with the next couple years that you have, the runway that you have, to try to find that big idea and pursue it.

Thank you, Quincy. Hopefully next time I have some good updates. We can talk about, you know, getting out of pivot hell, not remaining in pivot hell. I'd love to have you back on the pod once you've figured some more things out so you can share more, distill more knowledge and save people from having to necessarily go through as long a pivot hell period. Although I do think it's probably just like an inevitable kind of like rite of passage that most people have to go through if they want to build something great. That's right. Yeah. Seems that way.

Yep. Well, Sam, thanks again, everybody tuning in until next week. Happy coding.