#175 From electrical engineering student to CTO with Hitesh Choudhary
01:21:23 Share

Hey, Tej Chowdhury, welcome to the Free Code Camp podcast. Thank you so much, Quincy. Super excited to be here. Yeah, we're excited to have you here too, man. Like you have created so many awesome courses for the community over the years, both on your own channel, which has grown. It's like nearly a million folks tuning into that and then on the Free Code Camp channel. So thank you. It's been a long time coming having you on the podcast. Thank you so much. I was super excited for the whole day that, yeah, finally today we are meeting up

Yeah. Yeah. And you're over in Jaipur. It's pretty late there. Thank you for your flexibility in meeting with me. Thank you. Thank you. Yeah. So I wanted to go ahead and start off by talking about AI. You've been doing a lot of courses on leveraging AI and you have a pretty good grasp on the practical limits of using AI as far as building and maintaining a code base. What are your thoughts on the current state of AI coding tools?

AI coding tools, I'll be honest here, they are pretty awesome, but they do certainly have a limit. So the most people, and especially students, when they get introduction to the AI that I want to leverage it, I want to use it, they see like, hey, build me a to-do app or build me this XYZ and AI does all the job. And this is very scary for the students that, oh, now AI can do all the job.

But the reality is different when you move into a different code base. So we do have another company, one of the LMS, one of the pretty big one. We serve around 22 million users there. And the code base is pretty insanely huge, especially on Ruby on Rails.

And there we realize whenever engineers use all of the AI, it has its limitation. The context is not properly served. The code that comes out is not that much of a great quality, but I would not shy away to say that it has increased our productivity in delivering the features as well.

So by knowing the AI, surely you can become at least 1.5 or 2x of what you have been writing the code, but it can't do all of the job. AI is currently on a position where it's marketed heavily that it will solve XYZ every problem on the planet. But we all know it's not going to do that. It has its limitation, context size, and how much it can learn. And when you have to design the features, which was not designed anywhere in the past, AI is not going to help you. But certainly...

Certainly, it can do some of the job faster for me, which I don't want to do it anyway. So in that case, it surely helps you in increasing your productivity. Okay. So first of all, LMS, Learning Management System, you have one of the biggest, most popularly used learning management systems that are used at schools. We call it Learnest. It's serving currently around 22 million user base. Yes.

And it is one of the most secure 1DRMs. So the technology we deal around with that is cutting edge. And some of the basic features, like how do we collect the user's requirement, how we draft it into the user's stories, and then we pass it on to developer. That time cut has been cut down quite a lot. It was previously like... AI tools have helped speed up. So AI tools have sped up kind of that aspect of development. Yeah.

That aspect from seven days to it's almost like one or two days. So we are super happy with that. Yeah. And what are some areas where you don't think that AI can be really readily applied to your code base where you wouldn't trust AI to do a good job of implementing something?

When you have a large code base and you're designing the features, just don't mindlessly trust and hit the tab, tab, tab, or just don't allow it to write the code. Sometimes the tab feature, I'm much more on the leaning side that, okay, this one is better because I can select what kind of code I want in my editor. But compared to if you go just, hey, cursor, hey, Windsor, if you just do all of that, sometimes they do a good job. But most of the time, they're going to touch something.

the places in your code base, which you really don't want to touch. And there is no way of doing that. So that part I am completely avoiding in our large code base, but on the part where it says, okay, we want this. So sometimes it does the good job there. So yeah,

especially when you have a large code base, the context size needs to go into the LLM and they will obviously train on it until unless you have different plans. And on top of that, you will just get up with the tokens pretty quickly. It will say, okay, your work is done for today. Your tokens are expired.

Now, this is day off for a developer. That can't happen in the middle of the day. We still have a long day to run. We have to build those features. We can't just rely on the tokens or you'll keep on spending the money in buying the tokens more and more and more. So no, just relying on token is not a good idea. So that's like an economic consideration in terms of like you're going to run out of...

space in your context window and LLM can only kind of like carry so much in its working memory that it actually understands outside of like the actual underlying foundation model and like what it was trained on. So if you're, if you've got a novel code base that it's never seen before, it can only retain so much of that.

is what you're saying, while it's thinking about it and thinking about how to write code that does things within that context. And most of the people who actually interact with it and publicize it, like make publicity around it on the Twitter, they actually use basic examples most of the time and popular code bases, like on JavaScript, maybe on Python. But once you touch the code bases, somewhat like Ruby on Rails, which is not much discussed or something similar to this,

then you immediately see, okay, I do have shortcomings there, a lot of them. So it depends on the popularity of the language as well. It's going to give you the answers based on that as well. Yeah, so Ruby on Rails, of course, very popular web development framework 12 years ago when I started doing software development. It is used for a lot of tools. I think Reddit used it at some point. There were a few other big websites. GitHub still uses it. GitHub uses it, yeah.

And we are not small, 22 million user base. We also use it. Yes. But not as ubiquitous as just like Node.js, for example. Node.js and modern popular tools. So if I had to start a company today and I had to build some CMS, LMS, whatever, I would probably choose different stack now. Would the ubiquity of code examples and the likelihood of an AI being able to help you

create features, would that be a decision that would factor in? So would you choose like the most ubiquitous tools just because that's likely to have the biggest, you know, prior, the most priors in terms of... No, I wouldn't choose. Like AI is not going to affect my decision in choosing the tech stack. I would still choose, can it serve the purpose? Will I be happy with it? And especially most important thing, which a lot of new developers don't think,

Do my existing developer or will it be easier to hire a developer on this tech stack? This will be my number one decision thing. Yeah. Do you think that there's going to be kind of like a gradual thinning of the ranks of, you know, serious frameworks that would be used to build like a mobile app or a website as more and more people start to consider, hmm, how much training data is out there? Like what is the quality of AI agents in terms of creating?

you know, pull requests going to be based on this. And then of course what you mentioned, which is actually finding developers who are already familiar with tools or, or, you know, languages, arc, um, frameworks that have a lot of existing learning resources, those it's easier to onboard people into no JS than it is, you know, groovy on grails or something like that. Right. I,

I still think that's a very interesting perspective that will it impact the decision of the founders to pick the stack? I don't think that will happen in the next at least four or five years because the trust that my developer can learn this tech stack or move from X tech stack to Y stack if I give him

X amount of time is much higher in consideration than will the LLM will have the more context window. Because I know a human can learn faster and can deliver reliable results than just spending the token. So again, my still take will be that, okay, I'll trust the developers if they are much more trained compared to AI much more trained, I'll go on the human side. I know he can deliver.

Okay, so let's talk about the hype in AI coding tools, for example. There are a lot of IDEs. There are also tools that will just kind of spin up an agent, and it'll go into your GitHub repo, and it'll start opening up pull requests to address open issues and things like that. How far do you think some of these tools are from being sufficiently mature that they can work?

really widely be adopted and trusted within a large bespoke code base like your LMS's code base. Oh, we are very far from that, especially when the context window is pretty huge. We are very far away from that. You know how many lines of code are in your code base approximately? Like within an order of magnitude, is it like millions? Yeah.

Yeah, surely. Definitely. Because we have been developing this LMS from last 10 years. We recently celebrated our 10 year anniversary. So I think it's a pretty huge base and we have been moving from backbone JS to react. And, uh, the little bit for some part we wrote in node JS as well, then came back again to Ruby on rails. We use different databases as well. Uh, we are heavily relied on AWS as well. So it's, it's a lot of moving parts there, but I would come back onto the point, uh,

I don't think so we are anywhere near that I will allow an agent to touch whole of my code base and I'll just sit like this. Okay, you do the job. No, I'm still panicky on that side because the problem with these auto running tools is they touch the file. They just try to solve it automatically.

The problem, whatever the problem you give it, they just try to solve it without worrying that this code might actually break another part of the code base. They're just focused on, I will solve this no matter what part I have to touch. There was one incident where I was recording a video. I just trying to explore this code LLM. And I tried that, hey, let's do this in the fast API. And the code base say tried it two times and they said it's not working. And it switched to be from fast API to flask API.

And the whole code base is trying to write. It just moves you to a different tool. It's just like, I can't get this to work. So, hey, we're going to like executive decision. You should use Flask. And you don't want that to happen. And this was a very small to-do kind of application. I was just testing it around. Now imagine when the magnitude increases. Suddenly say, hey, it's not working on Ruby on Rails. You might want to try Node.js. I'll write the code into Node.js, scrap it. You don't want that to happen.

Yeah. Well, I am somewhat of a, like a, like coding AI skeptic, if you will. Um, I don't really use the tools that much. I use AI for a lot of other things, but like, I don't really totally trust it to write production caliber code that is not going to introduce tons of bugs. It's going to annoy the rest of the developer team. Uh, but at the same time, I talked to lots of very smart engineers who seem to be using this, uh,

I can understand the use case of like, I'm just building a tool that only I'm going to use and it's not a big deal. And other people aren't going to, I'm not going to be making work for other people by being lazy and using an LLM to write part of my code base for me. It also changes your style of writing the code. Every developer eventually develops a writing style. I like my variable names like long, or I like to have my variables names in this style or the LLM doesn't have a style. It has acquired style from thousands of developers. So,

And people don't understand initially that having a developer style is really important. How you do the things. Do you prefer if else in the long format? Or do you like the ternary operators? These small things matters. And when you are using AI, you don't have a style of writing the code. Recently, I did a video on this. I was teaching dependency injection. And the very first code that split out by the AI was exactly that.

that violates the principle of dependency injection. So I was teaching exactly that and surprisingly, I recorded that video as well. So again, if you are a developer who don't want to have dependency injection or have the dependency injection, you have a style of writing code. AI doesn't have a style. Yeah. So...

Give me an idea of a feature where you would use AI. Let's say you were working in the code base of your LMS and you were trying to, there was just a ticket. And like, how does AI actually help you?

In writing, writing user stories, when we ask a client that, hey, what are the features you're looking for? He gives us the vague idea. We sit down with other founders and look for that. Hey, this is what we exactly want. And we have to design the user stories around it so that we can have a story point. AI does that really good.

Yeah. So, so user stories for anybody who's not familiar with this, like this agile concept of basically breaking down as a user or as like a admin, I need to be able to access this type of account and modify this permission. Like it's a very specific granular kind of feature request.

So somebody has asked you that I want to have a to-do app. I'm dumbing it down. So I want to be able to create a to-do. I want to be able to edit my to-do. I want to be able to mark it as checked. I want to delete this. All these user stories, AI does a pretty good job in that. So it's good at kind of envisioning, okay, here are all the different things that would conventionally come with this type of feature and just break it out into a document that you can then hand to the client and say, okay, here are the user stories.

Yeah, that part, we love it. That isn't actually writing the code. That's just helping write the spec that serves as an input for the actual creation of the code.

Yeah, because for the developers, well, the challenge always has been, do we get enough requirement from the client? The requirement from the client always keeps on changing. This is the number one complaint by all the developers. And if we are able to solve that complaint, I think this is a good win. Yeah, absolutely. And it's not too hard to just generate that based because, I mean, GPT's, like whichever LLM you're using, has probably read 100%.

hundreds of thousands of those types of agile design documents. So it can probably kind of pattern match a little bit and do a lot of the legwork for you. So you don't have to just sit there with a blank piece of paper trying to think, okay, well, if I were a user of this type of app, what kind of things would I need?

So I can imagine it saves a tremendous amount of time. Another use case that we actually use is not through the game code, but we had to write a lot of code for it. So we have this internal computer lying around. We just fed a video to it. It's a pretty simple Python script which transcribes all of your video

And from that, all the transcription we use in LLM model, it generates a YouTube description for us. It's really helpful to pick up what should be the ideal title for the video and the description for it. Super helpful for us. We actually use it.

Yeah. And I believe on like Free Code Camp's YouTube channel, a lot of times if you see like these very granular like timestamps, a lot of times those are created by the course author. But like for the podcast, I don't go through and like try to say, OK, how am I going to title this section? Right. But there are timestamps on this podcast and a lot of that is done using LLMs.

So another thing this reminds me, whenever I send a video to you and to your team, I actually put them all together in a one long video. So it's all on FCP, Final Cut Pro. It's an Apple software to edit the videos. Yeah, very good one. Yeah. And on the left side, I have all the timestamps of my video. So instead of writing them manually, I take a screenshot, give it to ChatGPT. Hey, just extract these titles for me. It does a pretty good job. I take that and send it to you over the email. Yeah. And these are the kind of

Like things that like, I mean, even if you're like a real serious craftsman and you really care about the finished product, um,

do you really care about the exact title of the timestamp? Not necessarily. And this is where AI comes in and makes your life a whole lot easier, right? So I can see those as very actionable applications of AI. Are there any real code base applications? Is there any part of the code base currently that is being written with the help of AI other than just using tab completion tools?

No, nothing. So far, we don't have even... And also, we don't have any reliability that we can't do this thing without AI. That is not there. None of our large code chunk bases written totally by AI help, surely. Not entirely. Yeah. Okay. Well, that's very helpful. So there's always a human in the loop kind of proofreading and deciding, eh, I don't like that function. That's sloppy. Or if you do it this way, you're going to cause trouble down the road. Like, you've got a human who's kind of like...

sniff testing all the auto-generated code. But the great thing is, I would advise this to all the people who are watching, the moment you know the code...

You know that, okay, this is how I write the code. And then you see, oh, I have written this code probably 10 times. And that's where your productivity increase. So instead of getting scared away by the AI, I think this is the best time to learn to code because you will be able to deliver the product, which used to take like by four people team. You can actually stand alone, alone, deliver this entire product because you are now much more faster.

You previously were also faster, but it was taking more time to type that, write it out. So learn the code. Don't get scared of the AI. AI is your companion. It will help you to deliver the results faster. Awesome. Well, that's like a very balanced take, and I appreciate the nuance that you've put on it. I want to dive a little more into your history and kind of like how you became this person.

prolific teacher of developers because you've taught like millions of people through, uh, your free code camp courses through, through your YouTube channel. Uh, you've created some other courses. Otherwhere you've, you've literally worked one-on-one or like in small groups with students from IIT, which is the most prestigious kind of public university system in India. And I am curious how you got to this point that so many people were looking to you for kind of like excellent, uh,

Pedagogy, excellent teaching. So teaching is something which I absolutely love because I do have a history behind it. So when I was a kid, we were not doing financially well. And I was in this school and I had to drop out of this school for six months because we couldn't pay the fees of it.

It was really a bad time for us, the whole family. We couldn't put up enough money to put the food on the table, forget about the fees and everything. So I, from the day one, I was kind of a studious guy. I was, I enjoyed learning. The books are not like a work for me. These are like fun time for me. I enjoy that. So not having the education and then early childhood was probably a trauma for me. And one of the day I decided that this shouldn't happen.

surely I couldn't, I knew this fact, I'm very objective that I can't do it for free for everything because I have to survive in my life. I also have to make money out of it. Otherwise, how will I pull my family out of this poor situation? So that crafted me as a teacher that no, teaching should be accessible to whatever I'm teaching. I have taught physics as well, only for like five, six months, but I wanted to teach. That's all I wanted to

Then I saw the coding. It's very challenging. It's very fun. So I started teaching and still a lot of my time goes into teaching for free. So from that trauma to make sure that everyone who even can't afford it, just have a YouTube, watch my videos, learn from it and just get your life a little on the happier side. So that's where the teaching comes up. And I think it's been like 13 years or probably more. I'm still actively teaching free on all of all of the platform. But I do agree that I do paid courses as well.

And again, I have to survive somehow. I have built companies which are totally around education. On one side, I teach and another of my company, Learnest, where we help another educators to come forward and teach. We have teachers who are teaching farming. We have people who are teaching chartered accountants and we have the coders as well. So I think teaching is the only thing I want to do forever in my life.

So even when you're not teaching, you're kind of enabling other people to teach through your LMS company and, and again, LMS learning management system. So you think about like anybody could potentially take a out of the box platform for hosting, you know, video courses, text, maybe some interactive components, um,

Uh, but basically this is the system that like people can sign in and they could like imagine free code camp, right? Like our platform is open source. And in theory you could strip a whole lot of the content out and you can put your own content in there and you could host it and it would be using the same infrastructure, but it's not really designed to be used that way. But your software is designed specifically to be kind of like an empty vessel that people can pour their teaching into. And then people have this kind of a scalable, uh,

software as a service type tool that they can use to learn, as you mentioned, farming, for example. Yeah. And we give these tools to still to a lot of people for free as well. Like another guy who teaches farmers how to do farming, we incur all the cost from him. We don't charge him anything. So that still, that whole thing in me built in as I was not able to take the education at that early childhood. So I can't leave away all of the people and YouTube allows me to do that. A lot of,

my content, a lot of hard work. People have seen that these courses are professionally produced. So I put a lot of effort in producing them, high-end cameras and editors and all of that. I still say, oh, this needs to go out. This is a good information. This needs to go out publicly and should reach as much as mass as possible. So I give my videos to Free Code Camp and sometimes I give it to other universities as well. They also teach these from these videos in their classrooms as well. So I said, go ahead, enjoy this.

Yeah. Well, walk me through your kind of like early life. I mean, you mentioned the hardship that your family endured, how you had to drop out of school for like six months when you just couldn't afford the fees. But like what what was what were your early years like? You grew up in Jaipur. Yes.

And what were, what was kind of like your high school experience? And then the application process to get into university, I, my understanding is it's incredibly competitive because there are so many people in India, uh, you know, kind of competing for scarce seats at universities. Maybe you can walk us through that period of your life.

So I'll start from the early days of engineering. It's a good walkthrough of life, actually. So I started after my 12th exam. This is like final year of your school. I gave an exam in that time used to call it IEEE, which is equivalent of IAT. Now the exams are merged. I scored around 5000 or something rank all over India.

So I got a seat in another city of NIT, which is like IIT, then little lower than that NIT. But we couldn't afford at that time, we were going through the hardship. So I couldn't afford to move to another city and pay the fees. And then one of the college in Jaipur, based on that rank, offered me a whole year for free and said that if you can score this much, your second year will also be totally free. I said, done, we'll get that score no matter what happens.

I chose a stream of electronics and communication. So I studied all about how the hardware works. But studying that, I realized during one of the competitions that hardware gets fried up and you can break them easily. And that's where my interest was like, okay, we cannot afford as many hardware as I want to experiment my things. And the

The situation of college is not that good that they'll keep on giving you the hardware. So even labs, they are not really that good to be honest in India. So I switched on to software and the first time I saw the software, it was very difficult. Even the loops basic for loop took me probably two, three weeks to finally grabs my head. Okay. This is how it works.

But the best part was I can just scrape off the entire code base, start writing it. It doesn't cost me anything. So that was my idea. Okay, I can write the software as much as I like and I can keep on iterating over it. There is no cost. There is no penalty if I make a mess.

So that got me into the coding part. And I learned early on a language known as Perl. It's not really popular now. And Bash scripting, I have done that. And then from there, I learned a little bit of a Python. Python was not popular at that time. Perl was much higher, especially where I used to start. So what year were you learning that Perl was more popular than Python? I mean, Perl,

obviously a very important precursor to most modern scripting languages. A lot of them owe a lot of their features and their usability to Perl. Like Ruby, for example, is my understanding is basically based off of Perl and they just wanted to improve upon it. But what year was this approximately? I don't remember. I'm really bad in remembering the years because the only part, because I remember that I have been teaching for 13 years on YouTube because YouTube gave me a pop-up notification. That's how I remember it. Oh, 13 years. Yeah.

You're not one of those people that focuses on the past too much. Nah. What's done is gone. What can I do? So moving on. So Perl was good. Then I had to order a book. You can't find it. YouTube was not existing at that time. So I had to order a book. I have to request a lot of people. Please get me a Python book. Please get me a Python book. So that's how I moved. And after college, I was much more interested in cybersecurity.

So I moved into one of India's biggest company where we were doing penetration testing for some of the big giants in India, as well as I was still teaching and never left out that. So I requested my company that, hey, we can do a business out of it. Otherwise, the company was not going to allow me if it starts. Hey, we can do a business. Then only the conversation starts. Otherwise, I say, focus on your work. So I said, we can do a business of it on the weekends. I'm free. So fine.

On behalf of your company, I'll go into different universities. I'll teach them how the penetration testing is done, how actually it is there to work in the companies. So from Monday to Friday, it was corporate work. Saturday, Sunday in some part of the country. IIT Bombay, IIT Guwahati, almost all NIT, NIT Allahabad, IIT Roorkee, all of them, I was their guest lecturer for weekends. Used to teach for eight hours on Saturday and probably six hours on

on Sunday because I have to catch a flight, then catch a flight, go back. This was a routine for almost like for three years, I guess, two to three years. Wow. So you were like working a normal week and then you're flying, which is not like a casual thing. You have to get on a plane, you have to go somewhere else. You have to get to like wherever you're going to stay, unload, like be ready to show up and teach eight hours. And then I think you said like four hours the next day. And then, I mean, that's like a huge workload. You had to have been working at least 60, 70 hours a week.

Yeah. And I enjoyed every part of it. And one of the things that it gave me is, first of all, I can teach for long hours, like three, four hours doesn't give me a beat.

As well as I got a lot of command over communication. Once you do this much of communication with the real people in front of you, you need to make sure that your dialect is neutral as well as they are able to understand you because India is a big country. There's a lot of dialects within the India as well. And as the company grew, sometimes we had to teach people which are not from India, people from France, from Australia. You need to be very neutral and they need to be able to understand whatever you are saying.

So this was actually a pretty good outcome from those workshops. Yeah. And you've done a lot of traveling. I don't want to jump too far ahead, but I think you've said you've visited like 43 different countries or something like that? So far the count is 43. I'm planning to make them 45 very soon.

So that's a lot of cross-cultural communication. And as a teacher, like communication is everything. I mean, literally the job is to take what is in your head and figure out a way to get it into the other person's head as intact as possible without being too colored by language or cultural differences or other things like that. Like how do you transmit information essentially? That is the discipline of teaching. And how do you make it relatable and interesting? Well, not the guy –

One of the guy in, I was probably in France at that time or somebody, he taught me that teaching is a two-sided job. One is whatever you are expert in that you have to teach. Another side is learning how to teach. That opened up my mind. I was totally blown away that there are so many research paper. There are so many books about the pedagogy and you can learn about them. One of my favorite book is Geek Pedagogy. And another one is how the

best teachers do, somewhat like that the book name is. So you also have to learn how to teach as well as you have to be master in your craft. That can be coding, that can be teaching how to repair phones, whatever that is. So it's a two job at the same time. And now the digital thing is there. So you have to be on the three side. You have to learn how to record the videos. You have to record microphones and everything. So teaching is not just teaching. It's a multi-hatted job. Yeah. And a lot of teaching is not necessarily conducted in a lecture hall.

It's conducted through YouTube or it's conducted through, you know, a thoughtfully authored technical tutorial or a piece of inner active learning. You know, uh, you could, you've studied electrical engineering and you could probably think of lots of ideas for how you might create like a software environment for learning circuits, for example, or something like that. So there's a tremendous amount of, uh,

I call it like instructional design. I consider teaching as the actual kind of like discipline of like actually communicating and instructional design is creating like an artifact that other people are going to use for learning. But I 100% agree that you have to have the domain expertise in what you're teaching and you also have to have the meta skill of effective communication essentially and empathy, learner empathy and lots of other things.

So that is the thing which I acquired in early life, that communication skills, which has taken me to a lot of places. Communication, I would say that, hey, learn this first. This is the one thing that you want to master as early as possible. And

Whatever the language you are going for, try to get a little bit mastery over the English because it's like an API. It will help you to connect with a lot of people come from a different France, Germany, India, a lot of people. So after that, I got my interest in iOS development. I realized that, hey, this is really an interesting thing. I can just build an app and thousands and hundreds of thousands of people can actually use my app.

So I didn't jump into web development first. Before that, I was an iOS developer. So I started learning iOS development and whatever the money I saved, I bought a Mac mini with that because that was the easiest option to get into the iOS world.

and started making iOS apps for one of the company. I was just the second engineer that they hired. I said, I don't know much, but I can help you out. I can put relentless amount of hours in that. So even if I'm not that good, some else engineer will do eight hours. I'm okay with doing 16 hours. Just get me in. So they were very impressed that this guy can, even if he doesn't know that much, but he's ready to put in hours. So hey, he's a no brainer to hire. So they hired me and I learned a lot of things on the go.

And during that, we built a lot of good apps. They were like a service company which delivers apps for other clients, so learned a lot. And during that time, I discovered Udemy. Udemy, the popular website where people buy and sell courses. They'll create courses and list them, and a lot of them are pretty inexpensive. I know it's extremely popular in India. Yeah.

Yeah, it was not inexpensive at that time. At that time when it came out, it was around $150 and $200 per seat. It was expensive in the early days. So I put up a couple of courses there and enjoyed, but realized that Indian audience cannot buy the courses there because there was a credit card issue early days in India. It was a pre-era when the payment gateway system doesn't exist. Stripe was the only one. PayPal, Stripe, that's it. Yeah.

No other platform was there. And India is famous for, I think it's called like UDP or something. It's like a... Yeah, OTP. It's one-time password. It gets on your phone. You get the confirmation. Oh, I was thinking of the payment infrastructure thing.

Oh, UPI. UPI, yes. This is fairly newborn. This is fairly newborn in front of me. So OTP was the previous one, one-time password. You enter a credit card detail, message code will be given to you, and then you can only add that. This infrastructure was not supported by PayPal. This was not supported by Stripe. So at that time, we started a company, Learnist, that will be an LMS company, and I will be the first user of it to teach the people.

And fast forward, not to bore you a lot on this, that company, which I started LearnCodeOnline, got acquired for around $1.1 million, $1.2 somewhat million. It was a stock deal. So that got acquired. And the company who acquired it got acquired by PW, which is the unicorn in India. It was $30 million, somewhat here, there, $30 million acquisition. So we all got exit.

I served for a while in that company as well as the position of senior director. Before that, in that company which got acquired, I was CTO, chief technical officer.

And after that, I got acquired in that company. So $30 million exit. And after that, I served there as a senior director. Pretty fun. But then I came back and I realized, hey, I can teach in Hindi as well. And that bug inside me that I want to teach more. I want to teach more. So I said, hey, I'm retiring from the corporate. I want to teach full time. I want to do this. So I started this new channel, Chai or Code, which teaches programming in Hindi.

And it's getting pretty popular. We are almost like 600k within a few months. I'm enjoying this. Teaching is what I love. And so this has been a long story, what I've been doing.

Yeah. So that, thanks for putting that in the context. Cause a lot of people are wondering like, okay, how did you get to where you are, where you have, I mean, a lot of people would say like the relative luxury of being able to teach people because it is hard. It's hard to sustain yourself as a teacher. It's hard to sustain yourself making like YouTube videos, even with the YouTube revenue and things like that. Um,

You have basically used your career as an engineer to kind of vault yourself into this position where you can, you can focus full time on teaching. And is that all you do? Like, are you basically just teaching all the time?

Yeah, most of the time, like my 80% energy goes in just teaching. But yes, we do paid courses as well, because we have to survive with the team as well. Now there's a lot of team and 20% of time goes into the LMS, the company that we have Learnest. So there we go. And we still ask, we call a lot of teachers, hey, what is the difficulty that you're having in teaching? Can we build more features for you? Can we build more software to you? And this gives us a unique idea of what's happening latest in the tech. Do you need machine learning? Do you need AI? Why do you need AI? Right?

Is there a reason behind it? Don't tell you need AI. Tell us what is the problem that you are trying to solve and we'll build a feature for it. Whether it takes machine learning or whether it's take AI or just a pretty good old solid MySQL database can be solid like that. So this is where my most of the time goes. I imagine a lot of people just think AI is like the magic solution to everything. Oh, just use AI to fix it. Just AI it. I heard that AI used as a verb the other day. But in fact, a lot of off-the-shelf tools that are...

tried and true, uh, maybe much better for those types of situations. So now that we've established kind of like what your, your journey so far has been as a developer, like from an electrical engineering student to being CTO, to being a teacher and just kind of helping maintain this existing LMS project, learn this that you've been working on for the past 12 years, how many years?

It's 11th. That's when I remember because just we had a celebration for 10 years. That is the only reason I remember it. Yeah, there has to be some notification. Otherwise, Hitesh doesn't know how long it's been. Sometimes I begin hard time in somebody when ask me, what's your age? I have to do calculation. Yeah, let's see. I was born in... I've definitely been there as my age creeps upward. So I want to dive into...

some of the things that you've learned from teaching students. Because India, of course, has more engineering students than maybe any other place in the world, other than maybe China and India have comparable volume of engineering students. And of course, computer science is the most popular major, I believe, in India. And I think a lot of that is because, as you said, code is infinitely reproducible. It's cheap to screw up something with code and just go back and fix it or revert things.

state of get you you've started to get course on free cocaine by the way. Um, and as a result, there's like little, you know, risk in experimentation. And, and so you can just really treat it as a sandbox where you just knock it down in your castle, start building again. And it's not a big deal because you've got like the same sand, you just restructured it. Right. But with actual electronics, like you studied in school, um,

There's a physical cost. There are materials. Things can break in a state that is not salvageable, and you have to throw out the board. Any number of things can happen. They can impede your learning. But as long as you've got a computer, you don't even necessarily need an Internet connection. You can just be in there on a plane disconnected from the Internet, writing code and running tests against it and thinking about what you're building. So it is a much more accessible environment.

tool and then you can scale it you know without having to like do a production run and make sure that your chips are exactly as you want them and have them physically shipped to like people that are going to use them no you can just deploy through the internet you can deploy through the the apple app store i know that you've done a lot of mobile app development you can just put it out there and then people can start downloading and immediately using it so do you think that that partially a

or explains why computer science has become like the dominant major in India? See, India is a country which is still evolving and there's a lot of people who are still not financially stable. And, uh,

you just look for what is the first thing that can get me financially stable. And I think that is why most of the people are choosing computer science, because I do understand a lot of people think go for the passion, look for what interests you. But when you don't have food on the table, you look for the first option that can get you out of the poverty line. And that is why a lot of people look forward that if I can get into computer science and I can learn to code,

at least I can get this much of the package so that I can help my family and make myself stable. And a lot of companies which evolved in India over the last decade are mostly software-based companies. They have been outsourcing their software development skills to a lot of US-based client, to the Australian clients. So that is where the need was there. Companies require more engineers. The engineers were looking forward. What's my first...

way to get out of this poverty line and how can I make money? That is how it evolved. So I think that is where the computer science students are going there. And I think it's a good choice altogether. Computers are always good. There's no brockage, cost or anything like that. But one thing I would like to mention up here is

There are a lot of people who enters into the computer science. They study for it for a six month or so, but the drop rate that I can't do the computers is significantly higher. People move into different roles that I'll probably move into design. I'll probably move into product manager or something.

The year that they stay in actually writing the code, this duration is very, very small. And that bothers me. People should be writing code even if they have moved into more managerial positions. So please go ahead and write that. So that's the first part of your question that how is it going on? And another thing that I look forward is that, hey, don't worry about the schools. Just because I studied electronics or somebody studied computer science or

The value of what you get in the college or the professor is teaching in the college is

is not that significantly high at this point. You have internet access. You have a lot of great teachers teaching you online, a lot of stuff. So I believe now is the point that anybody can just jump in and start writing the code. So please don't hesitate. I'm a very biased teacher towards the code. I want everybody to learn the code, but it's fair. I'm biased towards that. So take my advice with a grain of salt.

I will always encourage you to write the code, but this is not the only thing. Look forward. What can actually give you

The nearest monetary advantage, if your ecosystem is designed for the companies like in India, it's all ecosystem design. We have the fastest food delivery system that delivers food in five minutes. Now imagine the scaling that they are working on, amount of orders that India is a country of scale. And on top of this, the greatest advantage that every single user in computer science is VCs, venture capitals.

they are literally sponsoring you not directly but indirectly let me tell you how you have a free way of writing the code you can write that code in vim or in vs code sponsored by somebody you just use the software for free you can write this you can use next.js or probably you can use ruby on rails somebody's sponsoring that framework somebody's actively building it once you build this let's just say a next.js app you can host it on warcell

Somebody has paid for Varsal. You don't have to do. There's enough generous free tier plan or maybe use render or maybe any XYZ service or get into AWS, get the free credits.

Somebody has already paid for you and you have the ability, this magical power of building something, build for something like your nearest gym or maybe your nearest juice shop or nearest food capital or whoever is selling the food. Just build for them and charge them that, I'll charge you X amount of money. You can literally build an entire business or software absolutely for free. Thanks to the venture capitalist. Yeah. And this is something that I encourage like there, like so many actionable pieces would like to give there. So I'm going to recap first. Yeah.

India is the country of scale. I love that because you're right. There's just a ton of people. They're trying to get a lot of stuff done. Everybody's busy. There are so many experimental startups going on, and many of these are funded through venture capital or they're using tools that were developed before.

With venture capital money, as you said, you mentioned some of the hosting providers, some of the web development frameworks and mobile development frameworks that are essentially paid for by large companies that are developing them. You can just grab these off-the-shelf tools. You can use them. And one of the things you said about going out to your local juice shop and just being like, hey, can I build something for you? And then essentially building a stable of contract work, a bunch of freelance clients, and being able to

pay, you know, to keep a roof over your own head, maybe feed your kids with that money. Uh, and essentially just leveraging the fact that there's so much free stuff in software or that's heavily subsidized. Uh, I really like that. So what advice would you give to a university student who didn't get into like the most elite institution, or maybe they got into a good institution, but they can't get into the computer science department. Like that is pretty common here in the U S for example, you get into a good school, but

there's like a waiting list to get into the CS department because everybody wants to be in the computer science department, right? Because again, that's where like the highest compensated jobs generally are is that they go to computer science majors. So what would be your advice to somebody who is at a school, but they don't necessarily have access to the very best profs or the best mentorship that the university could provide? How can they use the internet to supplement their,

So it's a really simple advice. See, the only advantage that you get in the top tier IIT or an equivalent in the US now is the peers, the ecosystem that you get, because you have worked really hard. That's why you have gotten to IIT. The other person also has worked really hard and he also got an IIT. So as long as you both are on the same level of putting in the effort and putting in the hard work, you will be gelled up quickly. You will become friends quickly.

And if for some reason, because of the competition, you are not able to crack IIT, the only thing that you're missing is peer group, your fellow students who are actually together with you, who are equally passionate about something that you want to do. So the only difference which I have personally seen in the IIT and the non-IIT is the

The amount of effort that IITians put on a regular day versus the people who just say, I don't want to study much. Let's just go and have fun outside. But in every college, there are some students who say, OK, I was not able to get in that college, but I'm equally hardworking. I'm equally passionate.

So the advice is find your peers. And this problem is solved by Twitter a lot, Discord a lot. And there are other community groups and meetings that happens offline. So find your group. Where are the people who are equally passionate just like you? And that's it.

You have supplemented yourself and you have moved yourself into a group who are equally hardworking, who are equally passionate. You can find them in the comment section of YouTube. Sometimes you can find them on a Discord channel. You can find them on Twitter. So follow them and the algorithms are pretty good. They will actually keep on pushing you in that zone where you find the like-minded people and just get into that.

And apart from this, the whole knowledge, even the best of the best teachers from IITs and from NYU and all these, they are also on YouTube. They are also sharing their information. They're tweeting it out. They are making YouTube videos. So whatever the best you can learn, the information is not restricted now. The only thing that is restricted still is the experience.

So get enrolled in yourself in that experience, whether that's through the Discord, offline hackathons, online hackathons. Get into that hardworking. I'll work really, really hard in that. And another thing which every, every single student should do if you are in computer science or you have a tiniest interest in the computer science and especially in the world of AI, build your own product, not a project.

product, at least attempt high, I will replace the Discord. I will replace the Basecamp. I will replace Twitter. Have a crazily passionate product which you are super serious. You buy a domain for it, not really that of expensive one, probably $3, $5, you can get enough of that. And just use leverage

the frameworks and hosting provided by the free or just invest a little bit. Once you invest a little bit, $5, $10, you will be more passionate about that project. But at least have one ambitious project that everybody says, oh, you can't beat Basecamp. But it says, oh, I will try at least.

The least that you will get out of it, you will learn a ton of things. You can use it as a portfolio project as well as if it works out, that's the best. You will have the money of it. If it doesn't work out, you will be hired in a company based that, oh, this guy, if he can build this much, he surely can build some of the feature for us.

And keep that project always alive. Work out on weekends or once in a month or something. Keep that alive. It will keep you alive in the latest of the tech what's happening, as well as will keep your interest as a software developer. And this is what I miss. Everybody keep on asking me, hey, what project should I build? Should I build three projects for my portfolio? Should I build a booking system or an OYO clone or...

maybe a hotel booking system i says no you don't need five or ten projects or three projects just build one but it should be crazy good and it should be alive for one year two year three year there should be an attempt that i want real users on this project so turn your projects into product and you will have a really good time in software so many things to unpack there so first and apologies if you can hear like the my neighbors mowing their lawn so there's like a rumble in the

Find your peer group. It doesn't really matter whether you got into the most elite school. If you can talk to those kind of like hardworking, motivated people who did, and you can do that through online communities. You mentioned like discords. You could go to hackathons. I've,

Been to a lot of student hackathons recently, uh, just serving as a judge, uh, serving as speaker, kind of hanging out, just meeting people, talking with them, learning from them. And yeah, like a lot of those people weren't necessarily enrolled at the school that was having the hackathon. A lot of times anybody can show up and just participate. Uh,

And you may be able to get into other events. I've gone to lots of universities and like just sat in as like kind of like auditing a lecture or like the professor will say it's open to the public. You find out about those things. You go there, you're sitting among the highly motivated, you know, uh,

hardworking, diligent students who got into that school. And you can basically just take notes alongside them and ask questions and talk with everybody too. So that is like a really cool kind of like observation that it's not the content of the lecture is it's not the actual information that is everywhere. And you can just get that on command. It's the people essentially, and what you can learn from those people and those relationships you can build.

The other observation you had is instead of building a lot of different projects, just build one and don't make it a project. Make it a product and make it with the intention of having a lot of users and maintaining it long term and continuing to maintain one code base instead of just jumping from one beginner project to another and kind of like reinventing the wheel every time.

Just get further and further. Small breakage here as well. Initially, it will take you two, three different projects to work on with because you don't know the code and building the full one. So by this means, don't interpret as that I'm saying that, hey, don't follow the tutorials which says, hey, build this and relax. Initially, it will take you two or three of those projects so that you get the whole idea of how to build a project and then try to have your own product. Okay.

So, so kind of like stair step your way up through the tutorial projects to the actual main event, which is the project that you're going to maintain longterm. Yep. Awesome. That is super actionable advice. I really appreciate you sharing that. So have you learned anything about the way that higher education works in India? That was surprising. Like, like, wow, that's really, that's like how it still works.

Very surprising, actually. So I still didn't submit my thesis. I am not still a PhD and I'm not intending, I have zero interest in submitting that because I'm

First of all, I didn't find much value in it. The value that you used to get having a master's was significantly higher like 10 years ago. Now it's not that much valuable. And I have a couple of reasons for it, and I'll be very open. Yeah, please. Feel free to judge me on that. Yeah, and I just want to emphasize, you have a master's, I have a master's. We're not like anti-graduate school, but I love sobering takes. Please, please.

Yeah, I still am a pro guy of education. I still say nobody should drop out if you have gotten finished that. I'm still on that take. But what I've realized is, first of all, a lot of colleges are there. But the amount of effort that should have been put in paying the teachers was so much cut off that the teachers don't have any incentive to update themselves.

If you're going to pay a teacher bare minimum and you're going to put all of your money in just having the best of the infrastructure, the building should look shiny and you're not even giving the teachers an assistant so that he can take the attendance and do the regular chores and he's doing everything and is being paid very less. He doesn't have the motivation to learn. If the teacher is not motivated enough to learn and explore the things and have his own free time, he's not able to deliver that specific knowledge to the students. What's the result? Student feel like I'm not,

really interested in sitting in this class. I probably will go out and watch a YouTube video that that guy has put more effort and more energy in delivering that video. So I also felt the same during the master's and PhD. Some of the teachers were just downloading the presentations from XYZ schools and give it to me without even reading it once.

Hey, this is the material. You just read it. I said, what's the point of being here? If I had to just get this PPT, the PowerPoint presentation, I can just download it myself as well. I don't come here for a 20 kilometers ride to get this PPT. You could have mailed me as well.

And you are not adding any value to it. You are not able to do all of it. So that's one take that it's not really good because the teachers and the mentors, they are not good. Apart from that, like a lot of time years ago, this was your second chance of getting into the IIT. If you haven't gotten to IIT in your bachelor's degree, you again can give an exam and get into the IIT and get a tag in the master's. But realize later on,

I just want to point – just stop you there and point out that it's very hard to get into Harvard as an undergraduate. Very hard. It's maybe like 2% or 3% of people who apply get in, and many of those people who apply have perfect –

uh, entrance exams, like what we call the SAT in the U S uh, and still don't get in. Right. But if you go to the Harvard, uh, graduate school of education, you can get into a master's of education degree program and they admit like half the people that apply. So it's much easier to get a Harvard degree if you're pursuing as a graduate degree. And of course it's generally more expensive and like the, the graduate programs are kind of like a cash cow. Certainly the masters are for a lot of these schools, but I,

When you said the graduate degree is kind of like a second chance at getting a prestigious degree, I immediately jumped to that disparity. Like 2% to 3% admission versus 50% admission rate. I do my observation. I have a lot of free time. I do have observations. Okay. So this is how it works at a lot of Indian schools as well? Or is it how it historically works? Is it still like this?

It exactly works in a lot of countries. I have traveled to a lot of them and I ask these questions to general public whenever I sit by near them in a park or in a food restaurant. I just like to ask these questions. It happens exactly same in all over the world.

It's much difficult as a graduate program. It's very difficult to get into a master's, like much easier to get into the master's program. So coming on. So once a student gives used to give 10 years ago this master's program, again, it was easier to get in. But still, you get the access to these great professors, their network and the companies who come to your campus to hire you. They also used to give a chance to the master's students and graduates.

It's an easy placement opportunity. Now the things have changed. A lot of companies are not allowing the master student to sit in their campus placement. They say, no, we only want graduates of this year. So only bachelor's programs. So that is one drawback. Apart from this,

The value addition by the teachers that has to happen is not that up to the mark now. Something needs to be done. It's not really a value addition. I didn't get much of the value addition into the master's. It was certainly the books were good. The syllabus is really nicely because it's being marketed.

So the syllabus which needs to be marketed needs to be really great. And I just followed the syllabus. I got the book from online and all these orders from Amazon, studied them. I enjoyed that. I learned a lot of my machine learning and neural network during my master's. Enjoyed it. But I realized that the papers that I'm giving, the exam papers and the teacher's,

they're not qualified for teaching this material they don't have that expertise they haven't put up this effort to actually build that there's no challenge in it and if you don't have a challenge you don't learn anything so i am not really a pro guy now i don't advise people to hey you should get it this is the only path no it's not the only path now you can build the best of the best product shoot it up on the twitter it can get real users

It's not a great value. In the previous time, education institution used to fund your projects as well. Like, hey, if you build good, we are doing that. Now, instead of doing that, they have become an incubation hub because the startup is the new thing. So whether you have studied from that campus or not, you can still enroll in that incubation hub. So again, if you can find peers through the other route, why to go this route? The whole idea is to get into the peers ecosystem. So that is what is happening now.

Another thing that is happening in India is now it's much there are different routes of getting into these IITs because in India, IIT is such a prestigious symbol that you flaunt it off throughout your life. This is kind of a thing.

And now the education newer system is you can enroll into different sort of ed tech companies. And these companies have a tie up with the IITs and you can get a degree from that. So you literally have never been into the campus. You have never been into that peer group. Wow. So they're like kind of pimping out their brands. Like this very prestigious brand, they're diluting it by just letting anybody give them a bunch of money to get it. That's wild. Yeah.

That's really happening?

Not even a campus. You can literally lease out a floor in a corporate office and call it as a campus. I don't know how it is working and how it is not working, but you can just get it. They will get exams and everything being processed.

Like you, obviously you give the exams, but it's not you are being in the campus. And same is happening, almost same is happening for the prestigious management degrees as well. We call it as IIM for management. Same is happening for them as well. You never step your foot in the campus. You have never met your peers. You have never grinded that, but you take this route. I don't know how much fruitful is it going to be in the future, but it is happening. Well, it seems like a horrible long-term, you know,

I can't imagine like free code camp. If somebody said, Hey, if you just pay this company a bunch of money, you can get the free code camp cert.

without actually having to go through free code camp. Like we would never engage in something like that. And yet that's pretty much, it sounds like that's what a lot of these prestigious university systems are doing. I do know that in the U S we had this thing called trilogy, which was this company that would like go and pay the university some money so they could like pretend to be like their coding bootcamp, even though it was like an unrelated company. And then the university, like you, you'd go to like a Rutgers state university,

It was a pretty prestigious school in New York, and you'd attend this program, and you didn't find out. It was in the fine print. This isn't actually Rutgers. This is a different company, but we're on Rutgers campus, and we gave them a bunch of money so we could basically use them in the marketing and stuff like that. But these kinds of agreements, they dilute brand equity, and it's just really bad for long term. I can't believe that's happening. That's crazy. Yeah.

People don't understand the whole point why IITs got popular is because not because they are IIT, because the people who studied from there, they were extraordinary. They understood the concept. They delivered the knowledge so well. That is why their school got popular. Now, imagine if same amount of people who don't understand the concept or not that hardworking or not that brilliant, they will get out what will happen 10 years from now.

Right now, you get a lot of advantage by saying that I am from IIT or I am from NYU or I am from Harvard. But eventually you find out that, okay, there are two kinds of people in this institute, one who have actually studied there and are brilliant and another one who are just there, but there is no way to find out how you're going to judge an IIT like from 10 years from now. Yeah.

Yeah. So essentially there's going to be like a dual kind of like system essentially where the, okay, I see. And then people have to put in parentheses, you know, undergraduate, like admitted, you know, or like some kind of way to differentiate. And then people who aren't able to claim that eventually employers will figure out a way with their application tracking applicant tracking systems or something like that to filter out people who, uh,

uh, see like maybe with machine learning models, they'll figure out, okay, how likely is it? This person actually tested really well and, and had met all the requirements that got into the real program. And how likely is it that this person just paid somebody a whole bunch of money? Let's use AI to solve this problem. Yeah. But, but it will eventually be solved. And eventually maybe the universities will be able to get off scot-free because like, uh,

It's just like a workaround. They never formally addressed the fact that their brand was diluted. But maybe the actual proof in the pudding is in the eating, right? If they're actually hiring candidates who went through the traditional path as opposed to the paid – the bribe path. I'll just call it bribe path. It's not like – I'm not sure how to word it. But I mean essentially you're bribing the university to give you some of their –

Brand equity to be able to have that halo effect. And there is always a reason behind it. What I loved about the people as a psychology and everything, see, my right is your left. Your left is my right. So there's always a perspective. Everybody justifies that system. So the way how justification comes from the companies who are doing this collaboration with the universities is, hey, these universities have outdated syllabus.

and their teachers are not really that well trained, which I agree with that. So they say is that you come onto our campus, we'll give you the real world knowledge, which we have, we'll give you the like a long four year running coding boot camps. We'll teach you all the machine learning and data science and data structures and whatever you is the long syllabus is. We will prepare you much more for the real world

So that you can have a degree as well, as well as you are prepared for the real world. So I like perspectives because in perspective, everybody's right in their own mind. Yeah. Yeah. But I mean, I can't help but think like, okay, what you just described, that was like, that used to be, you create your own university and you build a reputation up for just being better. Right? Yeah.

I mean, there are lots of engineering schools that cropped over the last 50 years in the United States that just aren't really good schools. And they didn't need to take some weird shortcut where they like,

took somebody else's name and made it. They just went through that path. But I guess in the United States, nobody's creating universities anymore. And maybe it's the same in India. Maybe it's just too long, bureaucratic of a process. It's a process to have universities, so it's better to have a tie-up. And the students who are getting into those four-year program, three-year program, they probably can have a degree as well as a safe side as well learn. So again, what I really like about it is

IITs might be saying, I don't know the official statement from them, but they might be saying, hey, not everybody can get into our program. So we are giving an opportunity to others to learn and have our degree, something. So everybody's right in their own mind. Students are saying, okay, if I'm not able to get into IIT, at least I have another chance. The colleges, the universities, the middlemen are saying, okay, we are teaching you the real world. So

Everybody's right in their own context. So I don't judge people now. It's extremely generous of you to like, try to like even empathize with the decision making going on. For me, it just seems...

brazenly, you know, cash grabbing. I mean, it's like they're getting their cake and they're eating it too. They're getting the money from these companies. Over the years, and especially with the traveling, it has forced me to understand the perspective from every side of it and see that, okay, if I am at your position, how do you make it correct? And I've realized everybody's right in their own mind. Everybody's a hero in their own head. So I always try to get into that picture as well and then see, okay, let's see how it goes forward.

Yeah, even history's biggest villains probably had some sort of justification in their mind for going down the path. Everybody is a hero in their own head. Even somebody who's writing the meanest comment right now in this just podcast, he has his own reason to write it. So I said, probably that gives you more dopamine. So probably it's okay. Well, this has been a lengthy discussion of higher education in India, and this was not planned. It's just kind of extemporaneous discussion. But-

I want to talk about your tutoring of students. You have all these prestigious university students. They've gotten into these programs, and yet they know that the pedagogy at the university and the methodology and all this is kind of dated, and it's not necessarily the best. But they know that you're an extremely experienced teacher, and they can find that out because they can watch your videos and get a taste of how you teach. And you had a bunch of them approach you

Uh, students at like an IIT university and they said, Hey, can you like teach like a seminar essentially? Like, can we give you, can we like fund you sitting down and actually teaching us? And my understanding is you're like, no, no, no. Just watch my videos. I don't do private teaching, but they were able to convince you to like essentially teach a cohort directly. Is that accurate? Yeah.

Yeah, it actually happened. So these were just like five or six students, probably six or seven, something like that. They asked me that, hey, we want to learn the subject and we want to pay you. And I said, money is not the problem. That has never been. So they just convinced me that instead of teaching us with the regular route, can you teach us like seven to eight hours? We'll be sitting with you live.

And can't we just finish this topic in just four days, five days, instead of taking a month long of it. And you just go through it and give us one day break. You teach one day, give us a break. We'll sit that. So what these students, they were, they're still within touch with me on the WhatsApp. What they did is they went into a single hostel room. That's why I'm saying peer learning experience is the most important. They all decided that we'll go up into single room and we will just have this crazy one week hackathon of learning. Okay.

We will hire the best mentors. So they approached me, hey, Tej, we want to do this program and you teach us this thing. And we'll pay you good enough. Each student will pay you one lakh rupees or something. So we are actually focused on that. I said, okay, you go ahead and do that. So five lakh is not really a big money for me. It's very, very okay-ish for me. But for a one student paying this much of amount of money, it's pretty huge. Yeah, this is equivalent to like 1,000 US dollars. So put things in perspective here.

If you were to get like a private tutor and they were really good in the United States, you might pay like $100 an hour. So like 10 hours, that would be like a lack essentially of private instruction. Now, if it's a smaller group, maybe people would do like a group class and they pay like $50 per student per hour or something like that. And that's across many different disciplines. So if you're doing piano class, if you're doing foreign language class, if you're doing ballet, any number of things you'd probably pay maybe like...

$50 to $100 per hour for a course. So it was just a five or six students. They all came in and started learning. They said, you teach us one day for six, eight hours, then give us next day as a break. We will self-study in the group in the hostel. Then again, next day you come up with that.

I said, I had never done this kind of a thing. Let's experiment it. How does it go? And I was so impressed by their progress because they were constantly in a group. Everybody was teaching each other. I realized, oh man, this is so good. And by the end of the like seven days or eight days, they were able to learn so much. I was so impressed. We built a good product out of it, a good project and everything. And then out of the end of this, I gave them a hackathon, online hackathon. Now all of you, five or six of you have to build this.

And they were able to build that complex thing within like three or four days. I gave them a week, but they were able to deliver that in three or four days. And all the prize money that they paid me, I gave that back as a prize money. So here's your special reward for this. So I said, if you give this, I'll give each one of you a MacBook. Wow. Yeah. I mean, that was like about balance out to how much they paid you. That's cool. Yeah.

I didn't do it for the money because money part, we have a good startup and we got an exit. So that is not a problem for us. But I was able to see this experiment that, okay. And the most important part is since they have paid me and they acknowledge it, that we have bought previous courses in the past as well from other. But the moment we paid you $1,000 or equivalent of it, we got serious. Now that our money is online, we have to get every bit out of you. Yeah.

We have to be very...

very aggressive with you that, hey, teach us more. We want to, if I have any question, I want to ping you and want to learn truly from you. So that seven days, I think, was one of my also very rigorous, don't want to do it again. The students were like pinging me at late night, two o'clock. We are stuck in this, what to do? They were figuring it out, all of that. But the good part is end of the day, I didn't charge them anything. I just gave it as a kind of a MacBook. Didn't give it literally the MacBook, just transferred the money back to their account.

So that is how it went. But it was a very good experience. I highly recommend students to do it. Even if you cannot hire other teacher, at least make a group, learn it from YouTube, pre-code camp or any other place, make a group and do a hackathon for seven days. Hey, it's a seven days learning hackathon in just one room, five crazy engineers. Let's do something. Yeah. So what you're describing is kind of like an accountability, a commitment device, essentially like an accountability to be an accountability mechanism that

Once people have skin in the game, once their money is at stake, like, wow, I paid for this. I better make good use of it. That's something Free Code Camp hasn't really been able to tap into because we're free. But a lot of people have had ideas like, oh, you should make this thing where I can... If I don't do Free Code Camp for a week, then I get penalized and I have to make a $20 donation or something like that. I'm like...

But we don't want donating to Free Code Camp to be like a negative thing. We don't want there to be a negative association. Damn, I screwed up. Now Free Code Camp has got my money. So we're still fine-tuning that. But I definitely have heard that from many people. Courses still have a negative vibe to them. Everybody will say, don't do courses. People will say, hey, do books. Read books. Books are good. They are very generous. And the moment you say, I have a paid course, oh, you are a course seller. Yeah.

Come on. Yeah, people will pay for books. They won't necessarily pay for a course. This is digital age. That is a phenomenon. The authors who have been writing books, they are now making courses. This is digital age. Yes, of course, there are bad books. Yes, of course, there are bad courses. But there are great books. There are great courses. Yeah, and I will say that there are paid books and there are free courses.

Books do, just like there are free courses. Free Cooking has published dozens of free books on different topics. Book is just a medium. It's text as opposed to video or interactive, which is kind of like our cup of tea is just like figuring out a way to get people actually fingers on the keyboard. But yeah, I 100% agree. And I can understand where you're coming from, like wanting to get people to actually read

uh, kind of put their proverbial money where their mouth is, even if it's a token amount, right? Like, uh, if it's the equivalent of like, you know what you'd spend on a night out or something, going, going to eating dinner and, and, and going to like a, a movie or something like that, right? Like, like that has weight, uh,

Um, and, and I certainly like every month, like, Oh, am I going to pay for Netflix this month? Am I actually watching it? Right. Uh, are we actually using it? Can I justify continuing to use it? And then, Oh, well I'm paying for Netflix, so I might as well sit down and watch the show with, with my kids. You know, that's just kind of like the way the mind works is once your skin is in the game, so to speak, uh, or once, once you've already paid to use it, you're more likely to use it more.

So that is a phenomenon that I want to figure out how to harness to get people to use free cocaine more, which is my real goal, not necessarily to get more donations, but to just help people learn more. Um, so please use it. Yeah.

Well, I want to talk about time management. That's something you have done a tremendous amount of work on. You create, you publish this TEDx talk a while back that I enjoyed watching. Uh, and you have a lot of strategies, a lot of observations about, you know, productivity and time management. What are some of the most profound lessons that people can learn if they're, they want to get serious about managing their time better? Oh, the moment you want to manage your time, the first thing you have to go is into the third person perspective mode, right?

A lot of time you are so much engaged and so much, as they say, the closer you are, the less you see kind of a thing. So step away a little bit. First of all, that's the first thing you have to do. You have to understand where you are engaged, what's taking more time, what's less. So step away from the things. As I say to my wife as well quite a lot, that the whole world is like a television. The moment you move your neck a little bit tiny degree, there's a new channel happening. There's a new channel out there in the world. So go sit out in the park,

have a notebook or something, remove yourself all from digitally and think about how does my work look like? How does my week look like? How does my month look like?

If you're not going to step away from it, you want to actively manage it. It's not going to work. You have to take one day off, two day off or probably three days off to actually figure out what the management looks like. You might get anxious that, oh, I'm super busy. I have to do this. You have to plug off everything and not just for a few hours, a few days, actually. So step one, do that.

And then go ahead, watch my TED talk as well. That is pretty good. All that I've summarized a lot of books that are published as of now. Not all of them are bad, but a lot of books are published as of now. They have been published in the pre-mobile era, pre-TikTok era, pre-Instagram era. They didn't face the problem of the reels.

So there is no solution of that reels in the book. The addiction of scrolling and endless crawl of Twitter or maybe Instagram, that was not there. So there is no book or no guide that can help you in that. So you have to be modern in that perspective. So what I said in my tech talk is if you cannot leave it, just master it.

So if you cannot leave the habit of scrolling it, at least be in that peer group on Twitter or in TikTok that at least you are getting some value out of it. So automatically the algorithm will push you towards watching business advices or maybe advices on how to teach better or get into the Twitter zone where

the spaces are happening where people are discussing about if you're interested in that politics or maybe if you're interested in business, go in that. So if you cannot leave it, at least master it so that you are in a better group. We all have tried to detox ourselves from the TikTok and I'll delete. Doesn't work. Doesn't work. All of our back in the YouTube game.

game instead of endless crawling at least be mindful of what you're doing it's not going to go away it's very well engineered by the people engineering team as well as psychologists so you cannot get away from the real so at least watch the good ones that's my number one advice and apart from this

uh, hit the gym. Uh, workout is the most important thing. And if you have the luxury, uh, to be in the gym at the odd places, I know a lot of you don't have it, but if you can be in the gym at the odd places so that you can work out without the headphones and nothing. So your brain will have the ability to think when you work out in the gym, you have a lot of thoughts and the thoughts will come when you're not listening to the music or not listening anything. Uh,

the whole quietness. You will find that you talk to yourself quite a lot than you think for it. So talk to yourself a lot. That is exactly the meditation for some people. For me, it's gym. I don't do active meditation. But for gym, I don't listen to any music. I'm just on my own, smiling, enjoying. Oh, this can be done. Oh, I can do this in the next boot camp. Oh, this could be a nice video for the YouTube. Talk to yourself. So these are some of the actionable advice you can have and rest are there in my talk. You can also watch that. Yeah, that

That's awesome. So I just want to recapitulate there because this is really good advice. Like even if you can't necessarily break away from using, you know, platforms like TikTok or YouTube or scrolling through, you know, Reddit or Twitter, at least take the time to like curate that experience so you're getting something positive out of it and you're not just, you know, doom scrolling essentially or just kind of like,

mindlessly kind of like flicking from one thing to the next, at least try to structure it to where if you're going to have that vice, like that it's as like that there's some positive usefulness for it. The other thing you said, go into the gym. It's interesting. You say, go to the gym without any sort of music and just be present. Just be there. Just be focused. And the thoughts are going to come out. Your inner voice is going to come out. You're going to start talking with yourself, um,

Using the gym as an alternative to meditation. I don't meditate either. Like I've tried it. It just isn't for me, but I love just going for long walks and thinking. And of course I go to the gym and usually I do listen to music from the gym because I want to pump myself up. But it works there.

Yeah. Well, uh, do you, I want to ask you one final question because we've learned so much from you during this conversation, but you mentioned earlier on that there was a time when your family was, uh,

you know, struggling to be able to pay for you to have tuition. And, and that was a time I imagine of great uncertainty for you about the future. If you could send some sort of message back to yourself, some advice, or just maybe like one sentence back to yourself that you think could help you make it through that time and maybe help accelerate your arrival at the destination you're in right now, what would it be? I, I have moved into a mindset and mentality that I,

If somebody asked me any time that, hey, can you go back in time and fix any one thing? I said, no, nothing. Is there anything advice you would like to give? I said, no, nothing. Whatever a decision I have made has moved me into a good position now. And I think any external advice or anything on top of it would have ruined the journey. It was such a beautiful journey when you had nothing in your life.

to a point where you now have a good life. You are able to help out thousands of people. I'm able to send my dad on a vacation. Any country he puts out a finger on it, I recently sent him to South Korea for just having a fun vacation. I think any external advice will ruin that whole journey. I wouldn't do it in any case. And I recommend to everyone as well. Hard times, the beauty about the time is it will pass. Whether it's a good time or bad time, it will pass.

So don't focus too much on, hey, it's a bad time for me or it's not working. It will work out. Just put in the effort. It requires probably, for some people, it requires less effort. For me, it required crazy amount of effort. Probably it requires more effort from your side. Probably 70 hours, probably 80 hours, probably 100 hours. But

with the hard work effort consistently, your time will come. Probably it will take five years, 10 years. Some people have seen the success of their startup when they were 45 of age, probably 50 of age. So they were relentless in doing that hard work. A lot of my videos still, and in the early days, nobody noticed me that Hitesh is a guy who posts YouTube videos.

Nobody cared about it. But it's been 13 years. I still have been constantly putting out effort. And even with my new channel, I didn't expect people to show up this early. But they loved it, enjoyed it. Or probably I had more experience of delivering it. But I'm not here to go any time out. Even if tomorrow's channel dies out or hacked or whatever happens, I'll create another one. I'll start another one. And I'll still post videos. So don't worry too much about advices and all that. Just move on. It will work one day.

Awesome. Well, that's great advice to end on just like the,

kind of faith that, that things will work out and, uh, that you, you can't really send anything back that won't maybe potentially alter in a negative way. How you got here, right? What if you accidentally sent something back in it? There's like this butterfly effect and Oh no, no. I'm like working as a janitor instead of working as a teacher to all these people or something like that. Right. Because, because you gave me bad advice or you get, it's not the, yeah. So I definitely can understand where you're coming from. And I just want to thank you for being such a positive, uh,

influence on so many people and being out there, putting yourself out there. I know not every single video blows up. You've got videos that have millions of views. You've got videos that have a few thousand views, uh, but you're putting your heart into every single one of them. And I, I've really enjoyed learning from you over the years learning. Uh, and again, I, I encourage everybody watching this to watch your, your, uh, time management.

TED talk, but also to check out like your gig course, your TypeScript course, check out all the many courses you're creating on your channel and just to continue learning from you because you're a great resource for the community. Thank you so much. You have been very kind in the words. Yeah. Well, everybody until next week, happy coding, happy coding.