Navigating a 20 year Career in Tech w/Sunil Pai

This content was created by taylor desseyn. This content was uploaded by taylor desseyn. This is a YouTube live/interview titled "Navigating a 20 year Career in Tech w/Sunil Pai" (published 2025-09-03). The recording is an informal, conversational livestream hosted by Taylor Desseyn with guest Sunil Pai (principal engineer at Cloudflare). They discuss Sunil's 20-year career (early embedded work, Myntra, React core, PartyKit, Cloudflare), the history and evolution of the web and developer tooling, practical career advice, anecdotes (Indian food, college), and current tech trends—especially AI and its effects on developers and tooling. Language: English; format is a hosted live discussion/interview.
Chapters
All right, we're live. All right, here we go, alerting the troops. All right, cool, awesome. All right, uh, good morning everybody, uh, afternoon where Sunil is across the ocean, across the pond. Um, uh, I excited to be back after, uh, the, the holiday yesterday. Um, and, uh, you know, it, it's always interesting. Um, so I, it's like I told Sunil backstage, like, you know Obviously the power of Twitter. I still call it Twitter, just like seeing people. And I've followed Sunir for a while and I don't know, I'm sure you posted something that prompted me to get into your DMs, but really appreciative that he said yes. And he and I were chatting backstage and I think it's gonna be a great conversation. I presented this talk 3 weeks ago in Raleigh at All Things Open meetup, and I think it's gonna be very similar to what we're gonna talk about today. Um, in that basically, hey, listen, like, the web's changing. You're not late. Sunil obviously has a ton of experience that we're gonna talk about his background. But before we dive in, please say hey. We'd love to hear from you. I know we're going live. People are— it's just now hitting the interwebs. So please say hey. We'd love to hear from you. Let us know where you're tuning in from. Um, but, uh, I want to give a kind of a few shoutouts, kind of housekeeping things as always. First off, I'm an Appwrite hero. Me and my guy Jason Torres on the Torque team are Appwrite heroes. Uh, I always say I'm the only non-technical Appwrite I'm the only non-technical hero probably on a technical heroes program. Um, I am— I'm the true definition of a personality hire. Um, so, uh, I don't know how to code, um, but I feel like Appwrite's great. You should go check them out. If you do check them out, tell them that TorqueBoy sent you. Me and Jason, like I said, are both heroes. And listen, we got a lot going on in the Torque community. Um, we have so much. We have lives, we have— we have public lives, we have Discord lives, we have— we're helping people write resumes, we're helping people do interviews, we're doing all this stuff. If you want job search, if you want to source job search Tips, jump in our Discord. But to get all that information, go to torque.community. We'd love to have you. Also too, if you want to be— if you want to know when I go live, shoot me a text, 615-235-5650. And of course, would love to have you there. I just shot out a text to my community that I'm going live with Sunil, so we'd love to have you. All right, Sunil, let's give the people what they want, man. Who are you? What do you do now? And we'll dive into, uh, your background.
Well, personality hire to personality hire.
Yeah, it's a thing. It's a thing, man.
100%. I mean, uh, engineers could stand to have more social skills, of course, in general.
Yeah, of course.
Uh, first of all, Taylor, thank you so much for having me on. Followed you for such a long time now, and I'm, uh, so happy to be here. Um, uh, my name is Sunil Pai. I'm currently a principal engineer at Cloudflare. I do a bunch of stuff around AI agents. I've done a bunch of stuff around real-time at Cloudflare. I helped rewrite their developer CLI from Rust to TypeScript, which is the opposite direction that a bunch of people—
I was about to say, I'm not technical, but that does seem like a very different thing.
It worked. It worked. It saved the product. So, for the 4 Rust engineers out there who are mad at me, send it to my DMs. So yeah, I am, that's what I do. That's what I currently do. I have a lot more behind me, hopefully a lot more in front of me as well, but that's what I currently do.
Yes, I love that.
I love that.
Well, man, you know, kind of really what the titles say is navigating a 20-year career in tech, right? I mean, there's, you know, we were chatting backstage, a lot of the waves, a lot of the trends, Man, just give the folks— and we have people starting to tune in now. Chris, good to see you. Naz, Jalen, it's great to see you this morning. Again, please continue to say hey as this hits the interwebs. But man, give a background. Where did you start? Talk to me kind of through your career at a high level to set the stage of the conversation today.
I'll try to keep it short because I'm from India. I'm from West Coast India, which is the best coast of India.
I love it.
I love it. Uh, which is to say, like, I have opinions on seafood and curry that are long and varied.
Wait, wait, wait. I wanna hear that. I wanna hear that. Right? 'Cause I, I have not been to India yet. It's on a bucket list for sure. I have a lot of friends out there. Never been to London, but I hear that's like, I hear England is like Indian food central.
Oh yeah.
Yeah.
100%. I, I don't, in fact, like people think that London is where the great Indian food is. And for context, I moved to Newcastle just in the north of England about 9 months ago. And the Indian food here in a town with 160,000 people is like S tier.
Is it really?
I was surprised. I didn't expect it. Like we went to like 7 different ones just to try it out and each one was amazing. They have a legit Indian style buffet restaurant here.
Oh dude, I could hurt myself on Indian buffet.
Bro, yeah, it's like £25 and you sit down for like 2 and a half hours.
Oh, sounds like my jam. If I'm ever in Newcastle, we're going.
It's crazy. It's like 40 dishes. They have a— and the thing about an Indian buffet is they have a little bit of everything. They'll have like a live pasta counter, a little Chinese, uh, 12 types of desserts, 40 other dishes, 4 types of biryanis. I'm like, who is this for? Why is this place here? But it's packed. Like, all the students, of course, love it. It's a uni town as well. Anyway, uh, I, uh, I'm from India, and I've done a bunch of things along the way. I was explaining to you that I got into a really good uni for computer science, and then I proceeded to discover— I had my first serious girlfriend. I joined the college band. I discovered substance abuse. So it took me about 7 years to do a 4-year degree.
Okay. There you go.
And I've now spent the last 20 years just making up for it, just to get the chip off my shoulder. about just to get the chip off my shoulder that I can do as well as my roommates. And suddenly it's been working out pretty well so far. Don't get me wrong. I know some people from my batch who have now spent 15 years in NVIDIA and I only found out like 3 months ago. I'm like, oh, buddy, you have like money, money. Yeah, yeah, yeah. Anyway, so people might know me from a bunch of things. In India, I used to— I was a senior architect at a place called Myntra.com. which was an e-commerce site, but it was also like the hotbed of a bunch of JavaScript stuff that happened. You might know me— you might be mad at me for making a CSS-in-JS library. It was called Glamour. At one point, it was everywhere. I think Office 365 used it for the longest time.
Seriously?
Yeah, for sure.
That's awesome.
And in the most unexpected places all over. You might know me for my 1 year that I spent on the React Core team. Again, as the personality hire. That team could use a little more personality. Hi to Ricky. I love him. He's my best friend. And he's probably— yeah, he's—
I love it.
I love it. You might know me from doing my time at Cloudflare where I've otherwise written their developer CLI and I now do their AI agents work. A bunch of people might know me from the startup that I built called PartyKit, which is a real-time multiplayer stuff. Hey, Riley, how's it going? Hey, I know Riley. I'm mad about the work I did on that.
Everybody's mad.
Yeah, yeah. I introduced the testing APIs, which pissed off a bunch of people. Anyway, so I've done a bunch of things. But the thing I think the theme around it is, and like I was telling you, I think between 2014 and 2018. See, in 2013, React launched and everybody got mad. Then they were like, wait, there's something special here. And then between 2014 and 2018, which I consider as the golden years of the React community, not necessarily of React. I think React is still— I mean, I use React. Only React? Well, I'm wearing the damn React t-shirt.
I saw that. I saw that. Had to rep it. Had to rep it.
I love it. That was not planned. Well, I saw it this morning. Anyway, and 2014 to 2018 is when a whole bunch of names in the community just came out of the woodwork.
Sure.
And that was, I mean, those nice ZURP days when people would send you to conferences all across the world. So, we had a lot of friends that we had a circuit of meeting people. I think that was interesting for me and it helped me get my name out there, the work I'm doing, the things I care about. And I made lifelong friends. So one of these things either made you mad or helped you out at work. And I assume that's how some people know me. And for those who don't, yes, it's been a varied and weird career. I think I was telling it to you. I think I've done all the major food groups. I worked in startups, big companies, design firms, worked for the government.
Wow.
My first, first job was writing embedded software for trains.
There you go, there you go, there you go.
Uh, it's, it's been a lot. Just let— like, so it's— now I just consider myself like the JavaScript guy. Like, it's running everywhere, so I'm, I'm usually around. Does that help? That's the short—
No, I love it. That's great. People, people, uh, people definitely resonate with the Indian food, uh, conversation. I want to go back to that real fast, and then, and then I'll hit you up some questions. What's your favorite What's your favorite dish, man? Like, like, like, what's, what's, what's, what's like 1, 2 on the list?
Easy. Number 1 is the chicken curry that I make. And the reason that I think it's good is, um, I have a recipe that I've modified over the few years, last few years. Uh, it's made with coconut milk and it's not spicy. Okay, like the base—
I'm a wimp. I'm a wimp.
No, no, no, that's why you like this. This isn't— it isn't spicy, but it's like super flavorful. And I love the recipe because it's so forgiving. Like, on a day you can— I drink when I make this curry.
It's like, you know, today, you know, it's good.
I'm going heavy on the garlic. I'm doing this. Yeah, I like listening to like Metallica and shit while I'm cooking this. I love it. My chicken curry is easily like my favorite Indian food. I've cooked and made it for friends. If you're around me, I will make that shit for you.
How long we marinating the chicken for?
No, no. This is a non-marination one. Okay.
Non-marination. Okay.
One pot. Yeah, yeah, yeah. I do this because it's like a one pot thing. Okay. My other favorite Indian food is coastal seafood, specifically around the Konkan Coast, which I want to say is like between Bombay and the very bottom. Okay. The reason that if people have ever been gone partying to Goa, which is like. the party capital of India. Yeah, they're very intense flavors. They're like 40 types of chilies, all with a different—
very—
there's no one type of chili, like it's all types and mixed with 600 types of spices. And you're going to Goa. The reason you go to Goa is the alcohol is tax-free, which makes it super, super cheap. And all the food is just shacks on the beach. Like you will start at like 4:30 in the afternoon and you'll end at 2 in the morning just eating and drinking down the thing.
That's amazing.
the experience. That is like what food does for you, right? Like, yes, I've done fine dining, I've done all of that, but I'm not joking when I say my top is the chicken curry I make because it usually means I'm making it for myself and other people and we sit around and drink. And the second one is walking down the beach eating seafood in Goa. Fuck it.
I love it. I love it. There you go. There you go. Um, all right, so I, I want to talk about just the internet. I mean, you know, you and I were chatting backstage. I mean, I've been recruiting the web since 2011. I've seen a lot of changes. You've been actually hands-on through the changes. Let's just talk super high level. We're going through a big change right now. We're going through another big change with AI, right? And of course, some people may roll their eyes. Oh, great, another AI talk. But it's here. We talk about it. But let's just talk about trends as a whole. What are your thoughts on them? What people should be aware of? Just give me kind of some Sunil hot takes on this Tuesday morning.
Okay, I'll just go for it. Feel free to interrupt.
Yeah, highly hot take. Yeah, unfiltered.
The first thing you need to know, and you need to remember this every morning when you wake up, because I know everyone reaches for their phone the moment they wake up, because I do. You have to understand that social media is adversarial.
Okay?
I mean this in a very specific way. You are hearing not just only the good things that happen to people, you're hearing from people who want you to think that things are going only great for them. So I mean this in terms of companies, CEOs, people shilling you their products, trying to tell you what is important. You'll miss— like AI thread boys on Twitter, right? Google just launched a new model yesterday. Everything has changed. And if you don't read and retweet my thread right now, You're donezo, my guy. You're donezo. So I know for a fact that everyone is in a state of confusion. I know that everyone feels like there's this daily need to keep up with what's going on. And I'm here to tell you that that's all bullshit. Don't get me wrong. I'm not trying to be a boomer about it and say invest in fundamentals, even though you should. But like I was telling you earlier, This has happened before multiple times. And for 2 of them, I was around.
That's crazy.
Right? So, I got out of uni around 2007, 2006, 2007. So, between 2005 and 2008, 2009, when JavaScript was just starting to become popular, there used to be a website called ajaxian.com. Now, any of the 35-plus, 40-year-olds in the chat might know what I'm talking about. To those 3 people, hey, what's up? How's it going? Uh, take an ibuprofen for your back, I guess.
I took ibuprofen for my back this morning, Sunil. Not sure.
Like, yeah, at this point everything's kind of broken anyway.
Yeah.
Okay. Uh, can I have my drugs to go? I need to be in bed by 10:30.
Seriously, seriously, I'll throw a party, but everyone's gotta leave by 9 PM.
Um, anyway, so, and Ajaxion was run by Dion Almere and Ben Galbraith. Uh, amazing, amazing people. And they did this. That was the website you would go to on a daily basis just to know what new things were happening. It wasn't even about hype. No one was paying them. I think they occasionally had ads and stuff. But I remember the feeling. To be clear, I want to give context to the people here. I was doing web development back when dev tools weren't a thing. Yeah, I am pretty— I'm also handsome as hell. You can say both things at the same time.
10 out of 10. Can confirm.
I'm funny and hot.
I wish I had that facial It's growing in.
I mean, it's a little patchy. There are guys who do like— anyway.
Anyway, I digress.
So, it was amazing because that was the time when jQuery was just starting. jQuery was 2006, I think. Somewhere around there. Prototype, Scriptaculous, Dojo, MooTools. MooTools, which is the progenitor of React. Like half the React team were from the MooTools From the MooTools group. Tom Okino, Sebastian Markboe, Chris Nakazawa, so many people. Hell, even Guillermo. Wasn't Guillermo doing stuff with MooTools? I forget. I don't remember. Anyway, I remember very— I was mentioning that there were no dev tools. You'd have to debug your code by using alerts. Like you'd have to write alert this value as a string that would pop up. And there would be bugs that showed up only when you did alerts. It was all a mess. And Joe Hewitt, he launched Firebug 0.1 in Firefox. It was the first way to do— to debug your code and do console.logs. And I remember that feeling like, oh, my God, this changes everything. Because until then, and maybe even now, JavaScript was very much considered a toy language. Point being, I'm older than most of you here, even though I have the great hair and the facial hair. I remember having a very strong feeling that I'm late to the game.
Even then?
In 2008. Okay?
Yep.
Geez, that's a while ago.
It's a while ago, dude. It's a while ago.
Oh, my God. You have to remind me to talk about that thing about Going from being the youngest person in the room to being the oldest person in the room. Happens every day. We'll talk about it in a minute. Anyway, so I remember thinking I was done for. What am I even doing in this space? All the big names, all the problems are solved. But I just, like, went with it because, like, it's such a— you're so lucky if you have a job in frontend development that you get paid for. Because It's so visually satisfying. The feedback cycles are so fast. It's always interesting because it's not just about algorithms, but about like human interactions and stuff. I remember feeling like that in 2008. The second time that I felt this, and this is a feeling that we both share, was in 2011, 2012. Angular was the name of the game.
Yep.
We were like, oh, man. Angular kind of particularly sucked on Indian networks at the time.
Oh, really?
Yeah, India solved their internet problem, but back then an Angular app would barely load. It was really bad. And I remember again thinking, well, it's kind of boring, but it's a job and I enjoy it. And I enjoy writing code and solving business problems. The reason that we all got into React wasn't because we thought it would get big. In fact, we thought it wouldn't. It was all the rebels and the weirdos who got into React. Like, what if you could render CSS in your JS? What if you made a server component? XML in JavaScript is so weird. And that is what is appealing about it. Every conference, every meetup was just about experimentation. And things that people had learned that could do— hey, Angular has actually gotten pretty good after Sarah Drasner got there, by the way. Just giving you a heads up. Sarah Drasner is a legend. And I think Angular is not bad anymore. I said anymore like it was bad. You know what? Yeah. Fuck that, dude. No, no, I like it. Anyway, I remember feeling that way then. And for us, React was an escape. Not into something bigger, but— just to keep things interesting. And I know for a fact everyone thinks about that about AI programming, like, today.
Yes.
And oh, my God, you are so, so, so wrong. You're so wrong. And especially just pattern matching on what's going on. It is so strange. And I guess that's the context you wanted me to bring, right? Like, there is all— Which is why I started by saying social media is adversarial. The reason you feel this way is not only because you think people around you are saying it, but you think everybody else is also feeling like it's done. And I'm here to tell you, it's not even close.
Yeah.
We are, like, less than— it's not even, like, step 1 of a 100-step journey. There are, like, thousands of things. Like, this is going to be Whether AI or not, I promise you that software development is an unsolved science and it's going to take another 20, 30 years. I don't care whether AI writes most of the code, whatever it is, it's just going to— if you've been keeping an eye on these things, you'll notice that everybody changes their mind every 6 weeks anyway. Everybody. Yeah.
So let me ask you this, man. I mean, I think obviously in your circles you've heard it, in my circles I've heard it. I think there's a lot of eye roll, a lot of groaning right now with where we're at with AI, right? You know, there's, again, like I said, social media pundits, it's a mess, blah, blah, blah. What, what would be your biggest encouragement or words, words of wisdom to folks right now who are maybe scared to get in on it, think it's a fad? I mean, I tweeted something the other day, and I had like 3 or 4 comments like, oh, it's a fad. It's, it's like everything else. What are your thoughts on that?
First of all, the tech is really old and good. Okay. It's only a fad if it's, well, if it's new, but Transformers, forget about the fact that Transformers paper only came out, I want to say 7 or 8 years ago. The AI space has been alive for a good 60, 70 years. 60 years.
Explain that. Explain that to me as a non-technical guy. Why has the AI space been alive that long?
So I mean, it's kind— okay, so if you go all the way back, uh, 20 BC. Jesus, not, not even that much. Uh, but I'm thinking more specifically—
There you go. We're pull— oh, I was about to say pulling books off the shelf.
No, no, no. Alan Turing at Bletchley Park wins the war against the Nazis with the codebreakers. If you've seen the movie Imitation Game, It's a great movie.
Great movie.
Bletchley Park, by the way, in the UK, 1 hour north of London. It's my— this is my temple. I go there like twice a year. I'm not joking. Okay. Okay. I recommend it. Bletchley Park is the OG. It's crazy. You get to walk around the campus and they tell you stories of all the things. Anyway, point being, particularly after the war got over, but I think they were thinking about— von Neumann machines, Babbage, they were all thinking about it even before the war, even before we had transistors or anything. There's always been the desire to create intelligence out of nothing.
Correct.
I don't know how many people here are aware of the programming language Lisp, for example. Lisp is a programming language you can learn in about 20 minutes. It's wonderful. JavaScript was invented by a Lisp programmer at— it was called It was called Mozilla at the time. Yep. Ben Eich. What's the name of the guy who invented JavaScript? Shit, why am I forgetting this? Invented JavaScript. Brendan Eich. Brendan Eich.
Brendan Eich.
Yeah. He was a Lisp programmer. And his research when he was working in Mozilla was to bring Lisp to the browser. And they entered into a marketing deal with Sun Microsystems where Sun was going to announce Java applets in the browser. And they needed a lightweight scripting language.
Interesting.
Did not know that. Which is why it's called JavaScript. There's no other reason. And Brendan Eich basically gave a C/Java-like syntax to Lisp. Which is why if you know JavaScript, it is so easy to learn Lisp, by the way. It's crazy. You're like, oh, they just copied JavaScript.
You heard it here first, folks. Lisp is going to be hot.
I'm just kidding. People have been saying that about Lisp since the '60s, and it's not happened. Lisp is very much the we invented it first. We had lambdas and we had— anyway, point being, it was invented by— I want to say his name is McCarthy. McCarthy, as a language for computers to generate for artificial intelligence. There's also a language called Prolog, which is a logic programming language, again, written in the desire to give machines thought. And the first thing you do is you almost give them a language to be able to speak. Point being, AI has been— people have been working on this space for decades now. And the idea of large language models being the solution for it is crazy. Uh, point being, to go back to your original question, it's not really a fad because the tech is very solid. It's very expensive because it says, let's— it's called the scaling law, which is, you know, chuck, uh, as much hard— the more hardware you throw at this, the smarter it becomes. That's basically where Sure. It comes from. And it's only until Ilya was like, yeah, let's throw like a billion worth, uh, billion dollars worth of GPUs at this and see how smart it can become. And then OpenAI kicked off the whole thing. Yeah, there's lots to go there. Point being, the tech is actually very good. The tech is unfamiliar for those of us who have been in the space for a while. By the way, it's why you'll see 19-year-olds and 18-year-olds get so much better at using AI tools than, well, people like me. Simply because they don't have preconceived notions of what programming should be like.
Interesting.
100%. In fact, a 19-year-old intern taught me how to use Cursor. I was like, it's not really a good tool. Then I was pairing with him and I saw him doing these things.
I was like, okay.
Okay. I need to do this. The tech is good. It's unfamiliar because there are new patterns. And that is where the opportunity for everyone, hopefully, who's listening to this should— yes, AI is a broad term. I'm talking about LLMs because that's the flavor du jour right now.
Sure, sure.
This is the time, just like JavaScript in the late 2000s, React in the mid-2010s, uh, where you are—
you can even say cloud engineering too. You could throw that in there because that was a trend as well.
I didn't even realize AWS invented the term Lambdas for functions, by the way. I was like, I didn't know that either. Actually a really good name.
I'm learning a lot from you. You're like a walking encyclopedia, dude. Damn.
I've been— you've been, you've been around a lot. Yeah, I'm like—
Those gray hairs, those gray hairs didn't just appear.
That's right, the salt pepper is showing recently, but otherwise No, no, like, I've been enjoying it because I started getting white hairs in the beard.
Listen, the gray is in, man. Gray is in.
I'm into it. It's working for me. Don't get me wrong, the rest of my body is just coins in a meat purse at this point. I'm just— I'm gonna go lie down after this podcast, but the rest of it is pretty decent. Your job as a technologist is not to call for a winner. Or not even to decide who the winner is. You can apply this to anything. I applied the same thought to crypto. I was like, okay, my job right now is not to decide whether it's good or bad. Or you know what? While I'm working, when I'm sitting in front of a laptop, that's not my job. My job is to become familiar. You are a technologist.
Yes.
You can go down to levels Normal person on the road usually wouldn't. You have to form very detailed opinions about what these things can do, can't do, where it's going. Like, can you look back at the 6 months and say, okay, fine, it was able to do this? An agent could run for 5 seconds. Now it can run for about a minute, which means probably in another 6 months it'll be able to run for 20 minutes without like falling apart and context and all of that.
Sure.
Your job is to become— well, you have to become a subject matter expertise at everything. I think a lot of people confuse a job with the work. I keep telling this to people.
That's interesting.
I love work. Okay? Work. And I don't know whether it's because, like I said, I have a chip on my shoulder because of university or because I'm an immigrant and work gives me validation and brought me Took me all the way from a tiny uni town in India all the way to here.
Sure.
Well, back to another uni town, but still pretty nice places along the way. You have to— the work is good. Your job is so nice. You get to like sit and drink little matcha teas and all of that. But the job is very different. The job is a responsibility. And you should never make the mistake of like confusing the two. Your job is to solve business problems. It'll be some version of this statement. You're usually working, uh, either you're—
you—
it's not even like you're an independent person working only on your problems. The entire point of technology is scale. So you're trying to solve other people's problems. You're trying to make it accessible to them. You're trying to make it cheap for them. Uh, in fact, you're trying to protect your users against themselves. I keep thinking about this, that if I was thinking about this back during my e-commerce days. If humans were perfect robots, then every e-commerce website would simply be a database that you open to the public. And you say, yeah, write SQL queries against this. Go for it. Cart, checkout, inventory, just write SQL.
Yeah.
But the truth is that users are not perfect robots. They're smelly human beings. And you want to not just, well, protect them against themselves. Like, oh, you can't actually add 600 of one item into a shopping cart. But you want them to feel good about it. Hell, you're a JavaScript programmer. After one point, it's even less about the data structures and algorithms, which we are famously bad as a group at. But about building human-computer interaction systems.
Correct.
Which sometimes can be measured by A/B testing systems, but a lot of it is vibes. You're like, oh, that looks pretty. That looks nice. The animation goes on for too long. Let's let them not have them scroll here. The red on yellow doesn't have good contrast. Let's fix these things. You develop those opinions by doing it 100 times.
Yes.
Right? And you have to be able to explain how you feel about it. So, look, I have my own bets about AI. And I'm not here to pitch them to you. Though you should build AI agents at cloudflare.com.
There you go. If Cloudflare wants to sponsor pod, let me know. Um, but yeah, no, I mean, but, but I want to hear your— where, where do you see this going, man? I mean, again, hot take central, right?
Yeah, I, uh, look, I don't know if the scaling— if the scaling intelligence with current systems means that it's going to be 10x smarter next year. All signs point to no, by the way. Right now we are in that winter of intelligence is what they're calling it.
Yeah. Aaron Francis's tweet the other day about ChatGPT-5 coming out, he goes, I'm actually thrilled that it's not great.
Right. It's a kind of plateauing.
Yeah.
Even look, let's say there were no model releases for the next 5 years.
Correct.
There is still 15 years worth of improvements to existing software systems you have. This is my belief. Okay? Every input box should have autocomplete. Every textarea should have autocomplete. There are no open source components that let you do this right now.
That's great.
You can look. You can look. Hell, people talk about how voice agents are the future of AI. There is no open source React hook or component that I can drop and make this work out of the box right now. That's kind of weird, right? I looked, I've even tweeted about this a couple of times. If you want open source glory today, build that voice component. It doesn't even have to do with LLMs, but you can wire it up to Whisper.
Sure.
You will become an open source legend today if you build that component. There's so much work to be done. My point being, even if nothing changed with intelligence, we have at least 5, if not 15 years worth of improvements to take. Also, It hasn't reached the normal person yet. Like, AI is not out there in the world. My dad started using ChatGPT. A lot of people in India use ChatGPT, by the way. He loves it, but he doesn't know the half of it. The analogy I have is when Google came out. Google came out 2001, I want to say. So by 2008— actually, you know what, way past that. 2007. There are 2 types of Googlers. There are ones who know how to use Google. So you'll know how to use keywords. You know how Google works. You're like, oh, if I get the keywords right, it'll find the page.
Sure.
And then there were people like my dad back in the day— he's actually pretty hardcore now— who would type questions into the Google input box. Hey, can you take me to, um, uh, the news about so-and-so? Like, it would be phrased as English questions, which You and I know that doesn't matter. In fact, you know that Google is stripping out all those words just to look for the keywords and then do a search on it.
Right.
Actually, it's weird because now we have AI that actually uses that context. It's going the other way around. So LLM systems actually do work better when you talk to it in English. I digress. But like this normal— like nobody is using AI out there to automatically Answer emails, let's say, the boring ones to clear out. You have Gmail maybe applying spam filters, but there's so many things that get past it. I think Gmail's been getting worse at catching spam anyway. You don't have— Apple tried and they failed miserably with Apple Intelligence.
Yes.
They tried so hard to make your phone smart and Everybody hated it and it was pretty bad.
Like, you know who hates it the most? Tejas. He talks about it all the time.
I bet Tejas hates it.
He talks about it all the time.
Hey, how's it going, Tejas? If you see this in the future, what up? Uh, uh, see, Tejas is brown excellence. Tejas is someone I look up to. He's a G. He's a G.
He is.
He is. Um, your job, your job as a technologist right now is definitely not to be an absolutist about this. You know what? You're looking for hot takes. I'll say the thing that everyone is afraid to say. If you're an absolutist about this to the point where you're like, I don't want to use AI, it's unethical, I don't want to work with people who want to use AI, I don't want to touch AI, you're making yourself unemployable today. Like, you're done for. Hey, my guy Casey B. What's up?
Good to see you, brother.
Again, I'm not saying AI is good or bad. You can absolutely have those opinions, but you work in an industry, trust me, I have problems with a lot of systems in the world today which are just absolutely unfair and biased towards billionaires and corporations and armies and empires. Take it from an Indian immigrant who's traveled the world. Yeah. Okay. That is not your job. Your job is not to have an absolutist view about tech, which is not done changing. Like if you made this decision about React in 2013, you would've made yourself unemployable.
Well, well, I'm gonna chime in. I know devs who have.
How, how did they manage? What did they do?
They've had problems finding jobs ever since. And then they wonder why they can't find a job.
I feel bad for them because they're like, I have 12 years of experience and I find it hard to find a job. And then you go into their social media and their posts and you're like—
And they're just bitching and complaining. It's just like, dog, what are we doing?
My guy, uh, what's— again, what's worse is they will have 400 people in their replies saying, yeah, bro, this is so unfair. How can people treat you this way? They will applaud them into unemployability, right? And I'm like, this is embarrassing. Like, you're seeing this thing happen in slow motion. This person doesn't have friends. Take it from me, okay? Like, being an absolutist in this industry that changes every 3 to 6 months is a death knell. You're kind of done for.
So what do people do? Like what do people do right now? And, and Ken, Kenzie does, Sunil writes down to earth, realist, practical. Makes bet. Yeah, 100%. So, so, so what do we, what do we do? What do devs do? Do we, do we, do we, do we all learn Cursor and we all can, we're all an AI expert now? Like, like, like what, what does it look like?
It's not a destination, it's about the daily routine. So the way I think about it is, love that, is, uh, uh, everyone likes saying this about the, California Gold Rush, right? They're like, oh, it was a great time to be selling picks and shovels. That's not true anymore. You know what a great time it is to be buying picks and shovels? Like every single AI company out there is losing money on their products right now. Which is to say your investment in this can be zero. But like I give myself about $30 a month to spend on Okay, so I spent $20— or no, I spent $40 and on only 2 products. Correct myself, $46. I'll tell you what I spent on. I spent $20— Do you have a call to his bank account right now? I was just thinking, I was just thinking. Uh, $20 a month on Cursor, uh, $20 a month on ChatGPT Plus, and $6 a month on Granola. Granola I love because—
Love Granola.
Such a good product, such a good product. Also, I know the team. And it's $6 because they very foolishly did a sale in December where they're like, oh, it's going to be $6 for life. I'm like, done. You're going to regret this.
Done, dusted.
Such a bad move on your part. But I did it because I— you can do it for a month and you can give up. There are other tools that are cheaper, better. It doesn't matter. You spend way more on I don't know, coffee and beer or whatever it is. The job is not to decide whether Cursor is the tool that you will use. You might decide that, don't get me wrong. Your job is to develop a routine on a daily basis, if not like weekly, if not daily, but go for daily where you're learning. You're learning about what these things are doing, what new patterns are popping up. That's what I meant about social media. Even though it's adversarial like this, a lot of alpha is still on Twitter. Taylor, it drives me mad, bro. I have to get my AI news from the most racist website on the planet. It's so weird, bro. I try so hard. On the one hand, and they don't even tell you who liked the tweets anymore, right? And I'm like, oh my God, somebody I'm following. Just like this absolutely horrible tweet. But like 2 tweets down, I learned about OpenCode's SDK launch, which they just had yesterday. Looks amazing. And you're like, wait, OpenCode did an SDK launch? Cursor is looking to do an SDK. Cloud Code is doing an SDK. Now I know people on these teams and they do it if their users or customers are asking for it. Which means more people are building tools with this. Immediately that's a signal for me to put aside 20, 40 minutes and, um, try to build something shitty with the SDK. I need to know what this is capable of, what it's not capable of.
Yeah, it's— it— so I, so I grew up sailing, which is, you know, kind of ridiculous to say out loud, but one of the things that— one of the things— so I, I raced competitively in college on, on sailing teams, right? Probably the widest thing I say. And so basically, for me, one of the things when you sail is you look at where the wind is going to be. Yeah. And you look at the way the wind goes across the water. Prime example, me and my dad, we put a, we put a, a B team together. We raced in a, in a regatta. There was 35 boats. It was a national qualifier, and we were pretty much dead last every race, right? Because it's the B team. It was my dad. It was, it was, it was, it was, it was a crazy Crazy weekend. The last race, the 10th race of the weekend, we saw the wind take a turn that very few people saw, and we finished 10th out of 35 boats.
Okay. Okay.
And I think in this case, you're looking at, you're looking at data points and where the wind potentially is gonna go here. And you gotta, you gotta get familiar with watching the wind. And I think that's very similar in tech right now where You see one thing, you see it now. It's like, okay, the wind may be going this way. Let me try it to get that experience under your belt.
I love that you brought up sailing because I wanna say roughly 2 or 3 weeks ago, I don't know if it came up on Reddit or something else. Sailing is fucking cool because, A, I love sports. I mean, I like watching sports, especially with friends. That's my thing. I don't play a lot of sports, but watching sports, sports movies, amazing.
Yeah.
Sailing, you need to be on the move constantly against an invisible force.
Correct.
And the best thing about sailing is if you do it right, you can sail into the wind.
Correct.
Bro, that's the craziest shit I've ever heard of. Totally unintuitive until you look into the science. You're like, oh, shit, that makes sense, right? And you kind of felt it when you're like riding a cycle against the wind and you're like, this is how I adjust constantly. Sailing is a bonkers good analogy for, like, how to deal with systems going on right now. When you think everything is against you, you can actually, like, make moves that— again, because social media seems like everybody's doing well, you would be surprised how if you put in a little bit of effort, a little bit of— by the way, I don't have any children. I'm not having any children. And I know a lot of people have told me, well, you get to do this because you don't have your time to spend on these things. There are so many people I look up to with families who are doing way more than I am. Kenzie Dodds is one of them.
Kenzie Dodds is an animal.
Is number 5 or number 6 on the way? No, I want to say number 6. I think like—
Yeah, we got number 3 on the way.
Right? You figure it out. One of my favorite founders is Steve Ruiz, TL;DRaw. He has a wonderful daughter. He tells me the thing that happened to him because of his daughter is that he makes sure that his 9 to 5 is super effective. Because past 5 PM, it's his daughter's time. He doesn't get to do anything. And if you're telling me TL;DR team is not shipping and learning new things and finding the barriers of human-computer interaction, bro, sorry, cope. Hard cope.
Yeah.
So, no, I keep going back to this. Your job is to learn. And what a wonderful job that is. How lucky you are, how, how, uh, how blessed your life is that you get to do this, and you're complaining?
Yep. Sorry.
Let's go to—
let's go to Nas's question here. How should a beginner approach this whole AI thing?
Well, the nice thing about being a beginner in AI, of course, is you actually get to sketch out your own journey right now. If you have—
And for the record, most people are beginners.
Most people are beginners.
Sunil, very senior. I mean, maybe you're not a little bit more than a beginner, but you were a beginner 6 months ago.
You know how I became tech lead of AI agents at Cloudflare? It's because in October last year, I had no idea what it was and I started writing internal blog posts like, hey, I just learned this new thing. This looks cool. Then I started writing external blog posts on my personal blog, sunilpai.dev.
This was literally in my talk. This was literally my talk tip 3 weeks ago. I'm so glad you're proud.
I just like, hey, I learned something new. The act of writing is also like learning. You're like, oh, like you'll be writing something, you're like, wait, that's stupid. That's not what it is. Let me write it again. I'm like, okay, now I know what I actually understood. Um, you can share with people. Don't, don't do it don't do this learning imagining that you're going to become a famous tech YouTuber, or like, do it for an audience of one. Do it like for yourself. Develop a little bit of a routine, like the discipline of doing it every day, every 3 days, every week, whatever it is. Pick, pick your thing. But the one thing, you know what, I have another hot take here. A lot of older people, and when I say older people, I mean people in my age group.
Like me and you. Yeah, yeah. I think you're only a few years older than me.
Uh, they will tell you about how important the fundamentals is. You better learn about closures and arguments array before you do all these cool AI things. Don't learn React before you know JavaScript really well. I'm here to tell you that's all bullshit. In fact, the only thing that has worked for me and so many, so many other people I know is being able to build something shitty and then making it better. And you have a magical friend in your computer now who will help you do that. So, the mistake you could make, and a lot of people make this mistake, is not reading the code that the LLM has written for you.
Sure.
It just wrote it for you. Sit down, read it, have a good time. And that's how you learn. You have a magical— when I was learning JavaScript, actually, I say learning JavaScript, but by the way, this is my reputation at Cloudflare right now, is I will bother you with questions, the most basic questions. It would be embarrassing.
Get out of town.
I'm not joking. I don't know shit about cloud computing, distributed systems. I bother the creator of Cloudflare Workers twice a week. I'm like, please explain this extremely basic thing to me. He's a very, very nice guy. He's like, okay, fine, I'll tell you. Then I get to take credit for all of his work on Twitter. So, win-win. So, man, like what a wonderful, wonderful job and life I have that I can be learning the entire time. I get to ask questions. Now, a lot of people get annoyed at— might get annoyed if you ask them for too much help. You're not doing it yourself. Well, magically in 2025, you have a friend in your computer who's answering all your questions without doubt. Now, some of those— it's a very helpful intern. Some of those answers might be complete bullshit. It'll make things up and it'll gas you up. Like, it'll gas you up. Like, What a wonderful question, Taylor. You are a mensch and a god among human beings for asking me about TypeScript types. Wow. So, get rid of that stuff. Put it in your prompt saying calm the hell down. We're British. We don't do this stuff. And just—
Be curious.
It's about the routine. That's just it. So many people want to be done with the learning. They're like, oh, if I can't be done with the learning, I'll never be good at something. But you will never be done. And the earlier you realize that and the sooner that you pivot to finding a routine instead of a goal, you will do really well.
Perfect.
So if I was starting programming today, I would Google what the most popular IDE is. Maybe I'll use Copilot. Maybe I'll use the free version. I don't care. I will just tell myself, oh, and I will try to copy things I see online.
Yes.
This is how I've always done it. Hey, _js looks cool. Maybe I should try building my own version. Hey, React looks cool. Maybe I should try building my own version. Hey, this website looks interesting. What if I try to build just the frontend HTML of it by myself? It will be really bad and the code will be bad. Nobody might ever see it. It doesn't matter because you do this 100 times. And the thing, if there's one takeaway you take away from this, like I said, like social media is adversarial. You will find yourself in the 1% of people, not even above average. You will be surprised at how few people give a shit out there. Giving even a little bit of a shit, like just a little bit of reading, a little bit of experimentation, Just trying to be better on a daily basis, you will magically get way better. By the way, this is— I'm now 41, Taylor. I'm turning 41.
Okay, I'm 36. So you're a little older than me.
Yeah, yeah. I am, I am. I buy CDs. I have a little CD player over there. Like, those are filled with CDs, by the way.
Oh, you're that old boomer, elder millennial walking around. on a Walkman or something.
I like Shappell Roan and Sabrina Carpenter, but I have no idea who they are. I know the 3 songs each that they have. No idea. Like I said, I work with interns who haven't seen The Matrix. And I'm like, if you haven't seen The Matrix, why do you want to program? I thought that's why everyone— anyway. Oh, shit. I kind of forgot the point I was making. My point is people look up to me now for opinions. Hey, how should I do this? What should I do here? And part of me is like, well, that's a question I've never heard. You want to figure it out together? And they're like, no, I don't want to take your time. I'm like, it's okay. I learned something new. It's okay. We can work together and do this. I've become fairly decent at learning new things. And that's through a lifetime of not being good at it. My— if I had to be very— put it down into a few sentences, I would say, Look to build, like, learning routines. You might think that the annoying thing about learning routines is that it's actually hard to measure progress because there are very few milestones. You can give milestones to yourself, but you got to trust in the process. Maybe, I don't know, maybe KCD is a way better teacher and educator than I am. I bet he— not only do I bet he has opinions, but I bet he has, like, 4 blog posts on this. Because he's somebody who writes down everything he learns.
Yes. Uh, well, Todd, I, I want to talk too because, because this was part of my talk and we're coming up on time. An hour went by way too fast with you. Um, but to talk about just putting yourself out there, like, like, so I, I actually mentioned Kent— I don't know if you're still watching, but I, I mentioned Kent C. Dodds, I mentioned Andy, Angie Jones. Kent C. Dodds, in my mind, React. Angie Jones, in my mind, Java. Guess what those 2 have done over the last 6 to 9 months? They have consistently put out blogs, content, what they're learning. And guess what? They're now getting called to speak on AI events, and it's only been the last 6 to 9 months.
It's insane. You're telling me things are done when in under a year? That would be weird. But no, that's like completely not true. And you're absolutely right. Kent C. Dodd's latest obsession is MCP. I know this because—
MCP everything. Everything that puts MCP this. If I had a dollar, I'd be rich.
You shouldn't be depressed because MCP is solved, because it isn't. You should be glad that all this shitty infrastructure JavaScript library TypeScript stuff is done. You get to learn these things now when it's a little more stable. Don't get me wrong, it's going to change again in 6 months. But what a wonderful time to sit down and be like, you know what, I'm going to make an MCP server. And I'm just going to talk to it with an LLM. Again, use Cloudflare Workers AI. That shit is— bro, our shit is so cheap, man. And I'm not even saying that as a shill. Maybe we're not making enough money. We should charge more money. It's absurd. You are— bro, I was telling you about this. When I was learning JavaScript, we didn't have compilers, bundlers. We had to copy paste these files together to form, like, little script tags on pages. And it was still joyful. And like I said, ajaxian.com on shitty Indian internet at the time. With no cell phones. I had to go sit in front of a computer to do these things. Are you mad? Forget about YouTube and all this stuff. And you should consider yourself, like, lucky and liberated that It's a little more easier to get access. What I also think is that I think social media has ruined people's attention spans, including myself.
That's a whole nother conversation.
Yeah. That is the big fight, by the way, being able to sit down for 30 minutes, 45 minutes without interruption, without checking what's going on on Twitter or Reddit or whatever it is. Honestly, I have no answers for that. And I genuinely believe that is the problem to solve.
Because yeah, well, I think, and I think what I wanna drive home too is we kind of wrap up here and I have some rapid fire questions for you too as we were to end it out. But I think the big thing is that you're not late and everybody, I would say now is more on an even playing field than ever.
I look forward to hiring someone on my team who says my shit is trash and they can do a better job. That is like the best outcome for me. And the reason I say this is that's how I got my job in Cloudflare. I tried out their stuff and I was complaining about how bad the DX is online. And I don't know if you know this about Cloudflare employees, they're extremely online. So, they got into my DMs and they were like, are you going to keep complaining or do you want a job to come fix this? I was like, okay, let's go. I've never done this sort of thing before, but I'll try. And that counts. Putting yourself out there, saying the things.
I, I, I thought, I thought I, I thought I have it scheduled. I thought I had it scheduled to go out today, but oh, I, here it is. Here it is. I literally have this tweet scheduled today. If you can create a demo of a product of the company you want to interview and lead with that in a DM to the manager, the response rate in your DMs is gonna be incredibly high.
Holy shit. It is a cheat code. Why don't more people do this?
That's what I don't understand. Like, like literally if I was like, hey Sunil, I, I don't know tech, but like, I mean, I do, but don't, but it's like, hey Sunil, I don't like the way Cloudflare does this. I created a quick proof of concept. What are your thoughts on this? I'm pretty sure you're probably gonna respond to that person 10 outta 10 times.
Oh yeah, absolutely. You are using my stuff. I'm so lucky that you bothered giving me 5 minutes out of your day. I know if you actually, dude, yeah, the resume, which is like, hey, here's all my feedback. I tried out these things. By the way, I already sent you 2 quick PRs. Your interview questions are not going to be, tell me about yourself. It's like, hey, let's talk about the projects we are working on.
Exactly. It changes the entire mode of the conversation. conversation.
100%. And by the way, this is not about senior, junior. I'm not joking. Like, you can be fresh out of college and this gets you like ahead of—
Yeah, I mean, you just said an intern taught you how to use Cursor.
Dude, it was crazy that day. I was like, yeah, that was—
it was like that day's etched in my memory. All right, let's, let's go ahead and wrap up here. I want to be respectful of Sunil's time. Again, thank y'all so much for tuning in. The, the comments have been off the chain. I'm— I, I don't know, I'm— I date myself when I say that, but The comments have been great. Thank you so much. So first question for you, Sunil, what's one tech tool or app you can't live without right now?
Oh man, what is it? Cursor or Granola? I think Granola because I think I have ADHD. It's so useful to just record. I'm doing it for this conversation and I'm going to go back and be like, what did I say? What did it— this cursor, one of those tools. Not Cursor. Granola and that family of tools, 100%. I sometimes open it up and I'll talk to it just out loud and then like—
I love that. I love that. What's— all right, let me ask you this. What's the very first job you ever got paid to do when you were a kid?
I was selling CDs that would hook you up to dial-up internet when I was 12.
Wow.
So that was in 1995.
Wow. That's amazing.
That was the year I also discovered Nirvana and I was like, oh, this is a great band. I wonder what else they're doing. And I was like, oh no.
Yeah. Oh dear. Oh dear. Let me ask you this. If you were to have your pick, right? Are you a fully remote guy? Do you like onsite? If you had a, I don't know what Cloudflare does onsite stuff. I know my preference is always one week a quarter having the whole kind of everybody in the office? I just always ask this to get folks' thoughts on this remote culture right now.
I work fully remote right now, but Cloudflare is extremely remote-friendly to the point they actually don't care whether you come into office or not. But I think those are 2 different types of work. I think the work you do when you're like, I'm— yeah, don't enter the door, I'm just locking in, writing a bunch of code, versus actually hanging out with teammates, whiteboards, those are 2 different types of work. There's a third type of thing which is not work, but equally important, which is having a beer and actually like—
Incredibly important. We don't talk enough about this. Maybe that's a whole—
Learning to like the people that you spend over 8 hours a day with.
It's also networking for your future too.
Come on. Oh, you know, oh, I— so by the way, despite my bad academics, one thing I'm extremely proud of is the relationships that I've built in this community. At a time when tech hiring is so bad, the relationships and network I've built mean that I can get a job with a single tweet. I'm not showing off. I'm saying it's taken 15 years of work to get to this point. You should invest in it. 100%. Which is to say, don't trash people online. Don't pick fights. Be intelligent and be social. Yeah.
Yep. 100%. Okay. So back in the— I mean, you were selling tech stuff when you were 12 years old. Did you always want to be in tech when you grew up, or did you have another profession in mind?
I thought I'd do computer science research. I didn't think I'd do it commercially.
Okay.
Some more research type stuff. Yeah. Yeah. I was always, always fascinated with computers. For a while I did think I'd become a writer and join a band and stuff. There you go. I actually, I thought that was what I was going to do. Then I did poorly in uni and I took the programming job. Uh, to pay the bills, uh, and I never stopped.
I love it, I love it, I love it. Uh, now comes an important question: pineapple on pizza, yes or no?
100%.
Oh, Sunil, where'd you go? Come back. He turned off his camera. Sunil, that question upset him so much that he had to log off. Sunil, you there? Sunil, you there? The pineapple on pizza was was was was such a big thing. Sunil, you there? What happened to his camera?
Can you hear me? I can. What happened to your camera?
That question upset you drastically. Yeah, I know.
Geez, like my no, like not me. My entire computer went on the fritz. I had to disconnect everything.
Pineapple on pizza, yes or no? Click.
The answer is yes, 100% yes. But obviously the gods don't want it. They're looking, looking at me.
That was such a great— that was such a great cliffhanger right there.
Am I gonna lose my citizenship over saying this?
I feel like I'm gonna— You upset somebody. You upset somebody.
Pineapple on pizza. Sweet, salty, spicy, crunchy, everything on the same. Of course, 100%.
There you go. I love it. Last, last 2 questions here. Um, what is a career mistake you're weirdly grateful for?
Uh, I did very, very badly on the React team, uh, and I left, and I thought my life was over because that was my dream team to be on. Uh, that brought me to Cloudflare.
There you go, there you go. All right, last question as we wrap up here. Um, the show is called Guidance Counselor 2.0 because it got— we— because we had guidance counselors in middle school, high school, and college, but now as adults we don't really have guidance counselor to turn to and ask for career advice. What would be your biggest piece of career advice to young Sunil? Young Sunil.
Invest in health and healthy routines. 3 square meals a day, 20-minute walk, maybe a little lifting. That has done way better for my code and my life and my career than any other form— static type systems, TDD, anything like that. The days I'm healthy and happy, I do a much, much better job than anything else. And I wish I could go back and do a little less grinding. Yep, I love that.
All right, y'all, well, thank you again for hanging out with us today. Sunil, thank you, my friend. Stay on so I can say bye to you backstage. Y'all have a good one. We have our fellow Torque Ambassador Will Klein on the show tomorrow. Going to talk about how he's overcome social anxiety and ADHD to present in conferences and navigate kind of the in-person uh, circuit. So pumped to see you guys tomorrow. Sneal, thank you, my friend. Y'all have a good one. See y'all tomorrow morning, 9:30 AM Central.
Peace.
More from taylor

tech doesn’t have to be the only thing that brings us together #techcommunity #techcareer

I want to give back to y’all that make the tech community so great #jobsearchtips #techjobs

spend a little time investing in yourself #linkedin #linkedintips #careeradvice







