How to Spend Your 20s in the AI Era

AI has upended the once "safe" CS career path. New grads are facing unemployment rates twice those of art history majors, and a CS degree is no longer a surefire ticket to wealth. At the same time, small, focused teams are scaling from zero to eight-figure revenue in months. In a special Lightcone Live at AI Startup School, Garry, Diana, Harj, and Jared discuss why it's now more important than ever to focus on building real skills, domain expertise, and agency rather than just chasing credentials.

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Garry Tan?

  Think about the area under the curve of utility that you could contribute to society. And everything else is simulacrum. It is not real. When you think about SBF, when you think about Theranos, when you think about the things that truly disgrace us as people who create technology. When you peel back a little bit, you realize there's nothing. It was a fucking lie. I don't want that for us. People outside of this room, the world at large, looks at tech and they hate us sometimes because those are the people who represent us. And I say, not for me. They don't represent us.

• 0:43Garry Tan?

  Welcome to another episode of The Light Cone. This time, we're doing it live. We're not used to doing it in front of a studio audience. So we thought we would start off with a controversial topic. This is something that a bunch of people who are at this conference have been, I don't know, just talking about, coming to us to talk about. Is this the last window to get rich? Are you worried about this? Are you guys worried about this? Is this the end of capitalism? What's happening?

• 1:10Harj Taggar?

  We're like, no.

• 1:11Diana Hu?

  Mom gonna stop to exist without EGI?

• 1:14Garry Tan?

  They won't admit to it, but in private conversations, this is one of the topics that certainly we've been debating. Well, you know, why is this coming up, actually?

• 1:24Harj Taggar?

  Seems like, at least when we speak to people applying to IC, who are kind of like members of the audience, there's a real sense of uncertainty created by AI right now, right? Like the things like the sense of, will the jobs that we thought would be there be available? And if we're not, if they're not, like, kind of what do we do? And if we're not sort of, if we don't have real ownership in something that's like valuable and growing, like what will we be left with? But that seems to be the thing that comes up a lot.

• 1:53Jared Friedman?

  I had dinner with some undergrads who are here last night, and they were saying that this is one of the things that people are talking about a lot on college campuses. It's just like, hey, the AI has gotten really good at programming now. What's gonna happen to all the programming jobs? Like, it used to be the case that if you were a CS major, there is a very clear path to like a very stable, like, upper-middle-class Background where you get like a good, stable job as a programmer. But like, are those jobs still going to be here in 10 years? Like, yeah.

• 2:24Garry Tan?

  Yeah, like my, my parents were really proud when I, you know, graduating, I, you know, got my degree. And then I got my job at Microsoft and I was a level 59 PM. You know, lowest of the low, but I had health insurance and my parents were really, really proud of me. And, you know, one of the fears frankly, like that we're hearing. And it's sort of, you know, coming out in the numbers is that will there actually be jobs? I think it's a tricky thing right now with the advent of intelligence. You know, some of the simplest things that people rely on entry level, people right out of college for. They're not hiring as many of them anymore. And, you know, the craziest stat, I think this came out of the New York Fed in February of this year.

• 3:11Garry Tan?

  Computer science majors. You know, obviously this is not the people in this room. This is just like, out of like, you know, a normal distribution of all computer science majors. 6.1% in unemployment in February of this year. Art history in contrast was only 3.0%.

• 3:29Jared Friedman?

  Wait, you're saying that the unemployment rate of art history majors is lower than the unemployment rate of CS majors?

• 3:35Garry Tan?

  Unbelievably, but that's what the stat indicated. Don't worry, guys, no one in this room, no one watching the hotline. We're talking about, you know, the, the median, which, you know, you guys are so, so many standard deviations above. Don't worry.

• 3:51Diana Hu?

  But that's concerning, right?

• 3:52Garry Tan?

  Yeah.

• 3:52Jared Friedman?

  Yeah, but like, but like this, this role of like, like level 59, you know, engineer at Microsoft. Used to be this like super stable job. If you do that job, all the adults in your life will be like, good job. Like you make the, you made the safe choice, the prudent choice. But like, one of the things I've been, like, I've been noodling a lot. It's like, is that actually the safe choice? Like, is it possible that the world has become inverted? And like the career path that seemed to be like the lowest risk, most safe path might not be anymore.

• 4:23Harj Taggar?

  Yeah, I think like one thing that's going to be interesting for this audience is that there's one theory, this um, Brian Kaplan has this theory on education, I think it's Brian Kaplan at least, it's like where um, it's basically all about it's credentialing, but it's actually a very specific thing that's being credentialed. It's like what colleges are credentialing to employers is that um, these people graduate our program, which means that they can like show up to a place on time and like, perform a series of instructions and you know, not do too many drugs and like kind of like, make it through, which is like the kind of people you want to hire and like they'll turn up, they'll do the job and.

• 4:59Jared Friedman?

  If you're a Microsoft.

• 5:00Harj Taggar?

  Yeah, yeah, like Fang, like I think like any company at scale starts like that's actually what they are hiring. It's like you went to a good college, which means that you can like do things reliably and follow instructions well. It's pretty clear now in the AI world that like the AI is very good at following instructions. And it's probably going to be hard for humans to compete with the AI on just like, following instructions reliably, in which case people here need to, think about what are they going to get out of their college experience? That goes beyond just kind of showing up, passing the test, following the instructions really, really well. Like it's going to require, how do you know to do things yourself? And how do you have like agency and independence? Because that's actually the stuff that's going to matter in like, I think, a post-AI world.

• 5:45Diana Hu?

  And I think the thing that happened is that Derek and I went on this college tour as well. And what's happening is that a lot of this CS curriculum is actually quite outdated. Like how many of you in the audience, if you're still in in college, do your courses even allow use of a cursor?

• 6:05Jared Friedman?

  Yeah, how many like forbid you to use cursor and like vibe coding tools in your CS classes?

• 6:10Diana Hu?

  Oh yeah, way more hands.

• 6:11Jared Friedman?

  Yeah, yeah.

• 6:13Diana Hu?

  And this is the future and those are the kinds of skills that are now.

• 6:17Jared Friedman?

  They're they're quite literally prohibiting the students from learning the tools that they are going to need in the future. It's crazy.

• 6:24Garry Tan?

  But when Google first came, when the internet first came out, um, a lot of uh, teachers would say you're not allowed to use Google.

• 6:30Jared Friedman?

  Totally, yeah.

• 6:31Garry Tan?

  Which is unfathomable today.

• 6:33Jared Friedman?

  Yeah.

• 6:34Diana Hu?

  And I think to Harj's point, a lot of the most crack students, as we were meeting them and all these, we had some events that were hosted. They had this sense of working on the side and working in a lot of side projects. To your point, Harj, about having a lot of agency. You learn a lot more in the process of building a lot of projects on the side rather than at school. How many of you have learned way more on independent projects than at school? Yeah. All the hands.

• 7:01Harj Taggar?

  All right, we picked the right people.

• 7:05Jared Friedman?

  What do you guys think is the answer to Gary's question? Is this the last window to get rich?

• 7:08Harj Taggar?

  Being intensely honest, as one of our colleagues, Paul Bluehyde, pointed this out, where it's sort of a flaw in the logic potentially. Like, if this is actually the last window to make money and get rich, then you're basically implying that, you know, that we're going to get some definition of AGI or ASI, whatever you want to call it. That's like a necessary condition for that to be true. Like, in which case, we're probably going to have like bigger, like, you know, there's going to be a lot more going on than just figuring out like how to make this like human money. I think that's a concept that Paul talks a lot about, is that in a world where the machines can do everything that's better than humans, like, what value will they even be in human money? In which case, why does it matter that you're going to race to accumulate like the human money now?

• 7:50Garry Tan?

  The game itself might change. The, you know, you sort of, you grow up, you go to college, you graduate, you go work, and you get a job, you buy a house, you have a mortgage, all this stuff. And then, you know, one of the weirder things that I see people critique San Francisco about is somehow this belief that San Francisco itself is about like credential maxing, which I don't, I mean, the part of San Francisco I want to spend time with is like not really about that. But I can see the critique.

• 8:21Harj Taggar?

  In general, I don't think people do their best work out of fear. Like, you do it out of like more positive motivations, because you're excited about stuff. And so I don't think my advice to anyone here would be you should drop out of college and work on an AI startup because it's going to be your last chance to make money before, I don't know, like, the event horizon hits us. I do think something that's interesting to note is just like the, how quickly like AI startups can grow is definitely something we've talked about. But if you think about Something I've been thinking about recently, like all of us, actually, when we were in college, like we would, you'd always have speakers come back, like startup founders who were like a year or two out of college, come back and speak and you hear from them.

• 9:06Harj Taggar?

  And I kind of remember that the milestone to hit when you were like a year or two out of college, having worked on your startup, was like raising a series A round of funding. And we'd have like the Dropbox founder come back and say like, yeah, I raised like a series A and it's like really, really cool. And then it sort of became, okay, well, actually, like maybe a couple of years out of college, you could raise like a series B or a series C. If you fast forward today, it's like you've got the Cursa founder, a couple of years out of college, coming back with a $10 billion company. Like it's like the, the pace.

• 9:36Diana Hu?

  Order of magnitude.

• 9:36Harj Taggar?

  Yeah, it's like the orders of magnitude difference, right? So I actually think a far more exciting reason to think about, um, you know, should I start a company or should I join like one of these fast growing companies is just like the time, like how much you can get done a year or two, like out of college is orders of magnitudes higher than it was even a few years ago. I think that's like for a certain type of person, a very like exciting moving factor.

• 10:02Jared Friedman?

  I like that a lot, Harsh. I think that's why Sam said at the beginning of the, of the event yesterday, this is the best fucking time in history to start a company. Yeah.

• 10:10Garry Tan?

  Well, the interesting thing about credential maxing and or what's happening now is that raising a series A is a credential that kind of gets bestowed by a fancy VC, uh, you know, driving a Ferrari down San Hill Road or something, right? Like that's something that's external to outcomes and often it's, you know, really like the shooting of a, a starting line gun as opposed to like something to celebrate in and of itself. The really big difference today is that the very best companies that we get to see day to day, they're like, I don't know, five people, 10 people. Uh, I think, you know, each of us on stage and all of the Y C partners are sort of collecting incredible startups that we get to work with that went from zero to 10 million, 12 million a year revenue.

• 10:55Garry Tan?

  Like that's net revenue, like it just goes in the bank. So you basically get the equivalent of an entire series A.

• 11:04Garry Tan?

  Instead of this fake credential thing where some fancy person on Twitter with lots of followers, you know, blesses you and suddenly like all these people, you know, TechCrunch says like the new hottest founder. You know what? Like, those are all external things that are not actually connected to real business or having an impact on anyone. It's fake, right? It's the fake credential. And then the cool thing now is that that is actually very directly being replaced by people making things that, you know, people not only really need, but they're willing to pay a lot of money for.

• 11:39Diana Hu?

  This is like a very good point. Sort of instead of taking the leap out of fear that this is the only time, taking it from a place of approaching and really going after something where this is really an exciting time to be a builder. We've seen crazy growth unlike anything, only possible with right now with AI. Like all these companies that we work with, zero to 12 million in 12 months. The cursor story where they went zero to one in one year, the next one one to 100. This is like unprecedented in tech history.

• 12:14Jared Friedman?

  For B2B SaaS companies, it didn't used to be the case that B2B SaaS companies were the ones that had hypergrowth. Like there were some like consumer social companies that got hypergrowth, but like B2B SaaS used to be this like, you know, clotting, slow growing, like, kind of thing. Now there's this weird inversion. It's the B2B SaaS companies that are like the hypergrowth ones.

• 12:34Diana Hu?

  I think what we're seeing is a lot of times is founders who are living in the future, the cutting edge, who are winning here, because you have to sort of build the taste to build something good. And you don't get taught some of those things in school.

• 12:47Harj Taggar?

  Another like on that front, like something very specifically we're seeing is that to build any products, you always need some combination of like domain expertise, which is really just like understanding your customer really well and understanding the space you're building in and understanding the market really well. And then technical expertise to actually build the product. And it feels like pre-AI thing shifted where.

• 13:09Harj Taggar?

  Sort of the technical expertise wasn't that important because it was real. Most of the software was like web software and it became fairly straightforward to build web software. And actually, all the value was in how much like domain expertise you have, like, do you have relationships with the customers you're going after? Do you have some edge on how to sell to them? Because everyone you're selling to has already got like 10 roughly equivalent products being sold to them. And that actually made it quite hard, I think, for college students to be able to like, go and compete for like, you can't like compete on so compete with Salesforce for like a CRM or go build like the best appointment booking software for healthcare practices. Like all of these things where it's very saturated. And now I think what we're seeing is with AI, there's this promise of, hey, like this is more than software.

• 13:54Harj Taggar?

  Like this can do like the work of people. It's like magic, but like it's actually quite hard to do that reliably. And so there's been this flip of where the technical expertise is now actually really like the missing piece for a lot of these things. And we consistently see at least in Y C that college students are actually at the forefront of this stuff. Like actually understanding how to use the models and how to squeeze the performance consistently out the models is something that even like, you know, PhDs and people are really experienced don't get. I think maybe that's why Elon had that sort of look yesterday when he was talking about researchers versus engineers. It's like, it's actually in the engineering and it's like working on the projects and like building real things is where you get the expertise.

• 14:37Jared Friedman?

  Yeah, I had a lot of college students ask me over the last two days, like, hey, I don't have domain expertise in any particular area because I haven't worked in industry that much. Like, what idea should I work on? And like, how do I basically like, how do, how do I get enough domain expertise to like, do something interesting? Or what advice would you have for folks in that position, Harj, based on, on that insight?

• 14:59Harj Taggar?

  I think Gary's got like a great point on this. It's basically like become like a forward deployed engineer, right?

• 15:05Garry Tan?

  Yeah, just, I mean. Go undercover, I guess, like go and figure out what people actually need. And yeah, there are just too many examples of billion dollar startups that we got to see. I mean, I always think about Flexport, you know, here's this guy who literally became one of the top importers of medical hot tubs. Like, I don't think anyone wakes up, you know, and graduates and decides like, hey, I really need to become one of the foremost, you know, import exporters of of medical hot tubs. But, you know, he did it, he did, they also, I think, were one of the first, the biggest e-bike importer. But then, you know, basically being in weird parts in the economy caused them to understand just things that the other person,

• 15:51Garry Tan?

  the sort of thousand, 10,000 other people who want to start startups, like, they didn't have that knowledge. And so sort of your ability, you know, if you're here, like, your inherent ability already is like one part of the Venn diagram. And then the other part is just something weird. It's literally just like, where does your interest come from? Like, I'm really taken by to what degree both OpenAI and SpaceX, for instance, were, you know, the genesis came from, like, interest and a hunch and just like not really any commercial intent. And yet, you know, coming out the other side, that was enough to attract the smartest people in the world,

• 16:32Garry Tan?

  attract capital, and then really create, you know, the most enduring businesses in the world.

• 16:38Jared Friedman?

  Yeah, the other thing that I've seen that's pretty cool is just, I've just seen a lot of college students go from having like no domain expertise in an area to being like total experts in like a month or two at Y C. And I think people maybe don't give themselves enough credit for how quickly you can become an expert in something. If you're just smart and you learn fast and you just make like a concerted effort.

• 17:00Harj Taggar?

  I think the door is more open now than ever. Like, you kind of go back to, yeah, in a world where like any domain, let's make it like, you know, if you're trying to build. Software for dentists is a random example. Pre-AI, it was just like people were being pitched with so many different software products that they weren't actually that receptive to like some college students promising some software and wanted to come like learn and like work in the office and understand like how it works. It's like, yeah, like twenty software vendors all like telling me the same thing. But now, because like AI has captured the imagination of everybody, everyone like wants to know what's possible and are consistently underwhelmed by what like the established software companies can offer them, but they're open to like college students just coming in and like.

• 17:44Jared Friedman?

  Well, because the college students are selling them pure magic.

• 17:47Harj Taggar?

  Yeah.

• 17:47Jared Friedman?

  So I had three founders in the last batch, actually, that are building quite literally like AI agents for dentists.

• 17:51Harj Taggar?

  Yeah.

• 17:52Jared Friedman?

  None of them, like I think their only experience with dentists is they went to a dentist.

• 17:58Jared Friedman?

  And but it was exactly what you said, Harsh. Like, they're literally selling these dentists like magic in a bottle. And so like, of course, the dentists will spend their time because if it works, it's like just incredible for their business.

• 18:09Harj Taggar?

  Which kind of just comes back to the agency thing because it's like the thing like in order to build these products, in order to go out and like build like the future big companies, you kind of just have to have the agency to be like, oh, yeah, like I'm actually gonna go do the like undercover agent or full deployed engineer and I'm just gonna go like camp out in like someone's office and just see how they do their jobs and learn how to do it and learn how to like build it with AI.

• 18:34Garry Tan?

  What about some pitfalls, like things that would prevent people from exercising their agency or exposing themselves to, you know, the real economy?

• 18:45Diana Hu?

  I think one of the things that keeps coming back to my mind is having a lot of these conversations with recent grads or college students. I think there's this arc of a lot of you trying to figure out what to do with your life. And through most of your life, you've been conditioned to kind of just pass tests, study for the exam, do the homework. And it's sort of like all these very constrained boxes that you have to check. And then you treat startups or your next job sort of like another test or exam that a lot of the rules are predetermined and you just have to go check the boxes. But that's the complete wrong mental model for it.

• 19:25Diana Hu?

  Because the problem is that when you go after building and tackling a big problem, it is an open wide space. There's no rules. You get to create it. I mean, the good thing about startups is. Plus and minuses, you have agency to design what you're gonna go after, set your goals. Instead of like some authority figure to like, oh, you need to do this and that. And we get asked questions like, oh, what should I look like in order to raise money? That is such a student question. Sort of like there's some sort of bar like by some higher power. Guess what? There's no adults in the room. It's you, you're in control. And you get to design those rules and you can go as fast as possible.

• 20:06Diana Hu?

  You don't have to have like, oh, we have to do this and this and check the marks and get there. It's really you design it, you're in control.

• 20:15Garry Tan?

  I think there are two very dangerous uh forms of like credentialism that you create for yourself. That we see that actually like we'd really like to warn you guys about. Uh, one is, I mean, I think we already talked about like making, raising money from investors. Like somehow the, the biggest goal, I mean, including us, by the way. It's like, you know, that we're just like people to help you and we think we can help you a lot. But like, once you turn that into like sort of uh the idol that you have to achieve. Then that's just missing the whole point. And I, you know, I think that that's quite dangerous. Um, the other thing that we were kind of concerned about is there are like entrepreneurship programs at some of your campuses.

• 20:57Garry Tan?

  Uh, some of them might take you to wild exotic places for retreats. We're not gonna name them, but like in full transparency, I'm very worried about them. Because what we're coming, we're coming to understand is they are teaching you to lie. And that is at a moment when literally all of software is changing. And that software is the most empowering thing in the world. Why do you have to lie? I understand in a world of like contracting capability, in a world where there's less money. Where there's, you know, fewer and fewer jobs. I kind of get it, it's very zero sum. We're at the most open, like sort of abundance oriented like mindset thing that is happening right now. Like literally everyone here is hyper, hyper empowered.

• 21:42Garry Tan?

  You don't have to play by those old rules anymore. You don't have to lie to investors. You don't have to like fake it till you make it. Like, you know, I worry that some of these programs are just literally trying to teach people to become more uh, you know, SBFs and Theranoses. And that's like, you know, that's a waste of time, like, and you're gonna go to jail.

• 22:04Jared Friedman?

  Also, a lot of these entrepreneurship programs, they do what Diana said, which is like entrepreneurship programs, especially ones that are not started, not run by founders, you know, all of us were, were, were started founders. They, they basically teach entrepreneurship like it was a course, like, like it was just a series of tests to pass, a series of checkboxes to, to check. Anytime you try to bottle up entrepreneurship and like teach it as a college course, that's kind of what you end up with is like basically like a, a sort of like cheap facsimile of entrepreneurship where like they teach you to like, you know, follow a particular method or a particular practice. And it's just not what startups are actually like.

• 22:40Garry Tan?

  I just think about that Jay Z line. It's like, everybody want to tell you how to do it. They never did it.

• 22:45Harj Taggar?

  True. A riff on that, and I'm curious to get people's opinions on, maybe especially Gary, actually. Something that is clearly different, I feel about the age we live in today versus say 10 years ago is just social media and using social media as a way to like amplify your message and your brand. This is actually something that came up at dinner last night is how much in the early stages when you're building a product, like, should you focus in on kind of like building the product and going one by one to get users, all of the kind of like traditional startup advice versus trying to cultivate sort of like a following or a brand or attention like online. And like, you know, 10 years ago.

• 23:25Diana Hu?

  Or spend thousands of dollars on a video.

• 23:27Harj Taggar?

  Yeah, yeah. Just like spend like high production launch videos and like lots of follow, like lots of followers on Twitter or X and

• 23:35Harj Taggar?

  certainly I think it's more confusing now because that wasn't even an option before. And you definitely see people succeeding at the getting the online like attention and people talking about like the company.

• 23:47Garry Tan?

  What do they call it? Aura farming?

• 23:49Harj Taggar?

  Is that what they're calling it these days? I can laugh. So maybe that's the, the phrase. Yeah. It's like aura farming, I guess. Yeah, I'm curious what you think, Gary.

• 23:57Garry Tan?

  All I care about is what's real and what you can, you know, touch and see and feel. And, you know, think about the area under the curve of utility that you could contribute to society. And you can always just look at that as ground truth. And everything else is simulacrum. It is not real. It is like media, it is fake, it is a credential, it is a thing that represents something. And yet, like, if you look deeper into it, it's nothing. Like, there's nothing. When you think about SBF, when you think about Theranos, when you think about the things that truly disgrace us as people who create technology. When you peel back a little bit, you realize there's nothing. This was just simulacrum. It was a fucking lie. I don't want that for us.

• 24:40Garry Tan?

  Like, you know, people outside of this room, the world at large, looks at tech and they hate us sometimes because those are the people who represent us. And I say, not for me. They don't represent us.

• 24:59Harj Taggar?

  So I think that's a no on social media.

• 25:01Garry Tan?

  I mean, I think social media is really great. I mean, I'm clearly extremely addicted to it. And it's done some really great things for me, some terrible things too. But, um, you know, I think that, you know, you do have to tell your story. Like, I think one of the more important things that is the gift is that you can tell your own story. Like, in fact, that you have to. Like, the second you rely on someone else to tell your story, it's gonna be great, great, great. And then when you don't have that voice, like, someone's gonna take that, you know, there's... The world wants to, you know, the only thing it loves more than like, you know, a, uh, a story of like, of becoming is one of unbecoming. And if you don't have that voice and you can't go direct, um, you know, they're gonna do that to you. And so, better to start, right?

• 25:46Garry Tan?

  Like, you know, I think working backwards, the thing that was most helpful for me, you know, obviously, for my startup, but we try to encourage all of our startups to do this, is you can sort of work backwards from the outcome that you want. Like, uh, I actually think Apple does this really, really well. Like, they don't commit to doing a feature. Until they have, you know, product manager who says this is who it's for. And this is the problem it's going to solve. Right. And you can actually turn that into going direct. So, you know, let's say you have a one week or two week sprint. It's all too easy to just say, look, like, here's my bug list. And here's my backlog and I'm going to fix these 10 bugs. But a much more powerful version of this is let's work backwards from what I want to put on X.

• 26:29Garry Tan?

  I'm going to make a very simple loom video showing off a feat of strength. A thing that I really want to share that I know our team can do. And then working backwards from that, the next two weeks, that's all we're going to do. Like you could storyboard it. It's like, it's going to do this, right? You can work like basically at that point, media and PM and design can be all the same thing. You know, it's connected to users. It's connected to communication. It's connected to what your product will do for people. And then you build it and you can just basically rinse and repeat like that. If you can do two week sprints of working backwards from a really powerful, not flashy. Uh, loom video of what you did in the last two weeks.

• 27:10Garry Tan?

  You can do this all for each other and we can create a culture that is not about flash, but about substance.

• 27:20Garry Tan?

  We got a question over here.

• 27:22Speaker 04

  I'm a big fan of Gary and Jared. So Jared actually inspired me. Brutal. For my startup. But I just wanted some advice because I'm kind of going through a dilemma right now. So I've been working on a startup for the past month and I'm, I'm going to my third year of uni. Um, for context, but, um. So yesterday at one of the after parties, which I'm not going to disclose the name. Cause you guys are probably going to apply to it, but, um. Uh, I was pretty much pitched my idea to one of the founders and he basically said like, you know, drop out of school and come work for me, uh, move to San Francisco. So, um, I'm really stuck between like, what do I want to do? Like, what's the right choice? Like, do I continue university and then go to, uh, San Francisco?

• 28:07Speaker 04

  And then, you know, grind the startup life? Or do I drop out now? Cause I'm already halfway done. Like, you know, I'm not like almost finished with college or anything. I'm already halfway there. So do I drop out and just work and then just move on from there?

• 28:20Garry Tan?

  I mean, I think the most important thing is do you trust them? And is it actually a good startup? So, which is, uh, kind of a hard question to answer. Like this, but like what, you know.

• 28:31Jared Friedman?

  Well, if it's one of ours, it must be a good startup.

• 28:33Garry Tan?

  Is it a YC startup? Yeah. Oh, you should probably do it. No, I mean, more seriously, I mean, I don't know. How would you, like, when you think about like where someone should go, like, what would you say?

• 28:45Diana Hu?

  I think Gary could answer.

• 28:46Jared Friedman?

  I'd add one, one third criteria, Gary. I dropped out of college to do Y Combinator and do a startup. I think in addition to the two ones that Gary said, the third one for you to consider is like, do you really like being in college? Which is like obvious, but like actually, yeah, okay. Like, like, I think, I think like, like Hart was saying earlier, when college students are thinking about dropping out and they're making it in a fear, they're making a fear-based decision. They have FOMO about startups happening in San Francisco. Maybe their friend dropped out and they're like worried that they're going to miss out. I don't know that those are the best decisions. Like, some of those people actually do end up regretting it. Versus like when I dropped out of college, I don't have a college because I was bored of college. Like, I'd done three years of college.

• 29:30Jared Friedman?

  I felt like I had gotten out of it what I wanted from the experience. I was just a lot more excited to like build real technology for real people. And I felt that regardless of whether the startup that I was working on succeeded or not, I wouldn't regret leaving college because I was just kind of ready to move on to the next stage of my life. So I think if you can, if you, if you can honestly feel that way, then maybe it does make sense for you to drop out.

• 29:53Diana Hu?

  I think if you feel you're done exploring living alternative life paths, what I mean by that, like you tried an internship working at a big tech company. You tried an internship at a startup. You tried another internship to start a company or you tried another one doing research. If you feel you fulfill the chessboard and the land of what your life could be and you explore everything, I think it's fine. But if you still have a bit of an inkling, it's like, oh, maybe I want to try what doing research, research is like. I think maybe not yet, but if you're super sure you want to have a career in tech or startups, then maybe it's fine to, to Jerry's point. It's like you already did that. Life, alternate life path.

• 30:37Garry Tan?

  I feel like I got lucky and then because I ended up at Palantir and these things that ended up being super successful, but in the moment, like, you know, I could have just, you know, Palantir could have been a bad startup, actually, and I didn't even think about it. So like, thinking back on what I should have been thinking of when I was, you know, 22, 23, it's actually really important to try to be at the most dominant places, actually. I mean, the power law for startups is so intense that if you're gonna go work at a startup, I do actually think that you should try to go work at

• 31:08Jared Friedman?

  At a really good startup.

• 31:09Garry Tan?

  A really good one.

• 31:09Jared Friedman?

  A really good startup, totally.

• 31:10Garry Tan?

  Like objectively, you should make literally a spreadsheet. You should, you know, go down and like, evaluate it the way an investor would. And then the difference is the investor has a portfolio and you just have one life. And likewise, when you start a startup, you should not try to start a startup just to be the median startup. The median startup is dead. Like, you actually, if you're gonna do it, you need to be like, you know, you need to work at superlative places with superlative people. And then that's the only way that like, good things happen. And, you know, I got lucky, I feel like. I, you know, later in my life, I became much more of a heat-seeking missile. Like, you know, I think that that's why I was drawn to YC itself, is like, this is a place that has an energy that I've never

• 31:53Garry Tan?

  experienced, even at Stanford or even at Palantir, that I just wanted to be here. And I became much more of a heat-seeker-seeking missile for like, like that type of, like, this is gonna be huge. But I wish, you know, at 22, I lucked out, you know, like, my friends went to college with, they started a company with Peter Thiel, so that was just very lucky. That's Diana's thing, is, you know, she likes to fund lucky startups, so.

• 32:19Diana Hu?

  Just get lucky. And I think you get a lot more lucky by being in San Francisco.

• 32:22Garry Tan?

  Yeah.

• 32:23Diana Hu?

  And working around really smart people.

• 32:24Jared Friedman?

  I want to ask Gary, you talked about like that level 59 job at Microsoft and like your parents are proud, you have health insurance and at some point you have to decide, okay, I'm done with this. I want to go start a company and like you work at something, maybe at night, you know, after work, you can't really like go out on LinkedIn and advertise it because like your boss would see it. But at what point do you say, I have enough here. I can really go quit my job, you know, start spending down those savings and go, go do something.

• 32:49Garry Tan?

  You're much more responsible than I am. Like I had, you know, $50,000 in credit card debt and I had the nicest apartment in Queen Anne and I bought a brand new Honda and it was very stupid. And so I had to go get a job and like, you know, I, I couldn't start a startup. I like, you know, wait, I had, I needed my friends to pull me out of that situation. So, I mean, I think you want, I don't know, at least six, maybe nine months minimum of like, just, you know, being able to live on ramen in the cheapest possible way. And then at that point, like the money in your bank is like just capital that you think of. And so that's probably what I would want. And then the other thing is, uh, I would try to bring on, I would want to work with the smartest possible

• 33:30Garry Tan?

  people. Like, I know this is a big internet debate. It's like, do you need a co-founder? Honestly, like if it's your first startup, I, I wouldn't, I would never start, like my second or third startup, sure, I could do it like alone. I, you know, have connections. I know who to hire, like all this stuff. If it were my first, I would not try to do it alone because there's just too much going on. There's like the, the gradient of things that you need to learn is too wide and you need to go, you need to go together.

• 33:58Jared Friedman?

  Yeah. And in my experience for people who are your age, who have like already graduated college or working at some company, that's the biggest limiting factor in practice for them actually doing a company is like they and their co-founder both need to be willing to quit their jobs and go in on a startup at the same time. It's, it's just like a timing problem. And like co-founders are hard to find and it's hard to make that timing line up. So my advice would be like if you had, if you and your co-founder, if that does line up for you and you and your co-founder are both in a point in your life where you're like able to quit your jobs and go all in, you should probably just do it because it literally might not ever happen again. It's actually that hard to do.

• 34:34Speaker 05

  Hey guys, well, first of all, thank you for the event. My voice is a bit cracked because of talking too much over the past few days. We're actually talking to the CEO of Strava yesterday at one of the after-parties. And he mentioned that they started being a very niche startup. And I've taken a look at all the YC commendator startups over the years, and it looks like they're getting increasingly niche. So I was just wondering, what do you think? Like, what's your take on being super niche at the beginning and then expanding? And how do you know how niche you have to be in the beginning?

• 35:07Harj Taggar?

  Being niche at the start has actually always been like the recipe to succeed. But even in the YC world, the current biggest company by market cap is Airbnb. And Airbnb was like the definition of niche when it started. Like, it was literally air beds in people's living rooms, right?

• 35:26Diana Hu?

  During conferences.

• 35:26Harj Taggar?

  Yeah, during conferences. So like, I'm not sure I can get more niche than that.

• 35:31Diana Hu?

  Democratic conferences.

• 35:32Harj Taggar?

  Yeah, democratic conferences, yeah.

• 35:35Harj Taggar?

  And it obviously, it turns out that that expanded into just being this like monster company that's taking over all of travel. But there's even less obvious examples of this, where people don't realize that things that seem huge now were niche at the beginning. Stripe, actually, in a sense, obviously, it's payments, of course, is a big market. But actually, when they first started, it was like an API for developers. And the only thing that differentiated it from Braintree was that you could take payments instantly. And people actually didn't think that that was much of a wedge. It was like, oh, okay, sure, if I'm working on a weekend project, I'll care a lot about that. But like, big businesses aren't going to care about that. They're fine to wait two weeks to get like their merchant account. And so it's always been the case that niches have been the right way to start, actually,

• 36:19Harj Taggar?

  and to dominate a niche and find ways to like expand into like adjacent markets and grow into a big company, I think, is like the recipe.

• 36:28Diana Hu?

  And Brian Chesky quotes a lot of some of the best advice he got from PG during the batch was to really hone in to find 10 people that love your product, much better than, I don't know, 100 randos. And a lot of companies start like that. You want to find those maximalist users that really obsess with you, and you iterate on those. Coinbase was also very niche.

• 36:55Harj Taggar?

  Yeah Coinbase was classic niche because crypto itself was small and fringe and even within that they were building for what people thought was a non-existent market. Essentially, it was just like regular people who wanted a nice user interface to like buy and hold Bitcoin.

• 37:10Diana Hu?

  It was a bit of an oxymoron, regular people in crypto.

• 37:12Harj Taggar?

  Yeah, exactly. Yeah, the conventional wisdom was just people who want to use it to like launder money and buy drugs and it's like, okay, no, there's like other use cases for it. But with AI more than ever, by the way, like I think that niche is the way to go because no one really knows how big the markets are and it does seem like things that seem like they were niche before AI, you can get people to pay you a lot, lot more money for because they're not just buying software, they're buying like work from you and so find the niche you're really interested in, optimize for your like passion and interest in it and just pull on that thread.

• 37:46Garry Tan?

  This is like actually a really powerful moment because literally you have 130, you know, I think of O3 as basically about 130 IQ, maybe O3 Pro can be even smarter than that. When I really think about that, it's like, oh yeah, like a lot of the people who I've ever hired in my lifetime are like, yeah, O3 is smarter than that person now, so. And then you can basically take that and you know, connect it to the proprietary data systems of almost any niche and the more weird and unlikely for someone like someone in this room to know about it, the more likely that will be durable enough mode that you can get, you know, you can basically get, you can wedge yourself in there and then

• 38:29Garry Tan?

  basically all you need is a wedge and then you just basically expand that wedge until you have the pie. Thank you guys so much for coming out.