Paul Graham, Founder, Y Combinator

22:38Paul Graham:Carolyn, were there any moments, this is a very naughty N-A-U-G-H-T-Y question, were there any emails that happened that Paul sent that you were like, oh, geez, oh, gosh? Well, probably, but no. Actually, I don't think I thought that. Like, oh, this is going to get us into legal trouble kind of emails. No, I never would say things that would go into legal troubles to make people mad. Yeah, just like, oh, that's really going to piss someone out. Oh, I'm sure that I was like, this is exactly like the harsh truth that this founder needs to hear. Like, I don't think I ever had the reaction like, oh, that's going to go over poorly.

23:20Okay, that's good. I think people sometimes just need sort of the unvarnished truth from someone who's like, look, this is what you need to hear. You know, I wanted them all to succeed. I wasn't going to be shooting them down.

23:31Paul Graham:Paul, no one worked harder than you. I swear, at helping startups who are failing. You would not give up on them. Other people would be like, whoa, just let that one go. And Paul would be like, no, I'm going to help them with a new idea. Or you would just work so hard. This time will be different. When did you start taking office hours? Remember when you built the software that allowed startups to book office hours? The booker. You built that, right? I wrote every bit of our software. Don't you realize, like, up until the point where I retired, I wrote every bit of our software except the JavaScript that did the up and down buttons on Hacker News, which PB wrote because I didn't know JavaScript.

24:18I wrote everything. If we had software, I wrote it. And I wrote every word on our website.

24:25Paul Graham:Yeah. Yeah. Our website was awesome. I loved our website. That's why my hair is white now. One of my memories of that era is how you would always – well, you would frequently help startups with their name. And I think – we might have talked about this in place. I still do that. No, no, no. I know. So my question for you is do you have a name you're particularly proud of where you're like, why don't you call yourselves this? And they're like, amazing. Imgix. I-M-G-I-X. Five letters. Carolyn has no idea who Imgix is. No, I do. I remember the name. What batch is it though? Oh God, I don't remember batches.

25:03Find, P-H-I-N-D. Oh, I remember that one.

25:06Paul Graham:I feel like if I had a dollar for every time you were at instant domain search with founders during an office hours. In the kitchen, right? Like right on your kitchen counter. Because you need them to stand around and look over your shoulder. So I do it right on the island in the kitchen. So many names, so many startups names have been found at that one corner of the island. So you're kind of famous for blowing up people's ideas, like taking their idea. In the good sense. In the good sense. In the good sense. Making them, here's how you can be the next Google. And you used to do this and some startups would take your advice, some others wouldn't.

25:44Paul Graham:Can you remember an instance of that? Because you'd come home and you'd be like, oh, wow, if you had been there, you wouldn't believe the idea we cooked up. Yeah. It's going to be the next Google. You know, I can't remember specific instances because I would do it so much. I mean, I would do it like maybe one third of the office hours I did. We would figure out how to make an idea much bigger. Yeah, it happened all the time and still does happen. Yeah. When I do office hours now, I still do that. And, you know, I see these startups right in the beginning of the batch sometimes. And so there's a lot of room to expand their idea.

26:23Paul Graham:I want to also talk about your, Matt, you know, you're really good at working. You were good at working with the startups on their demo day pitches. Tell the listeners about sort of what used to happen before demo day with the founders. I would write their demo day pitches. One of those little secrets. We would do this thing called vertebrae office hours where we would figure out what were the vertebrae of their Demo Day presentation. Because the thing about Demo Day is you're up there with 50 or now like 300 other startups. And so the investors are not going to remember anything about your presentation by the time they get to the end.

27:01They're brainily wiped clean. And so you really can only hope that they'll remember about five sentences worth of stuff in the very best case. And so you should decide in advance what those sentences will be. And those are the vertebrae of your Demo Day presentation. The rest is just filler, essentially. And so we would try and figure out what are the five sentences that everybody ought to remember about this startup that would cause someone to invest in them. There are five sentences with a specific purpose. So like, this is going to be a giant market. We are already solving the problem better than anybody else.

27:39It's ours to lose. And when we get there, there's network effects. And so that means it's winner takes all. Right? That's a pretty, you're already wanting to invest, don't you?

27:49Paul Graham:Yeah, sounds great. That's the thing. Demo day presentations shouldn't be like fast talking marketing speak. They should be almost like logical proofs of why this company is good to invest in. Like you follow, if you agree with this step and this step and this step, they all follow from one and the other. And the end result is the expected value is really high. That's a good demo day presentation. But I feel like you did more than just vertebrae. You would spend hours and hours and hours at YC and people would just come in. I would do the vertebrae with them and sit down and think, okay, this is what the core of your presentation should be.

28:30And then they would go off and make something that said what I said to say. And then I would sit there and they would practice before demo day. They would come to YC and take turns standing there projecting their slides and talking. And I would sit there in the audience, essentially, almost like a playwright saying, no, no, cut that. Right? Say that. that.

28:52Paul Graham:You were like a playwright. You'd be like, no, say this. Don't use that word. Say this. It was so, such specific information. Blow that part down. We were always trying to get people to talk slower. Oh gosh. Yes. Yeah. See, you remember. I can remember. Wasn't there somebody who wrote talk slower on their shoes so that when they looked down, it was like talk on one shoe and slower on the other or something like that. I think online, I think there's a recording of me sitting, talking to Brian Armstrong about his Demo Day presentation. I think that might be online on YouTube. Oh, that's pretty cool.

29:32Paul Graham:And then I think you'd sit, you'd be like a nervous parent sitting in the green room at Demo Day the whole time. No, no, no, I didn't sit in the green room. I was only in the green room during the breaks when the, when they were presenting, I would be out there taking notes. Um, and I would be wondering like, will they get this part right? Right. Cause I knew what was coming in all of their presentations, you know, I mean, there may be a few who didn't come into practice at YC. And so I wouldn't know what they were going to say, but 90 % of the time I knew what, what they were going to say. And what I was hoping, will they get this line right?

30:08Will it land? Yeah, will it land? And so I'd be sitting there over on the side, watching every presentation, seeing if it landed. There was a lot of stuff in retrospect. Yeah. Right? In addition to being attacked constantly in the press, there was all this actual stuff we were doing.

30:25Paul Graham:I used to say, and I've believed this, that even if your startup failed, I truly believe that anyone who came out of the YC program felt like it was a great experience and valuable for them. Well, you had to learn a lot. If you were even awake, you couldn't avoid learning a lot from the experience of trying to build something and get it in front of users and figure out how to make it grow and how to explain it to investors. And then we would do a post-batch survey. And I remember you saying, if all they complain about is the food on Tuesday nights, We're doing it right. Yes. We were always happy when they complained about the food because that meant there was nothing more substantial to complain about.

31:09So we were delighted. I mean, the food wasn't a deliberate canary in the coal mine, but it ended up effectively serving that purpose. I think the food's better now, though.

31:19Paul Graham:Yeah, it's all different now. Do you remember that time dinner basically consisted of hot orange juice? I knew you were going to say the chickpeas with the orange juice. well someone at one point we had this cook who was like adventurous in a bad way um and made this sort of orange flavored chickpea dish but the founders the founders took out all the solid bits of the stew and all that was left was this sort of hot orange cheese with a few it was orange chicken orange chicken carolyn which seems a little bit more normal and then the veggie the veggie option was same sauce orange chickpeas and the chickpeas were disgusting oh no i feel like paul you you you absolutely were like we've got something's broken here we've got to change there the complaints about the food are legit this time and we actually need to change something yeah no i mean the complaints about the food often were legit oh yeah yeah well you know this is um the The other thing that YC I think still does, but you guys did a lot, was inflection point office hours.

32:26So not just these vertebrae office hours, but inflection point office hours. And I'm sure there's some stories from those. I got to sit in on one of them, and it was really interesting. Well, inflection point was sort of an optimistic name for what was going to happen, right? This is when a startup is currently on a trajectory that will crash into the ground. and we're hoping. And inflection point is when a line bends in another direction, right? And so this is when the line's going down and we hope it's going to start going up. So inflection point office hours is when a startup was in terrible shape.

32:58We would get them together with like multiple partners and we'd sit there and figure out how do we save this thing?

33:04Paul Graham:But I'm going to interject in terrible shape, but had promising founders. Yes. Right. Because if they didn't seem like they had any hope, Should we talk about palliative care? I was just thinking, should we actually talk about this? Maybe it's time. Maybe it's okay. Okay. Because it was our thing. The new partners don't use this expression. Yeah. Do you know about this, Carolyn? Palliative care mode? Of course. Okay, okay. So if a startup was doing really badly and it seemed like they had hope, then we would bring them in and we'd say, all right, come on, let's figure out how to change things to make this startup succeed.

33:41If it seemed like they were doomed and they'd be better off just having jobs working for somebody else, we would have this thing we called palliative care mode where we would just be really nice to them. So at least as their startup was dying, they would be cheerful and happy about it. So no tough love. Just love. Yeah. No tough love. No. Yeah. We would do tough love if we thought it might work. Right. That's what I'm saying. Like you wouldn't bother with tough love for the palliative care startups. You just give them like gentle care. You just be really, really nice to them.

34:16Paul Graham:And gentle care. And so we knew. We knew which companies we hoped we could turn around and which ones we thought were doomed. And we were just going to be real nice to them. I love that you brought up inflection point office hours. That was such a fun thing that we did. I know. What else, Carolyn? What else do you remember? Well, I remember we all went to lunch. we probably had partnered lunch and we came back and we had inflection point office hours with adora and she probably talked all about i mean i know she already talked about her whole home joy journey on when we talked to her um but then there was a second one and i'm afraid i don't remember the name of that startup but when the founder walked in we had a 15 minute conversation first because you told him he looked like um a roman emperor i think you told him he looked like nero so we had this random tangent where we all talked about roman that's not the emperor you want to look like?

35:07Well, I think that's why it was somewhat controversial. And we're all like chuckling and the guy and the founder was chuckling. Then we're all started like Googling Roman emperors. And anyway, that was, I don't even remember the startup or how things turned out, but. Do you remember the meme that I only like to found, I only like to fund founders who look like Zuck? I do remember that meme.

35:28Paul Graham:I do remember that meme. That persisted for years in the press. I wouldn't be surprised if it's still out there. All because we once met this, because now you know me. I see something. If there's something unusual about their face, I notice it and even say something. And so we were once interviewing this guy who looked very much like Zuck, right? And we even funded him too. We did fund him. And then the company, I don't think they even made it to the end of YC. No. They had some co-founder dispute. Yeah. Wow. And so I, like a reporter said, there was a reporter like embedded in the batch, that batch, I think.

36:08And he asked like, can you be fooled by anyone? Well, and I jokingly, self-deprecatingly said, apparently I can be fooled by people who look like Zuck. And man, like for like five years afterwards, I was only funding founders who look like Zuck. It probably set back hairstyles in Silicon Valley by 10 years.

36:26Paul Graham:Do you also remember when someone asked you about like, what advice would you give? I'm bringing these foreign, I forget where they're from, founders over. What advice would you give to them? I said, like, make sure they speak fluent English, right? Honestly, that's the most important thing you could do. That persisted for years, like that you discriminated against anyone with an accent. Which the answer to that is, come to a demo day. Tell us if we actually discriminate against people with accents. Oh, my word. Oh, my God, yeah. And back to the demo day thing. Do you remember you used to work with people at demo day who had such thick accents that it was hard to understand them?

37:12Paul Graham:There was one founder. There was one founder that I eventually got to memorize her speech phonetically. like ABBA, the group ABBA used to do. You know how ABBA's original, their early songs sound like they're not native speakers, like they don't know the meanings of the words they're saying? That's because they didn't know the meanings of the words they were saying. They were just singing it as sounds. And there was one founder whose accent was so strong that she had to memorize her Demo Day presentation phonetically. It's amazing. I want to talk a little bit about Sam Altman too, today. Okay.

37:49As long as we're talking about famous lies everyone tells about YC, let's talk about the lie that we fired Sam.

37:56Paul Graham:Oh, we're going to do some myth busting. Well, let's first reminisce about Sam because Sam's like part of our history, you know, the whole time. And actually, Carolyn wouldn't be sitting there if it weren't for Sam because Carolyn was Sam's lawyer. That's right. That was how we met you. He wasn't seen me. He was Sam's lawyer because Sam was like the first startup that was so successful. He actually had a lawyer. That's right. So Sam, we had known for years and Sam was always in the picture, even when he was working on Looped. He would still come to dinner. He was always around. He was always around and he definitely was our go-to.

38:38Paul Graham:Like you need fundraising advice, go talk to Sam. Yeah. He was the best. Even then, I mean, now, my God, he's the God of fundraising. But even early on, he was like among the best fundraisers in Silicon Valley. Tell me what your opinion of him was in the very early days, in the first batch. Well, Sam, some people that YC has funded, I gradually realized were a big deal. With Sam, it was instant. Like Sam, like in the first minute of talking to him, I can remember thinking. I can remember what I thought at that time. I thought, okay, this must be what Bill Gates was like when he was 19 because Sam was 19 then.

39:20And I had always wondered, like Bill Gates ended up starting Microsoft when he was 19. What could he have been like? Because I certainly was not in a position to start Microsoft when I was 19. And then I met Sam and I'm like, okay, this is what one of these 19-year-olds is like, right? Um, he seemed, he seemed like he was 40, you know, just like this look 19. And so even that batch, like he was, he was the big dog, you know, he was the one who got Sequoia to invest in him. He was the one who raised serious money. Um, already he sort of acted like he was, if that, if that had been a high school class, he would have been the class president, you know, that's sort of what it was like.

40:04even during the batch. And then after that, he was always around, you know, we would go and visit, we'd go and visit looped and, or he would be around YC and then he became available. You know, he's like, I think he got bought, he got bought and he had to go and work for the acquirer for some time, but then, then he was free. Um, and so Sam being available at the same time, um, I was thinking of retiring. I thought the obvious choice, um, get Sam. And I was, right about one thing. I thought to myself, Sam gets whatever he wants in life. If the only way he can succeed is by making Y Combinator succeed, then Y Combinator cannot fail.

40:45And so it's not like I considered lots of options for who to be president, right? Like Sam was obviously it. And I even spent a year convincing him because at first he didn't want to do it. And the mistake that I made, the mistake that I made was that if Y Combinator is the only way Sam can succeed, right? Because he has gotten what he wanted. But what I didn't, what I should have said when I hired him or when I got him to replace me, what I should have said, by the way, I want you to take over YC. And also you could only do YC. You're not allowed to work on anything else. Well, he might have said no if you had phrased it that way.

41:25No, no. I think you would have taken it. I think at that point, because at that point, he didn't know what the other things he wanted to work on were. All these other schemes he cooked up were after taking over YC, right? So at that point, it was just like, and there's lots of room to become giant by taking over YC.

41:40Paul Graham:And some of his schemes were YC schemes. Carolyn's familiar with this. Yes. The basic universal basic income thing or rebuilding cities. There were all these side projects. Yeah. Yeah. One of which was OpenAI. OpenAI, right. I know you don't like us mentioning this. Oh, no, I don't mind that. It makes me sad because my heart's at YC. And I totally get Sam. And I totally get why he got distracted by other things he wanted to do. He was young and he had like a million ideas. And some of them are amazing. AI is actually a huge deal. It's a huge deal. Like who are we to sit here as armchair quarterbacking this piece of it?

42:32Like I'm, I'm glad, but I also, you know, it, at the time it made me sad. Okay. I'm going to circle back to the shoemakers, children, never having shoes. It does. It always feels that like the little bit, but it felt like that a lot at the time because Sam was so distracted by these other things that really, really interested him that it felt like we were like the barefoot child that didn't get very much love. That's how it felt, honestly. No, no, I know, I know, right? And like the myth, the myth out there is that like we fired Sam, right, from YC, right? And really all like if Sam had said, okay, this AI thing, I'll get somebody else to do this so I can focus entirely on YC, we would have been psyched.

43:18We would have He did pick. He picked OpenAI.

43:22Paul Graham:The truth was, also, he never thought the OpenAI thing would grow to the point where it was. And all of a sudden - You know how you can tell he never knew how big it was going to be? That he started it as a nonprofit. I mean, what a disaster that has been for him. I think there's people out there who think like Sam was pulling a fast one by starting it as a nonprofit. My God, his life would have been so much easier if he started as a for-profit company. And it's not like he needed it to be a nonprofit to get the money, Sam was the world's best fundraiser. Sam could easily have raised enough money to do it as a for-profit thing.

43:56It just, you know what it was, and this is actually quite revealing. Nobody knew then, because this was like 2015 or something when he started OpenAI, right? Nobody knew then that it would be worth spending so much on AI. Isn't that interesting? This was before there was anything that you could do with a million GPs, right? Now AI is so valuable, it's worth spending a vast amount of money on it. And therefore you have to raise more than you could possibly ever raise as a nonprofit. But the fact that you would even think you could start, you could even do this as a, you could be an AI thing as a nonprofit.

44:36It just shows you how small it seemed back then. You could fund it with donations, right? Ridiculous.

44:44Paul Graham:So OpenAI was getting bigger than he had imagined. YC was getting a little bit neglected. Press release goes out saying OpenAI is now a for-profit company and Sam's the CEO. And I remember thinking, okay, now we need to just have a heart-to-heart with Sam. And we had a heart-to-heart with Sam. We said, can't do both. We totally get that you want to work on AI. Do you agree that we need to find a successor? I remember saying that. And he was very fine about it. He said, I totally agree. And we worked on that. And we found Jeff, who was a partner at YC. And we both liked Jeff. Jeff was sort of the obvious choice, right?

45:27Because Jeff was already a YC partner. He was sort of established and legitimate and serious, you know. And he was the obvious person to take over then. But then he only did it for three years. and do you remember how bummed we were when he told us he wanted to quit like like we thought what you've only been doing it for three years right but i guess i mean he's older than me he wanted to retire yeah but then it turned out we could get gary and we never realized that was even possible but right because he was running initialized yeah gary like had one one i mean Jeff couldn't help this, but he had not done YC.

46:10And really, I think it's good if the president of YC has done YC, right? So they know what it feels like to be the founders, you know, trying to raise money or get growth or have Demo Day coming up. So Gary had done YC and was like popular with LPs. Like he was a known quantity to the LPs who trusted him because he made so much money for them and initialized. And he was like young and energetic, ready to spend too much time working on YC. I think you found out that he works too hard. So now, like, I don't know about you guys, but like, I feel like this is the best I've ever felt about YC. It used to be for like 10 years there, there was always some worry in the back of my mind, you know, some nagging worry and now it feels like don't you remember we had a we had an off we had a like board meeting and we were like wait a minute there didn't seem to be anything wrong you know that was so novel oh my god we have shoes yes we finally have shoes the shoemaker's children

47:14Paul Graham:finally got some shoes right paul carolyn you probably don't know this but earlier today paul tweeted that he was going going on the show and did anyone have any questions for him oh my God, how fun. I wonder if I should look at a couple. What do you guys think? Definitely. Can't hurt. We know that Twitter reply threads are full of fabulously valuable insights. Maybe there's some good questions in there. Okay. Well, there are 149 responses. So I'm going to just read a couple, Paul, and you can sort of choose. I'll read. Okay. One said, can you talk about how you balanced your desire to have a family and kids with the demands of your professional pursuits.

47:56Well, that was one of the reasons I retired in 2014 because we had had a second child, right? In 2012. And I was worried. I gave, I gave our first child a huge amount of attention and I was worried I wouldn't be able to do the same for the second one, right? Now that there were two, how was I going to do this? I didn't realize that you can't give your second child a lot of attention because anything you do with the second child, the first one wants to come along as well. So I was sort of, I didn't get that then yet. But I did want to make sure I could spend a lot of time. And it's funny, you know, when people get fired, they always say like, now they want to spend more time with their family.

48:35But that really was a big part of the motivation. I wanted to spend time with family, but I also just wanted to work on other things, you know, and that was the trouble with YC. If you're a YC partner, you can have side projects, But if you're president, it's like it takes up all your time and more. So the way I would combine it. But, you know, the thing is, YC was never 9 to 5. You didn't, like, go to this office and stay there all day long and then come home and, like, kiss your kids and then they go to bed. Right? And so I was often at home. I was often at home, like, hanging out, playing with the kids.

49:10It was well-suited to that, well-suited to having kids. And I want it to continue to be well-suited. for having kids, right? This is something like running a startup is not super compatible with having kids. I mean, people do it, but boy, they don't fit together nicely. I would like it if being a YC partner did fit together nicely with having kids. Yeah. All right. What else you got on that Twitter thread?

49:35Paul Graham:Okay. Here's a good one. How do you keep a sane mind in the crazy world we live in? Keep writing smart essays and stay relevant with my skills at the same time. How do I keep a sane mind? Well, it's important to be married to someone sane. Yeah. I mean, it sounds like a strange compliment to describe someone as sane, but the older you are, the more you realize that's actually a fairly unique quality, right? And so if you're married to someone, And as long as you don't both freak out at the same time, then there's always someone to calm the other one down, right? That's the advantage. So I recommend to everyone, marry someone sane.

50:27Paul Graham:Here's a good one. I'd love to know what you've been surprised about in the last few years that goes against your essays. You've spoken about vibe coding being surprising, for example. Yeah. Yeah. Well, everyone is surprised at the rate of progress in AI. I didn't think this would be happening in my lifetime. It didn't look in 2010 like we would have. I mean, it's, you know, we've gotten to the point where you have to go back and read the definition of the Turing test. We're that close. So, I mean, we have stuff that seems to me, maybe you'd have to really try hard to make it fail the Turing test.

51:05You know, at any given time, it's pretty much passing the Turing test with me. In fact, the weird thing is how normal this now seems. Because when I was a kid, I was like very interested in AI. And I would dream about this future where you'd have this computer that would pass the Turing test, you'd talk to it, it would be like talking to a person. And now you ask ChatGPT some boring question, it gives you the answer. And it's like, okay, it's so weird how we can adjust to everything. It just seems like not a big deal that we're like passing the Turing test all the time now. Has there been anything else that falsified?

51:41Well, I'm shocked at how young founders are. I don't like this trend at all. I don't like this weird thing where they like want to go start a startup instead of going to college. I don't like that at all. So when you say young, in this case, you mean like - 18. Yeah, like 17, 18 versus 22, 23. That's what you mean by young. Right. Like, I mean, Sam dropped out after two years, right? So Sam was 19 when he started the startup. Big difference between that and like not even going to college. So that's a bit shocking. I sort of wish I could put the brakes on that one. Maybe I can. Maybe I can. I'm thinking of writing an essay called Go to College that explains why it's a good idea to go to college.

52:24so what else has surprised me ai the youth of founders i'm surprised at valuations like demo day valuations are like 60 million dollars now right oh my gosh i don't think they're that high but yeah probably they are they are i've heard of multiple startups raising money at a 60 cap um that's 20 times the value the post money valuation of airbnb 20 times yeah Yeah. There's just so much money now. It's a demo day.

52:56Paul Graham:This is a related question. Which of your previous startup advice is different, if any, given AI? It hasn't changed things super much yet. I mean, it might get to the point where starting a startup is structurally different. At the moment, the main effect AI has had is it's a source of startup ideas. It's a source of things that founders work on. They work on some AI version of some product for some market, right? Or AI is the thing that gives them enough leverage to displace the incumbent. The world has changed enough because of AI. We haven't gotten to the point yet where you can't start certain startups or you can't start startups because as soon as you start them, the model companies set some AI founders to duplicate what you're doing.

53:45What were those German guys who would copy startups in the hope that you would buy them?

53:49Paul Graham:Oh, the Samwar brothers? Samwars, yeah, yeah, yeah. So, you know, now we have to worry about the Sams instead of the Samwars. Um, like, ha ha ha. Will, you know, the model companies start, like, sending out armies of, armies of AIs to, like, copy your startup? One day, maybe. One day, you know, that would be worrying. I sort of wonder what YC's model would be in that world. But here's something encouraging. You know what the best source of startup ideas is, your own needs, right? Think about it. This was Facebook, Google, Airbnb. All these companies were originally started to satisfy the founder's own needs.

54:43AIs don't have any needs. the AIs aren't like desperate to pay the rent in their apartment and there's this conference and they think okay let's rent out some airbeds on our floor right that's right AIs aren't desperate to know like what the other kids in their year at Harvard are up to so maybe maybe however good AI gets at being intelligent it won't replace all the human founders because it doesn't have the needs that lead to startup ideas. Isn't that interesting?

55:14Paul Graham:I have a question. Obviously, we live in England. We went back to America for a few weeks this summer in August, for the month of August, basically. What was your impression of the startup world? Well, every time we go back, there's more AI. A larger percentage of the companies are doing some AI idea and they're using AI to write a larger percentage of their code. And OpenAI is bigger. Every time we go back, it's like the same but more. The big surprise was something I already mentioned, how many super young founders there are. It seemed like I was constantly doing office hours that began with 15 minutes of, are you sure you wouldn't like to go back to college?

56:04It's funny. People imagine YC is determined to force all these people to do startups so we can make money out of them. And here I am constantly trying to talk them into ditching the startup and going back to school. What do they answer you when you say, please go back to college, you should go back? Are they just like, no, the time is now? Because that's what I'm imagining. No, it varies. It varies from one extreme to the other. There's sometimes people who are like, no, this is what I'm going to work on. Like that 16-year-old guy, Raghav, right? And he was probably right. Yeah. He was a very unusual guy.

56:42He was just meant to do this. You could see. And then there would be other people I'd talk to and they were like, yeah, we were sort of wondering if we should go back to school.

56:51Paul Graham:I think none of them will go back to school unless their startup fails, Paul. I hate – I thought that all the time. I'm like talking to our older son being like – Some of them should. He's trying to convince them to go back to school and they're not going to. Some of them should. And I wish the ones who should would. There are people working on this B2B thing where you have to, I mean, B2B, you have to be older, right? You have to be able to like sound, seem convincing. There are these women starting this startup who are 18 and like, they looked like they were 18. You know, they're going to have a hard time walking in trying to convince some CIO to purchase their, you know, purchase seat licenses for their thing.

57:29Whereas if they just did the same idea in a couple of years when they could seem older and more convincing, they could do the same thing and it would work better. Unless someone else does it. Yeah.

57:44Paul Graham:Yeah. That's always the risk. Other people are always doing it. That's true. They could work on it in the meantime. They just don't have to try and be a company. And there's other ideas. People say, sometimes, I'll tell you, sometimes people would say, oh, I feel like if I don't do it now, it'll be too late. like the AI revolution is happening now. And I said, you know what? People said that two years ago. Were they right? Is it too late now? They'd say, well, no, it's never too late. You know, if it ever is too late because of the AI revolution, if it's really too late to ever start any kind of startup, like we're all screwed.

58:20Think what that would have fun. Yeah, I was going to say, we're in trouble. You better get like a bunch of freeze dried food and a shotgun in a bunker somewhere. nervous laughter.

58:31Paul Graham:And we cannot end on that. You know how I feel about ending on a like really freaky note. How should we end C. Levy? How should we end on this? How should we end? We could talk about how fabulous YC is now. One reason it's interesting to talk about how well YC is doing is because YC is sort of like Reddit in the sense that You know, Reddit had a long period where it wasn't really progressing because the founder, because like Steve wasn't even there. Right. But it was such a powerful model that it didn't die. It kept growing anyway. And YC was kind of like that. There was like long periods where no one was paying that much attention to running it.

59:13Right. And because it was so fundamentally strong, it survived. And so everybody thought they had Reddit pigeonholed. You know, they thought, OK, it's like sort of a big deal, but not really. It's like a small sovereign state like Luxembourg or something in the sovereign state. But almost doesn't matter. Boy, all the Luxembourgers are going to be mad at me.

59:34Paul Graham:Catching strays. All 300 ,000 of them. We make it back to California and you in the kitchen all day long are having office hours with startups, much to my dismay when I'm trying to make lunch for the kids. Or when they see how messy our house is. Let's not tell our listeners about that. That's my little secret. It's my secret. Yes. So we come back. What makes you – because you have this infinite amount of enthusiasm and energy when it comes to advising startups. What do you get excited about when we come back and you re-engage with Y Combinator? Well, the reason I have so much energy is not because I'm benevolent and I have infinite desire to help them.

1:00:22I mean, I want to help them. But the thing that really makes me energetic is that I'm interested in their ideas. It's like imagine if you were into puzzles of some kind and people came and supplied you with a constant stream of new puzzles, right? That's what it's like. They're like, here's what we're working on. How can we get users? What could this idea be? They don't usually ask me that. I just tell them. And so that's why I have so much energy. That's why I always liked working on YC. That's the part I like. I don't like arguing with people who are trying to maltreat the startups.

1:01:01Paul Graham:That's Gary's job now. Yeah, yeah. Poor Gary. only someone who's done it knows what it's like. But I love, I don't think I would ever get bored of talking to founders about their ideas. Yeah. And YC, it's so energizing. YC's doing so well. It's really fun. And there's a sort of, a lot of startup mojo floating around YC that's not all that common in the English countryside. Exactly. It seems excitingly exotic. Well, I think this was a really fun conversation. We did some reminiscing, some predictions, some myth busting. There's something in it for everyone. Thank you for coming on again. It's always fun to talk to you.

1:01:53Paul Graham:Like you do every day. Yeah. I was going to say, I'll see you downstairs for supper. What are you cooking me? What do you want? Orange chicken. He's making you chicken. Orange juice chicken. Never. What do you want me to make? I would like something very healthy. Carolyn, I know this is probably way too much detail for you, but we have parsnips from our garden. Nice. And leeks from our garden. Oh, that's great. And so I was sort of thinking you could make something with parsnips and leeks. Yeah, sure. So let's get going. Okay. Sign off so you can get cooking. Bye, Carolyn. Bye, guys. Bye. Bye.