WEBVTT 00:00:00.000 --> 00:00:01.620 Okay, so if I were to ask you to close your eyes 00:00:01.620 --> 00:00:04.980 for a second and picture the single most influential 00:00:04.980 --> 00:00:08.759 entity in the global economy, say, over the last 00:00:08.759 --> 00:00:11.380 20 years. That's a big question. It is, right. 00:00:11.519 --> 00:00:14.279 But I don't mean a specific person, not Elon 00:00:14.279 --> 00:00:16.719 Musk or Steve Jobs, and not an app on your phone. 00:00:16.780 --> 00:00:19.820 I mean the actual engine room, the machine behind 00:00:19.820 --> 00:00:23.579 the scenes. What image comes to your mind? Well, 00:00:23.620 --> 00:00:25.820 I think for most people, the immediate image 00:00:25.820 --> 00:00:29.969 is probably a trading floor in New York. You 00:00:29.969 --> 00:00:32.829 know, Wall Street. Right. Or maybe a huge server 00:00:32.829 --> 00:00:35.130 farm humming away somewhere in the desert. Right. 00:00:35.189 --> 00:00:37.149 Or maybe the boardrooms of a central bank, you 00:00:37.149 --> 00:00:38.890 know, setting interest rates. Exactly. Something 00:00:38.890 --> 00:00:41.289 institutional, something massive and, I guess, 00:00:41.329 --> 00:00:43.810 old. But what's so fascinating, the reason we're 00:00:43.810 --> 00:00:46.149 doing this deep dive, is that if you actually 00:00:46.149 --> 00:00:48.149 look at the data, if you look at the sort of 00:00:48.149 --> 00:00:51.590 source code of how modern life has been reengineered, 00:00:51.670 --> 00:00:54.990 the answer isn't a bank. It isn't a government. 00:00:55.210 --> 00:00:58.140 No. It's an organization that started. Basically, 00:00:58.200 --> 00:01:00.479 as a summer project, it's just two letters, Y 00:01:00.479 --> 00:01:03.399 and C. Y accommodator. Exactly. And look, I know 00:01:03.399 --> 00:01:04.959 everyone listening has probably seen the name. 00:01:05.019 --> 00:01:08.599 You've seen that bright orange Y logo on Twitter, 00:01:08.859 --> 00:01:10.819 or you've heard it dropped in conversation at 00:01:10.819 --> 00:01:13.019 a coffee shop. Oh, yeah, for sure. It's part 00:01:13.019 --> 00:01:15.299 of the lexicon now. But the reason we're digging 00:01:15.299 --> 00:01:17.780 in today is because I honestly don't think people 00:01:17.780 --> 00:01:21.040 fully grasp the sheer scale of the impact here. 00:01:21.500 --> 00:01:24.340 We are not just talking about some exclusive 00:01:24.340 --> 00:01:27.140 club for coders. No, not at all. We are talking 00:01:27.140 --> 00:01:30.530 about an entity that is the engine behind. over 00:01:30.530 --> 00:01:33.650 5 ,000 companies. 5 ,000? And to put a really 00:01:33.650 --> 00:01:36.329 concrete number on that impact, if you look at 00:01:36.329 --> 00:01:39.109 the combined valuation of those companies, as 00:01:39.109 --> 00:01:41.549 of the latest data we have, which is coming into 00:01:41.549 --> 00:01:46.629 2025, it is sitting at $600 billion. $600 billion. 00:01:47.349 --> 00:01:48.969 That's just a staggering number. I mean, that 00:01:48.969 --> 00:01:51.450 is larger than the GDP of many, many countries. 00:01:51.569 --> 00:01:53.409 It is. It's a nation -state level valuation. 00:01:53.769 --> 00:01:55.489 And it was generated essentially from scratch 00:01:55.489 --> 00:01:58.480 starting just 2005. And what makes that number 00:01:58.480 --> 00:02:00.700 even more impressive, I think, is the breadth 00:02:00.700 --> 00:02:03.159 of it. We aren't just talking about, you know, 00:02:03.180 --> 00:02:05.620 obscure enterprise software that nobody's ever 00:02:05.620 --> 00:02:07.400 heard of, although they have plenty of that, 00:02:07.439 --> 00:02:09.900 too. We're talking about brands that have become 00:02:09.900 --> 00:02:14.400 verbs. They're woven into our daily lives. Let's 00:02:14.400 --> 00:02:16.879 just run down a quick list because it's basically 00:02:16.879 --> 00:02:19.379 a greatest hits of the Internet age. You have 00:02:19.379 --> 00:02:22.439 Airbnb. Which completely, I mean, just completely 00:02:22.439 --> 00:02:25.020 revolutionized the hospitality industry and how 00:02:25.020 --> 00:02:27.159 we think about travel. Totally. Then you've got 00:02:27.159 --> 00:02:29.900 Coinbase. Which, for better or worse, brought 00:02:29.900 --> 00:02:31.879 cryptocurrency into the mainstream conversation. 00:02:32.060 --> 00:02:34.960 It was the front door for millions of people. 00:02:35.180 --> 00:02:38.580 And DoorDash and Instacart. Yeah, which fundamentally 00:02:38.580 --> 00:02:41.539 changed the logistics of food and grocery delivery. 00:02:41.740 --> 00:02:43.259 You know, especially during the pandemic, that 00:02:43.259 --> 00:02:46.360 was a lifeline. It was. Then you have Dropbox, 00:02:46.460 --> 00:02:50.379 Reddit, Stripe, Twitch. I mean, just stop on 00:02:50.379 --> 00:02:53.180 Stripe for a second. Stripe alone powers a huge 00:02:53.180 --> 00:02:55.439 percentage of all Internet commerce. It's the 00:02:55.439 --> 00:02:58.099 plumbing. If Stripe goes down, a big chunk of 00:02:58.099 --> 00:03:01.719 the online economy just stops. And Reddit is 00:03:01.719 --> 00:03:03.699 the front page of the Internet. So this isn't 00:03:03.699 --> 00:03:05.699 just a portfolio of companies. It's like the 00:03:05.699 --> 00:03:07.879 infrastructure of the modern digital economy. 00:03:08.000 --> 00:03:09.919 It really is. I saw a thought experiment once 00:03:09.919 --> 00:03:12.219 that said if you woke up tomorrow and every single 00:03:12.219 --> 00:03:16.180 YC company was just gone. Yeah. The Internet 00:03:16.180 --> 00:03:18.439 as we know it would fundamentally break. A lot 00:03:18.439 --> 00:03:20.830 of it just wouldn't work. So that is our mission 00:03:20.830 --> 00:03:22.389 for this deep dive. We're going to move beyond 00:03:22.389 --> 00:03:24.590 the buzzwords. We're going to try to ignore the 00:03:24.590 --> 00:03:27.229 valuation hype and the celebrity founder worship 00:03:27.229 --> 00:03:29.629 for a bit. And we're going to try to understand 00:03:29.629 --> 00:03:32.469 the machine itself. How does a firm that started 00:03:32.469 --> 00:03:36.330 in 2005 basically industrialize the concept of 00:03:36.330 --> 00:03:38.930 startup success? We want to unpack the mechanism. 00:03:39.389 --> 00:03:41.169 We're going to look at their deal, how the money 00:03:41.169 --> 00:03:44.050 actually works, you know, the fine print that 00:03:44.050 --> 00:03:45.650 every single founder signs. We're going to walk 00:03:45.650 --> 00:03:48.330 through the fascinating evolution of their leadership 00:03:48.330 --> 00:03:52.310 from the founders to the big expansion era and 00:03:52.310 --> 00:03:54.770 then back again. And this is important. We're 00:03:54.770 --> 00:03:56.889 also going to look at the experiments that didn't 00:03:56.889 --> 00:04:01.409 work, the glitches, the failed expansions, because 00:04:01.409 --> 00:04:03.930 I think those failures tell us just as much about 00:04:03.930 --> 00:04:06.550 their philosophy as the huge successes do. Oh, 00:04:06.550 --> 00:04:08.939 for sure. The pivots and the dead ends are where 00:04:08.939 --> 00:04:11.599 you learn the most. And for all this, we've got 00:04:11.599 --> 00:04:13.860 a great set of source materials. We're working 00:04:13.860 --> 00:04:16.660 off a really comprehensive historical and operational 00:04:16.660 --> 00:04:19.319 overview of the company. And it tracks the timeline 00:04:19.319 --> 00:04:23.220 right from the founding days in March 2005 all 00:04:23.220 --> 00:04:25.579 the way up to the present day restructuring. 00:04:25.699 --> 00:04:28.060 So we have the full arc of the story. We do. 00:04:28.399 --> 00:04:32.000 So let's rewind the clock. It's March 2005. The 00:04:32.000 --> 00:04:34.899 dot -com bubble burst a few years ago. The tech 00:04:34.899 --> 00:04:38.110 world is actually... Kind of quiet. Yeah, it's 00:04:38.110 --> 00:04:40.709 not the booming gold rush atmosphere we think 00:04:40.709 --> 00:04:42.750 of today. Capital had really dried up. It was 00:04:42.750 --> 00:04:44.930 a tough time to raise money. And into that environment, 00:04:45.050 --> 00:04:47.290 Y Combinator is born. So who are the architects 00:04:47.290 --> 00:04:49.009 of this thing? Who's in the room at the beginning? 00:04:49.339 --> 00:04:51.459 So you have a founding team of four people. You 00:04:51.459 --> 00:04:53.660 have Paul Graham, who is, you know, often the 00:04:53.660 --> 00:04:55.339 public face of the group. He's kind of the philosopher 00:04:55.339 --> 00:04:58.180 king. He's the essayist, the guy who articulates 00:04:58.180 --> 00:05:01.600 the vision in these long, thoughtful posts. Exactly. 00:05:01.600 --> 00:05:04.100 He sets the intellectual tone. Then you have 00:05:04.100 --> 00:05:06.319 Robert Tappan Morris and Trevor Blackwell. And 00:05:06.319 --> 00:05:08.670 these two are the, like... The deep technical 00:05:08.670 --> 00:05:10.730 heavyweight. Absolutely. I mean, Robert Tappan 00:05:10.730 --> 00:05:13.329 Morris is a legend in computer science. He's 00:05:13.329 --> 00:05:16.170 famous or, you know, maybe infamous for creating 00:05:16.170 --> 00:05:18.470 the Morris worm back in the 80s. Which is basically 00:05:18.470 --> 00:05:21.310 the first major worm to spread across the Internet. 00:05:21.430 --> 00:05:23.250 So he knows networks inside and out. He knows 00:05:23.250 --> 00:05:25.089 them on a fundamental level. And Trevor Blackwell 00:05:25.089 --> 00:05:27.569 is similarly brilliant, especially in robotics 00:05:27.569 --> 00:05:30.689 and engineering. So you have this core of deep, 00:05:30.889 --> 00:05:33.930 deep technical expertise. But then you have the 00:05:33.930 --> 00:05:36.879 fourth founder. Jessica Livingston. Yes. And 00:05:36.879 --> 00:05:38.680 I want to pause here and really dig into this 00:05:38.680 --> 00:05:41.720 for a second, because our source material explicitly 00:05:41.720 --> 00:05:45.480 mentions that in a lot of the early retellings 00:05:45.480 --> 00:05:48.720 of YC's history, Jessica's role is often overlooked. 00:05:48.980 --> 00:05:51.480 It's true. People focus on the hackers, the code, 00:05:51.540 --> 00:05:53.740 the algorithms. But when you look at the facts, 00:05:53.920 --> 00:05:57.339 overlooking her role seems like just a massive 00:05:57.339 --> 00:06:01.050 misunderstanding of what YC actually is. It is 00:06:01.050 --> 00:06:03.930 a huge oversight. I mean, if you talk to the 00:06:03.930 --> 00:06:06.290 alumni from those really early days, or if you 00:06:06.290 --> 00:06:08.310 look at how the organization actually functioned 00:06:08.310 --> 00:06:10.829 day to day, Jessica Livingston was absolutely 00:06:10.829 --> 00:06:13.529 pivotal. In what way? Well, while Paul Graham 00:06:13.529 --> 00:06:15.490 was writing these philosophical essays about 00:06:15.490 --> 00:06:18.089 lists and the future of web applications, Jessica 00:06:18.089 --> 00:06:20.129 was the one handling the day to day operations. 00:06:20.629 --> 00:06:23.550 But it was way more than just logistics. She 00:06:23.550 --> 00:06:25.800 was the social engineer? That's the perfect term 00:06:25.800 --> 00:06:29.279 for it. She shaped the culture. She was heavily, 00:06:29.279 --> 00:06:31.860 heavily involved in selecting the founders. And 00:06:31.860 --> 00:06:34.480 you have to remember, YC isn't just buying a 00:06:34.480 --> 00:06:37.019 piece of software. The software changes a dozen 00:06:37.019 --> 00:06:40.660 times. They are investing in people. In human 00:06:40.660 --> 00:06:43.199 beings and having someone with that high level 00:06:43.199 --> 00:06:46.000 intuition for character, for resilience, for 00:06:46.000 --> 00:06:48.899 group dynamics. That was Jessica's superpower. 00:06:49.259 --> 00:06:51.480 So while the guys were looking at the code and 00:06:51.480 --> 00:06:53.600 the technical feasibility, she was looking at 00:06:53.600 --> 00:06:55.899 the founding team and asking, are these people 00:06:55.899 --> 00:06:57.680 going to kill each other in three months when 00:06:57.680 --> 00:07:00.860 the stress hits? Exactly. Do they have the grit 00:07:00.860 --> 00:07:03.920 to survive when everything inevitably goes wrong? 00:07:04.410 --> 00:07:07.550 That social radar was the secret sauce. It's 00:07:07.550 --> 00:07:09.529 what turned YC from just being an investment 00:07:09.529 --> 00:07:12.470 fund into a real community. And you could argue 00:07:12.470 --> 00:07:15.050 that the reason YC alumni are so willing to help 00:07:15.050 --> 00:07:17.029 each other today is because of that cultural 00:07:17.029 --> 00:07:19.550 DNA that she implanted right at the very beginning. 00:07:19.790 --> 00:07:22.170 I think that's 100 % correct. She set the tone. 00:07:22.430 --> 00:07:25.490 Now, speaking of those early batches, the operational 00:07:25.490 --> 00:07:28.709 model they started with sounds... Well, it just 00:07:28.709 --> 00:07:30.870 sounds exhausting. It wasn't the streamlined, 00:07:31.129 --> 00:07:33.410 centralized campus we think of today. Not at 00:07:33.410 --> 00:07:35.209 all. They were running a tale of two cities. 00:07:35.449 --> 00:07:37.569 It really does sound like a logistical nightmare. 00:07:38.009 --> 00:07:40.389 Initially, they were running two concurrent programs. 00:07:40.670 --> 00:07:42.930 They had one based in Cambridge, Massachusetts, 00:07:42.930 --> 00:07:45.170 and one in Mountain View, California. They were 00:07:45.170 --> 00:07:47.389 trying to straddle both coasts. You have the 00:07:47.389 --> 00:07:50.850 sort of academic, intellectual hub of Harvard 00:07:50.850 --> 00:07:53.350 and MIT on the East Coast. Right. The old school 00:07:53.350 --> 00:07:55.740 tech scene. And then... the burgeoning venture 00:07:55.740 --> 00:07:57.699 capital heartland on the West Coast. It makes 00:07:57.699 --> 00:07:59.879 sense on paper, I guess. It does, but you have 00:07:59.879 --> 00:08:02.399 to contextualize it. It's 2005. We didn't have 00:08:02.399 --> 00:08:05.379 Zoom. We didn't have Slack. Remote work, as we 00:08:05.379 --> 00:08:08.040 know it, just didn't exist. True. If you wanted 00:08:08.040 --> 00:08:10.399 to manage a cohort of startups and build that 00:08:10.399 --> 00:08:12.639 culture we were talking about, you had to be 00:08:12.639 --> 00:08:14.980 in the room with them. So the partners were literally 00:08:14.980 --> 00:08:17.959 flying back and forth across the country constantly. 00:08:17.980 --> 00:08:20.920 They were moving their entire lives twice a year 00:08:20.920 --> 00:08:23.540 to manage these two different groups. Yeah, it 00:08:23.540 --> 00:08:25.899 was incredibly draining. And the source material 00:08:25.899 --> 00:08:28.899 puts it pretty politely. It just calls it operational 00:08:28.899 --> 00:08:31.839 complexities. Which is corporate speak for this 00:08:31.839 --> 00:08:35.659 is killing us. Pretty much. So in January 2009, 00:08:36.100 --> 00:08:39.200 they made a really critical decision. They closed 00:08:39.200 --> 00:08:41.679 the Cambridge program for good. That feels like 00:08:41.679 --> 00:08:43.820 a watershed moment. It's like they had to choose 00:08:43.820 --> 00:08:46.340 a home. They had to plant a flag. They did. And 00:08:46.340 --> 00:08:49.159 they chose Silicon Valley. They realized that 00:08:49.159 --> 00:08:51.100 if they wanted this thing to scale, if they wanted 00:08:51.100 --> 00:08:52.860 the machine to work really efficiently, they 00:08:52.860 --> 00:08:54.620 had to centralize. And they had to be where the 00:08:54.620 --> 00:08:56.980 money was. More importantly, they realized where 00:08:56.980 --> 00:08:59.580 the center of gravity was. The venture capital, 00:08:59.759 --> 00:09:02.080 the big tech companies who might acquire you, 00:09:02.340 --> 00:09:06.019 the whole ecosystem. It was all clustering in 00:09:06.019 --> 00:09:08.179 the Bay Area. It's interesting to think what 00:09:08.179 --> 00:09:10.730 might have happened if they'd stayed split. or 00:09:10.730 --> 00:09:13.450 if they had chosen Boston, the entire trajectory 00:09:13.450 --> 00:09:16.269 of the startup world might look really different 00:09:16.269 --> 00:09:18.750 today. I think that's true. By consolidating 00:09:18.750 --> 00:09:20.789 everyone in Mountain View, they created this 00:09:20.789 --> 00:09:23.610 incredible density of talent. They forced everyone 00:09:23.610 --> 00:09:26.049 into the same pressure cooker. And that density 00:09:26.049 --> 00:09:28.629 is what really sparked the network effects that 00:09:28.629 --> 00:09:31.389 YCU is so famous for. Okay, so let's talk about 00:09:31.389 --> 00:09:34.700 the fuel for this engine. The money. In the very, 00:09:34.740 --> 00:09:38.100 very early days, YC was funding these companies 00:09:38.100 --> 00:09:40.059 essentially out of the founders' own pockets. 00:09:40.120 --> 00:09:42.299 Yeah, it was very small checks, almost like a 00:09:42.299 --> 00:09:44.639 stipend. But you can't fund 5 ,000 companies 00:09:44.639 --> 00:09:47.480 on personal savings. So when did the big money 00:09:47.480 --> 00:09:49.720 finally enter the picture and validate this whole 00:09:49.720 --> 00:09:52.419 experiment? The first major external validation 00:09:52.419 --> 00:09:55.500 came in 2009. That's when Sequoia Capital led 00:09:55.500 --> 00:09:58.299 a $2 million investment into Y Combinator itself. 00:09:58.659 --> 00:10:00.740 And for anyone who doesn't follow the venture 00:10:00.740 --> 00:10:03.700 capital scorecard, getting money from Sequoia 00:10:03.700 --> 00:10:06.659 is a very big deal. It's the ultimate seal of 00:10:06.659 --> 00:10:08.360 approval. They're like the New York Yankees of 00:10:08.360 --> 00:10:10.860 VC. Yeah. They backed Apple, Google, Oracle. 00:10:11.120 --> 00:10:13.639 If Sequoia invests, everyone else pays attention. 00:10:13.980 --> 00:10:16.120 Exactly. And it wasn't just the cash, though, 00:10:16.139 --> 00:10:19.399 that... $2 million allowed YC to start funding 00:10:19.399 --> 00:10:22.279 about 60 companies a year, which was a big step 00:10:22.279 --> 00:10:24.700 up. It was the signal it sent to the rest of 00:10:24.700 --> 00:10:26.820 the market. It said, this weird little accelerator 00:10:26.820 --> 00:10:29.100 thing, it's legitimate. This is the future. And 00:10:29.100 --> 00:10:31.019 then they doubled down pretty quickly after that. 00:10:31.059 --> 00:10:34.759 They did. In 2010, just a year later, Sequoia 00:10:34.759 --> 00:10:37.419 led another round, this time for $8 .25 million. 00:10:38.279 --> 00:10:41.440 This gave YC the firepower to really start scaling 00:10:41.440 --> 00:10:43.279 the number of companies they could accept into 00:10:43.279 --> 00:10:46.000 each batch. But the real game changer. And I 00:10:46.000 --> 00:10:47.720 remember reading about this when it happened 00:10:47.720 --> 00:10:49.960 and thinking it had to be a typo was the start 00:10:49.960 --> 00:10:53.820 fund era in 2011. Oh, that was a bombshell. It 00:10:53.820 --> 00:10:56.820 seemed just too good to be true. It was. You 00:10:56.820 --> 00:10:58.960 had the investor Yuri Milner and the firm SV 00:10:58.960 --> 00:11:02.100 Angel stepping up and making this just unprecedented 00:11:02.100 --> 00:11:05.600 offer. They said, we will offer every single 00:11:05.600 --> 00:11:09.139 Y Combinator company $150 ,000 convertible note. 00:11:09.620 --> 00:11:12.299 Automatically. You get into YC and boom, there's 00:11:12.299 --> 00:11:14.639 an extra $150 ,000 in your bank account. Essentially, 00:11:14.700 --> 00:11:17.389 yes. No extra pitching, no extra due diligence. 00:11:17.649 --> 00:11:19.710 If YC said you were good enough to be in the 00:11:19.710 --> 00:11:22.330 batch, Yuri Milner and S .V. Angel said, here 00:11:22.330 --> 00:11:26.049 is $150K. Wow. Later on, when the start fund 00:11:26.049 --> 00:11:28.570 was renewed, that amount changed to $80 ,000. 00:11:28.970 --> 00:11:31.450 But the precedent was set. It was a firehose 00:11:31.450 --> 00:11:33.970 of capital and directly at these brand new companies. 00:11:34.250 --> 00:11:36.590 That completely shifted the dynamic, didn't it? 00:11:36.750 --> 00:11:39.070 Before this, startups would go to YC, they'd 00:11:39.070 --> 00:11:40.899 get a... Tiny bit of seed money and then they'd 00:11:40.899 --> 00:11:42.919 have to spend months begging angel investors 00:11:42.919 --> 00:11:45.080 for enough cash to survive. Right. They're always 00:11:45.080 --> 00:11:47.100 fundraising. This meant they had real runway 00:11:47.100 --> 00:11:49.870 from day one. It industrialized the seed stage. 00:11:50.190 --> 00:11:53.289 It meant founders could focus entirely on building 00:11:53.289 --> 00:11:55.330 their product and talking to users for those 00:11:55.330 --> 00:11:57.750 crucial three months, rather than worrying about 00:11:57.750 --> 00:12:00.870 making rent. It just made YC the single most 00:12:00.870 --> 00:12:03.049 attractive option in the world for a young founder. 00:12:03.269 --> 00:12:06.230 Okay, so that sets the stage perfectly. We have 00:12:06.230 --> 00:12:08.850 the location settled in Silicon Valley. We have 00:12:08.850 --> 00:12:12.090 the backing of Sequoia. We have a mechanism to 00:12:12.090 --> 00:12:14.350 flood these new startups with initial capital. 00:12:14.470 --> 00:12:16.830 Let's open up the hood and look at the machine 00:12:16.830 --> 00:12:19.580 itself. How does the program actually work? The 00:12:19.580 --> 00:12:23.179 core rhythm of Y Combinator is the batch. Historically, 00:12:23.179 --> 00:12:25.559 for years, they ran two batches a year, a winter 00:12:25.559 --> 00:12:28.000 batch and a summer batch. Right. But what's interesting 00:12:28.000 --> 00:12:30.139 is that our current source material indicates 00:12:30.139 --> 00:12:33.200 that they now interview and select for four batches 00:12:33.200 --> 00:12:35.639 per year. Four batches. That is a significant 00:12:35.639 --> 00:12:38.000 increase in tempo. It really is becoming a factory 00:12:38.000 --> 00:12:40.620 line. It is. So what happens once a company is 00:12:40.620 --> 00:12:42.899 accepted into a batch? Is it like a university 00:12:42.899 --> 00:12:44.940 semester? What's the day -to -day like? Is it 00:12:44.940 --> 00:12:47.850 like a college semester? Sort of, but maybe without 00:12:47.850 --> 00:12:49.990 the grades and with a lot more existential panic. 00:12:50.429 --> 00:12:53.809 The source describes it as a mix of office hours 00:12:53.809 --> 00:12:56.490 and weekly meetups. Office hours sounds very 00:12:56.490 --> 00:12:59.769 collegiate, very academic. It does in name, but 00:12:59.769 --> 00:13:02.830 in practice, it's incredibly intense mentorship. 00:13:03.250 --> 00:13:05.809 You have these individual meetings with the YC 00:13:05.809 --> 00:13:08.210 partners, who are often very successful founders 00:13:08.210 --> 00:13:10.029 themselves. And they don't pull any punches. 00:13:10.210 --> 00:13:12.850 Yeah. They grill you. Yeah. What is your weekly 00:13:12.850 --> 00:13:15.769 growth rate? Why isn't it higher? How many users 00:13:15.769 --> 00:13:17.889 did you talk to this week? What is the single 00:13:17.889 --> 00:13:21.149 biggest blocker to your growth? It's just a relentless, 00:13:21.289 --> 00:13:23.990 obsessive focus on execution. And then the weekly 00:13:23.990 --> 00:13:25.789 meetups are the famous dinners, right? Where 00:13:25.789 --> 00:13:27.970 they bring in the big guns. Exactly. Every week, 00:13:27.970 --> 00:13:30.049 the entire batch gathers. Historically, it was 00:13:30.049 --> 00:13:32.009 for dinner. And they bring in a guest speaker. 00:13:32.190 --> 00:13:34.409 And we are talking about the absolute titans 00:13:34.409 --> 00:13:36.870 of industry. Mark Zuckerberg, the founders of 00:13:36.870 --> 00:13:40.750 Airbnb, top VCs from Andreessen Horowitz or Sequoia. 00:13:41.019 --> 00:13:43.259 And they speak off the record. That's the key. 00:13:43.480 --> 00:13:46.399 They share the war stories, the unvarnished truths, 00:13:46.659 --> 00:13:49.399 the secrets of how Silicon Valley really works. 00:13:49.720 --> 00:13:52.179 It's like downloading the collective wisdom of 00:13:52.179 --> 00:13:54.600 the valley directly into the brains of these 00:13:54.600 --> 00:13:57.419 22 -year -old founders. And what's the ultimate 00:13:57.419 --> 00:13:59.659 goal of all this? In those three months, what 00:13:59.659 --> 00:14:01.799 are they trying to get them to achieve? The source 00:14:01.799 --> 00:14:04.440 material highlights a few specific goals. First, 00:14:04.659 --> 00:14:06.559 teaching founders how to market their product. 00:14:07.000 --> 00:14:09.899 Second, refining their business model. But the 00:14:09.899 --> 00:14:12.889 most important one. The holy grail is achieving 00:14:12.889 --> 00:14:15.590 product market fit. Which basically means you've 00:14:15.590 --> 00:14:17.330 built something that a group of people desperately 00:14:17.330 --> 00:14:19.350 wants. Right. And you can prove it with data. 00:14:19.490 --> 00:14:22.090 They want you to have a graph of your key metric 00:14:22.090 --> 00:14:24.190 that is going up and to the right. Because it 00:14:24.190 --> 00:14:27.070 all leads up to one single high stakes day. Demo 00:14:27.070 --> 00:14:29.669 day. Demo day. The graduation ceremony. But with 00:14:29.669 --> 00:14:33.230 much, much higher stakes. This is the day where 00:14:33.230 --> 00:14:35.429 all the startups in the batch present their business 00:14:35.429 --> 00:14:38.149 and their technology prototypes to a room. Where 00:14:38.149 --> 00:14:40.690 nowadays it's often a Zoom room full of... hundreds 00:14:40.690 --> 00:14:42.909 of potential investors. And it's designed to 00:14:42.909 --> 00:14:44.909 create a frenzy, isn't it? It's not a relaxed 00:14:44.909 --> 00:14:49.289 affair. It is engineered FOMO, fear of missing 00:14:49.289 --> 00:14:53.250 out. The entire event is structured to make investors 00:14:53.250 --> 00:14:55.470 feel like if they don't write a check today on 00:14:55.470 --> 00:14:57.830 the spot, they are going to miss out on the next 00:14:57.830 --> 00:15:00.399 Dropbox or the next Airbnb. It's social proof 00:15:00.399 --> 00:15:02.860 on a massive scale. Completely. So let's get 00:15:02.860 --> 00:15:05.100 to the brass tacks. If I'm listening to this 00:15:05.100 --> 00:15:07.080 right now and I have a billion dollar idea and 00:15:07.080 --> 00:15:10.500 I decide to apply to YC, if I get in, what is 00:15:10.500 --> 00:15:13.000 the standard deal? Because I know this has changed 00:15:13.000 --> 00:15:15.500 a lot over the years and understanding the money 00:15:15.500 --> 00:15:18.110 is. absolutely crucial. It is. And the current 00:15:18.110 --> 00:15:20.429 structure, which was introduced in January of 00:15:20.429 --> 00:15:24.029 2022, is quite specific. And it's important to 00:15:24.029 --> 00:15:25.730 break it down. The headline number they talk 00:15:25.730 --> 00:15:28.809 about is $500 ,000. Okay. Half a million dollars. 00:15:29.070 --> 00:15:31.769 That is serious seed money. That's not just ramen 00:15:31.769 --> 00:15:34.029 money. You can hire a team with that. You can 00:15:34.029 --> 00:15:35.870 build a real product. For sure. But obviously 00:15:35.870 --> 00:15:37.690 they aren't just giving it away for free. Right. 00:15:37.850 --> 00:15:39.610 What do they take in return? It's split into 00:15:39.610 --> 00:15:41.730 two distinct buckets. And this is where it gets 00:15:41.730 --> 00:15:44.639 really interesting. The first bucket is $125 00:15:44.639 --> 00:15:49.500 ,000. In exchange for that $125 ,000, YC takes 00:15:49.500 --> 00:15:52.940 7 % equity in your company. And this is done 00:15:52.940 --> 00:15:55.860 on what's called a post -money safe. OK, let's 00:15:55.860 --> 00:15:57.259 just pause there and do the math for a second. 00:15:57.340 --> 00:16:01.980 If they're getting 7 % for $125 ,000, that effectively 00:16:01.980 --> 00:16:05.419 values your brand new, maybe pre -product company 00:16:05.419 --> 00:16:09.019 at roughly $1 .8 million. Roughly, yes. That's 00:16:09.019 --> 00:16:11.600 the implied valuation. Now, for some really hot 00:16:11.600 --> 00:16:13.519 startups coming in, that might feel like a low 00:16:13.519 --> 00:16:15.120 valuation. They might feel they're giving away 00:16:15.120 --> 00:16:17.059 too much equity. So what's the justification? 00:16:17.789 --> 00:16:20.669 YC's argument is that the 7 % is the price of 00:16:20.669 --> 00:16:23.549 admission. It's the tax you pay to get into the 00:16:23.549 --> 00:16:25.769 network, to get that Sequoia seal of approval 00:16:25.769 --> 00:16:27.889 we talked about, to get access to the alumni 00:16:27.889 --> 00:16:30.230 database, the advice, the demo day platform. 00:16:30.450 --> 00:16:31.950 You're paying for the brand and the network, 00:16:32.029 --> 00:16:34.210 not just the cash. Exactly. But that's only the 00:16:34.210 --> 00:16:36.149 first part. What about the other chunk? That 00:16:36.149 --> 00:16:39.289 leaves $375 ,000. Right. How does that part work? 00:16:39.450 --> 00:16:41.230 That is the really fascinating part of the new 00:16:41.230 --> 00:16:44.710 deal. The remaining $375 ,000 comes on an uncapped 00:16:44.710 --> 00:16:47.309 safe with an MFN clause. Okay, you just... dropped 00:16:47.309 --> 00:16:48.950 a bunch of jargon on us. We need to unpack that. 00:16:49.049 --> 00:16:51.950 Let's start with MFN. What is that? MFN stands 00:16:51.950 --> 00:16:55.610 for Most Favored Nation. And yeah, it sounds 00:16:55.610 --> 00:16:58.309 like something from an international trade treaty. 00:16:58.490 --> 00:17:00.990 It does. What does it mean for a founder in a 00:17:00.990 --> 00:17:03.309 hoodie who's just trying to write code? It's 00:17:03.309 --> 00:17:05.750 actually a very, very clever mechanism for YC. 00:17:06.279 --> 00:17:09.680 Simply put, an MFN provision means this. We are 00:17:09.680 --> 00:17:13.079 giving you this $375 ,000 now, but we are not 00:17:13.079 --> 00:17:15.559 going to set a price evaluation on it yet. Okay. 00:17:15.640 --> 00:17:17.720 Instead, we agree that whenever you go out and 00:17:17.720 --> 00:17:19.880 raise your next round of funding from other investors, 00:17:20.119 --> 00:17:23.599 we get the same best terms that you give to anyone 00:17:23.599 --> 00:17:25.940 else in that round. Can we walk through a concrete 00:17:25.940 --> 00:17:27.859 example? Because that's still a little abstract. 00:17:27.960 --> 00:17:30.079 Sure. Let's say you're a founder. YC gives you 00:17:30.079 --> 00:17:33.759 that $375 ,000 on an MFN safety. You go through 00:17:33.759 --> 00:17:36.240 the three -month program. You build your product. 00:17:36.420 --> 00:17:38.640 You get traction. You get your chart going up 00:17:38.640 --> 00:17:40.579 and to the right. Exactly. You get to demo day. 00:17:40.799 --> 00:17:43.220 Yeah. And you find a lead investor. Let's say 00:17:43.220 --> 00:17:45.500 it's a big firm like Andreessen Horowitz. And 00:17:45.500 --> 00:17:48.220 they love your company. They say, this is amazing. 00:17:48.319 --> 00:17:50.920 We will invest at a $10 million valuation cap. 00:17:51.099 --> 00:17:53.420 Okay. So that's the deal they offer me. That's 00:17:53.420 --> 00:17:56.400 the deal. Because of that MFN clause YC has, 00:17:56.720 --> 00:18:00.740 their $375 ,000 automatically converts into shares 00:18:00.740 --> 00:18:03.599 at that same $10 million valuation cap. They 00:18:03.599 --> 00:18:07.480 don't have to negotiate. Ah, I see. So YC doesn't 00:18:07.480 --> 00:18:10.220 have to waste time haggling over the price with 00:18:10.220 --> 00:18:12.740 hundreds of companies individually right at the 00:18:12.740 --> 00:18:15.259 beginning. Precisely. They just say, take the 00:18:15.259 --> 00:18:17.859 money. Go build something awesome. When you find 00:18:17.859 --> 00:18:20.220 a professional investor later who does all the 00:18:20.220 --> 00:18:23.259 hard work of pricing your company, we will just 00:18:23.259 --> 00:18:25.359 piggyback on whatever deal you negotiate with 00:18:25.359 --> 00:18:27.740 them. Exactly. It saves an enormous amount of 00:18:27.740 --> 00:18:30.440 time and friction. It lets YC operate at scale, 00:18:30.519 --> 00:18:32.240 and it lets the market decide the fair price 00:18:32.240 --> 00:18:35.079 later on when the company is more mature. But 00:18:35.079 --> 00:18:37.400 isn't there a risk for the founder in that? What 00:18:37.400 --> 00:18:40.720 if I negotiate a really bad low valuation because 00:18:40.720 --> 00:18:43.180 I'm desperate for cash? Well, then YC gets that 00:18:43.180 --> 00:18:45.519 great deal, too. Most favored nation means they 00:18:45.519 --> 00:18:47.819 get the best deal you give to anyone. So if you 00:18:47.819 --> 00:18:49.740 give your uncle a sweetheart deal at a super 00:18:49.740 --> 00:18:52.460 low valuation, YC gets that same deal. So it 00:18:52.460 --> 00:18:54.500 keeps the founders honest. It keeps them honest. 00:18:55.519 --> 00:18:57.900 and incentivizes them to get the best possible 00:18:57.900 --> 00:19:00.640 terms, which ultimately benefits YC as well. 00:19:00.900 --> 00:19:03.859 It's an incredibly shrewd and efficient system. 00:19:04.019 --> 00:19:06.599 That makes a lot of sense. Now, not every organization 00:19:06.599 --> 00:19:08.859 that goes through YC is trying to be a billion 00:19:08.859 --> 00:19:11.680 dollar unicorn. Some are nonprofits. I was surprised 00:19:11.680 --> 00:19:14.200 to see that on the list of alumni. How does the 00:19:14.200 --> 00:19:16.539 deal work for them? It's obviously very different 00:19:16.539 --> 00:19:19.380 because a nonprofit doesn't have equity. You 00:19:19.380 --> 00:19:22.160 can't give away 7 % of nothing. Right. So if 00:19:22.160 --> 00:19:24.700 a nonprofit is accepted, and they have accepted 00:19:24.700 --> 00:19:26.819 some really notable ones like WOTC, the healthcare 00:19:26.819 --> 00:19:30.500 platform, or Our World in Data, they simply receive 00:19:30.500 --> 00:19:34.380 a $100 ,000 donation. So it's just a grant, basically. 00:19:34.380 --> 00:19:36.940 No strings attached in terms of ownership. Exactly. 00:19:37.000 --> 00:19:38.920 And honestly, for a lot of those nonprofits, 00:19:39.259 --> 00:19:41.599 getting access to the YC network, learning how 00:19:41.599 --> 00:19:43.799 to scale an organization, how to pitch donors, 00:19:43.980 --> 00:19:46.180 how to operate. efficiently is often way more 00:19:46.180 --> 00:19:48.119 valuable than the cash itself. So that's the 00:19:48.119 --> 00:19:50.160 machine as it runs today. It's a factory for 00:19:50.160 --> 00:19:52.839 success with a very standardized, very clever 00:19:52.839 --> 00:19:56.160 deal structure. But YC wasn't always just this 00:19:56.160 --> 00:19:59.480 steady focused accelerator. There was a period, 00:19:59.579 --> 00:20:01.960 specifically starting around 2014, where things 00:20:01.960 --> 00:20:04.700 got... Well, they got really expansive. Yeah. 00:20:04.779 --> 00:20:07.480 This is the Sam Altman era. Right. So Paul Graham, 00:20:07.660 --> 00:20:09.680 the original founder and philosopher king, he 00:20:09.680 --> 00:20:13.220 stepped back in 2014 and he appoints Sam Altman 00:20:13.220 --> 00:20:16.059 as the new president. And Sam was an alum himself, 00:20:16.200 --> 00:20:18.819 right? He was. He was in the very first YC cohort 00:20:18.819 --> 00:20:21.779 back in 2005 with his company Looped. So this 00:20:21.779 --> 00:20:24.059 was a real passing of the torch to the next generation. 00:20:24.339 --> 00:20:26.960 And under Altman, YC's ambition just seemed to 00:20:26.960 --> 00:20:29.740 explode. It really felt like they stopped wanting 00:20:29.740 --> 00:20:32.500 to be just the world's best accelerator and started 00:20:32.500 --> 00:20:34.859 wanting to be a global empire that touched everything. 00:20:35.099 --> 00:20:36.799 Empire is a good word for it. I mean, look at 00:20:36.799 --> 00:20:39.160 2016. The YC partners went on this massive world 00:20:39.160 --> 00:20:41.680 tour. They visited 11 different countries, Nigeria, 00:20:42.039 --> 00:20:44.690 Denmark, Russia, India, Israel, Mexico. They 00:20:44.690 --> 00:20:46.309 were trying to see if they could find brilliant 00:20:46.309 --> 00:20:48.009 founders anywhere and everywhere on the planet. 00:20:48.089 --> 00:20:51.230 And alongside this geographic expansion, they 00:20:51.230 --> 00:20:53.390 started something called Startup School. This 00:20:53.390 --> 00:20:56.430 was their attempt to kind of. Democratize the 00:20:56.430 --> 00:20:58.250 YC knowledge, right? Exactly. It was launched 00:20:58.250 --> 00:21:01.970 in 2017. Startup School was essentially a MOOC, 00:21:02.009 --> 00:21:05.210 a massive open online course. The idea was, look, 00:21:05.289 --> 00:21:07.910 YC rejects thousands and thousands of companies 00:21:07.910 --> 00:21:10.369 every year. Why not try to help them anyway? 00:21:10.609 --> 00:21:12.750 So they offered online courses, video lectures, 00:21:12.890 --> 00:21:15.529 and some level of coaching. Yep. And in the very 00:21:15.529 --> 00:21:18.670 first year, more than 1 ,500 startups graduated 00:21:18.670 --> 00:21:21.109 from the online program. It was a huge scale. 00:21:21.630 --> 00:21:23.609 But speaking of startup school, we have to talk 00:21:23.609 --> 00:21:25.950 about the glitch. This is one of my absolute 00:21:25.950 --> 00:21:28.750 favorite stories in YC history because it's just 00:21:28.750 --> 00:21:31.670 so chaotic and so human. It is the ultimate oops 00:21:31.670 --> 00:21:35.329 moment in tech. So picture this. It's 2018. A 00:21:35.329 --> 00:21:37.269 new batch of companies is applying to startup 00:21:37.269 --> 00:21:40.569 school. 15 ,000 startups have submitted applications. 00:21:41.109 --> 00:21:43.049 15 ,000. They're all sitting at their computers, 00:21:43.170 --> 00:21:45.069 you know, refreshing their inboxes, just praying 00:21:45.069 --> 00:21:47.529 for that acceptance email. You know. And then 00:21:47.529 --> 00:21:49.690 due to a software error, the system sends an 00:21:49.690 --> 00:21:52.170 acceptance email to all 15 ,000 of them. Oh, 00:21:52.309 --> 00:21:55.869 no. Oh, my God. So you have 15 ,000 founders 00:21:55.869 --> 00:21:58.049 all over the world who think they just won the 00:21:58.049 --> 00:22:00.849 lottery. They're popping champagne. They're calling 00:22:00.849 --> 00:22:02.809 their parents. Maybe some of them are quit their 00:22:02.809 --> 00:22:05.150 day jobs on the spot. Precisely. The celebration 00:22:05.150 --> 00:22:08.029 is on. And then a few hours later, they get a 00:22:08.029 --> 00:22:10.890 second email and it says, actually, so sorry. 00:22:11.029 --> 00:22:14.089 That was a computer bug. You were rejected. Ouch. 00:22:14.839 --> 00:22:17.460 The emotional whiplash there must have been absolutely 00:22:17.460 --> 00:22:20.599 brutal. It was. The outcry online was massive. 00:22:20.700 --> 00:22:22.660 People were furious. They were heartbroken. It 00:22:22.660 --> 00:22:25.420 was a complete PR disaster for YC. So how did 00:22:25.420 --> 00:22:27.059 they handle it? What do you do in that situation? 00:22:27.420 --> 00:22:29.900 In a very YC fashion. They didn't just issue 00:22:29.900 --> 00:22:31.880 a standard corporate apology. They looked at 00:22:31.880 --> 00:22:33.900 the situation and said, you know what? Let's 00:22:33.900 --> 00:22:37.220 just do it. They decided to accept all 15 ,000 00:22:37.220 --> 00:22:39.400 companies into the startup school curriculum. 00:22:39.539 --> 00:22:42.910 That is just... That's such a bold move. Instead 00:22:42.910 --> 00:22:44.750 of trying to fix the mistake, they just said, 00:22:44.849 --> 00:22:47.630 OK, our mistake is now the plan. We're going 00:22:47.630 --> 00:22:50.130 to own it and just scale the program to fit the 00:22:50.130 --> 00:22:52.150 mistake. It really highlights their mindset, 00:22:52.269 --> 00:22:55.190 doesn't it? Why bother filtering if you can just 00:22:55.190 --> 00:22:58.430 build software to handle the scale? If the software 00:22:58.430 --> 00:23:00.990 can teach 15 ,000 people, then let's just teach 00:23:00.990 --> 00:23:03.809 15 ,000 people. It's an engineer's solution to 00:23:03.809 --> 00:23:06.480 a human problem. But not every big experiment 00:23:06.480 --> 00:23:08.720 in this era worked out so well. There was the 00:23:08.720 --> 00:23:10.960 big push into China. Right. This was another 00:23:10.960 --> 00:23:13.920 major play in 2018. They hired Qi Lu, who was 00:23:13.920 --> 00:23:16.359 an absolute heavyweight. He was the former CEO 00:23:16.359 --> 00:23:19.259 of Bing at Microsoft and Baidu in China. A huge 00:23:19.259 --> 00:23:21.180 name. And he was brought in to run YC China. 00:23:21.559 --> 00:23:24.259 The goal was to establish a permanent on -the 00:23:24.259 --> 00:23:25.880 -ground foothold in the world's second largest 00:23:25.880 --> 00:23:28.220 economy. It seemed like the logical next step. 00:23:28.259 --> 00:23:30.160 If you want to be a truly global institution, 00:23:30.559 --> 00:23:33.359 you have to be in China. On paper, yes. Yeah. 00:23:33.690 --> 00:23:36.230 But in practice, it just didn't last. By November 00:23:36.230 --> 00:23:38.269 of 2018, I mean, we're talking less than a year 00:23:38.269 --> 00:23:40.490 later, YC announced they were pulling the plug. 00:23:40.549 --> 00:23:42.250 They were not going to pursue the China program. 00:23:42.509 --> 00:23:45.069 And Qilu departed and the entity that was supposed 00:23:45.069 --> 00:23:47.650 to be YC China morphed into a separate independent 00:23:47.650 --> 00:23:50.710 accelerator called Miracle Plus. It just goes 00:23:50.710 --> 00:23:53.130 to show that even for a powerhouse like Y Combinator, 00:23:53.309 --> 00:23:55.650 you can't just copy paste the Silicon Valley 00:23:55.650 --> 00:23:57.650 culture and playbook into a completely different 00:23:57.650 --> 00:24:00.230 market environment and expect it to work. Not 00:24:00.230 --> 00:24:02.869 at all. The regulatory environment, the funding. 00:24:02.990 --> 00:24:05.630 dynamics, the sheer speed of the Chinese market. 00:24:05.730 --> 00:24:08.150 It's just a different universe. And I think it 00:24:08.150 --> 00:24:10.569 signaled a retreat from the idea of physical 00:24:10.569 --> 00:24:13.410 international outposts. They decided it was better 00:24:13.410 --> 00:24:15.430 to bring the world's best founders to Silicon 00:24:15.430 --> 00:24:18.309 Valley rather than trying to build mini YCs in 00:24:18.309 --> 00:24:20.490 every country. Now, running parallel to all this 00:24:20.490 --> 00:24:23.250 expansion of the core accelerator was a very 00:24:23.250 --> 00:24:26.069 different and maybe even more ambitious project, 00:24:26.170 --> 00:24:28.930 something that wasn't about quick returns or 00:24:28.930 --> 00:24:32.140 7 percent equity. Let's talk about YC Research. 00:24:32.539 --> 00:24:35.380 This was launched in October of 2015, and it 00:24:35.380 --> 00:24:37.900 was personally seeded by Sam Altman. He donated 00:24:37.900 --> 00:24:40.059 $10 million of his own money to get it started. 00:24:40.180 --> 00:24:41.759 And the mission was fascinating. What was it? 00:24:42.079 --> 00:24:45.259 It was to fund long -term fundamental research. 00:24:45.420 --> 00:24:47.380 The researchers were paid as full -time employees. 00:24:47.599 --> 00:24:49.579 This was not about building a product to sell 00:24:49.579 --> 00:24:51.680 next year. This was about funding moonshots. 00:24:51.960 --> 00:24:53.940 And when you look at the projects they actually 00:24:53.940 --> 00:24:56.740 picked, moonshot is definitely the right word. 00:24:56.839 --> 00:24:59.460 For sure. I mean, the very first project explicitly 00:24:59.460 --> 00:25:01.579 mentioned in the source material is one that 00:25:01.579 --> 00:25:04.779 everybody knows today, OpenAI. Wait, wait. OpenAI. 00:25:05.500 --> 00:25:08.359 The company that created ChatGPT. The company 00:25:08.359 --> 00:25:11.079 that is currently defining the entire AI revolution 00:25:11.079 --> 00:25:14.900 that started as a YC research project. Yes. It 00:25:14.900 --> 00:25:17.640 was the very first project undertaken by YC Research. 00:25:17.900 --> 00:25:20.839 It started as a nonprofit research lab with the 00:25:20.839 --> 00:25:23.059 goal of ensuring that artificial general intelligence 00:25:23.059 --> 00:25:25.720 benefits all of humanity. That's incredible. 00:25:26.019 --> 00:25:27.440 Yeah. I don't think most people know that connection. 00:25:27.700 --> 00:25:29.660 It's a huge part of the story. But they didn't 00:25:29.660 --> 00:25:32.259 stop there. In January of 2016, they announced 00:25:32.259 --> 00:25:35.019 a massive study on universal basic income. Or 00:25:35.019 --> 00:25:38.660 UBI. So just giving people free money to see 00:25:38.660 --> 00:25:40.859 what happens to the economy and society. Exactly. 00:25:40.859 --> 00:25:44.180 A real world experiment. They also had a project 00:25:44.180 --> 00:25:46.119 dedicated to new cities, which was researching 00:25:46.119 --> 00:25:48.859 how to design and build better, more efficient, 00:25:48.980 --> 00:25:51.599 more humane urban environments from the ground 00:25:51.599 --> 00:25:54.559 up. And there was one called HNRC, right? The 00:25:54.559 --> 00:25:56.920 Human Advancement Research Community. The name 00:25:56.920 --> 00:25:59.500 alone is pretty ambitious. That one is particularly 00:25:59.500 --> 00:26:03.460 interesting and maybe a little abstract. The 00:26:03.460 --> 00:26:06.119 mission statement for H. Ayer Arc was, ensure 00:26:06.119 --> 00:26:10.160 human wisdom exceeds human power. Wow. Okay. 00:26:10.809 --> 00:26:12.769 What does that mean in practice? It was inspired 00:26:12.769 --> 00:26:15.529 by a conversation between Sam Altman and the 00:26:15.529 --> 00:26:18.650 legendary computer scientist Alan Kay. They brought 00:26:18.650 --> 00:26:21.150 in other luminaries like Brett Victor. It was 00:26:21.150 --> 00:26:23.970 really focused on creating new tools for new 00:26:23.970 --> 00:26:27.630 kinds of modeling, visualization, and programming 00:26:27.630 --> 00:26:30.730 languages to help humans understand complex systems. 00:26:31.660 --> 00:26:33.460 So what happened to HR? I don't hear about it 00:26:33.460 --> 00:26:36.140 today. It was shut down in 2017. And what about 00:26:36.140 --> 00:26:38.720 YC Research as a whole? Did it continue? Eventually, 00:26:38.720 --> 00:26:41.380 it spun out completely. In 2020, YC Research 00:26:41.380 --> 00:26:43.859 officially disaffiliated from Y Combinator and 00:26:43.859 --> 00:26:46.539 was renamed Open Research. So it seems like there's 00:26:46.539 --> 00:26:49.460 a realization that running a nonprofit blue sky 00:26:49.460 --> 00:26:52.460 research lab and running a for -profit hyper 00:26:52.460 --> 00:26:54.880 growth startup accelerator are two fundamentally 00:26:54.880 --> 00:26:56.559 different businesses. I think that's exactly 00:26:56.559 --> 00:26:58.460 right. They require completely different incentives, 00:26:58.740 --> 00:27:00.859 different timelines, and just different mindsets. 00:27:01.000 --> 00:27:03.279 Trying to house them both under one roof is probably 00:27:03.279 --> 00:27:05.440 too distracting for both sides. And this brings 00:27:05.440 --> 00:27:07.680 us to the most recent history, the post 2020 00:27:07.680 --> 00:27:11.079 era. And here we see a really distinct shift 00:27:11.079 --> 00:27:13.339 in strategy. It's a move away from expansion 00:27:13.339 --> 00:27:16.420 and toward focus. Yeah, a lot happens in a short 00:27:16.420 --> 00:27:18.839 period of time. First, obviously, COVID hits 00:27:18.839 --> 00:27:22.640 the world. The summer 2020 batch goes fully remote 00:27:22.640 --> 00:27:24.900 for the first time ever. Which must have completely 00:27:24.900 --> 00:27:27.640 changed the dynamic of the program. No more in 00:27:27.640 --> 00:27:29.420 -person dinners in Mountain View. Everything 00:27:29.420 --> 00:27:32.329 is on screens. It was a massive change. Then 00:27:32.329 --> 00:27:34.829 we have another leadership transition. Sam Altman 00:27:34.829 --> 00:27:37.009 had already left to focus full time on OpenAI. 00:27:37.309 --> 00:27:40.089 Jeff Ralston had a tenure as president. And then 00:27:40.089 --> 00:27:43.890 in January of 2023, Gary Tan takes over as president 00:27:43.890 --> 00:27:47.269 and CEO. And Gary Tan is another YC alum. He 00:27:47.269 --> 00:27:49.680 founded a company called Posterous that... Was 00:27:49.680 --> 00:27:52.180 NYC. He is. And he was also a highly successful 00:27:52.180 --> 00:27:55.160 VC in his own right with his firm initialized 00:27:55.160 --> 00:27:57.960 capital. And his leadership really marks this 00:27:57.960 --> 00:28:00.180 period of consolidation. The source material 00:28:00.180 --> 00:28:03.259 frames it as a return to roots. So what were 00:28:03.259 --> 00:28:05.140 the actual signs of this consolidation? What 00:28:05.140 --> 00:28:07.380 did he change? Well, one of the first and biggest 00:28:07.380 --> 00:28:10.180 moves was shutting down YC Continuity. What was 00:28:10.180 --> 00:28:12.359 Continuity, exactly? That was their growth stage 00:28:12.359 --> 00:28:14.640 fund. It was a separate pot of money designed 00:28:14.640 --> 00:28:16.819 to keep investing in YC companies as they got 00:28:16.819 --> 00:28:19.380 bigger in their Series B, Series C, and later 00:28:19.380 --> 00:28:21.220 rounds. So they were trying to compete with the 00:28:21.220 --> 00:28:24.039 big Sand Hill Road VCs. They were. But in the 00:28:24.039 --> 00:28:26.920 spring of 2023, they shut it down and laid off 00:28:26.920 --> 00:28:29.119 the staff associated with it. So that's a huge 00:28:29.119 --> 00:28:31.819 statement. They're basically saying we are not 00:28:31.819 --> 00:28:34.019 a growth stage investor anymore. We are an early 00:28:34.019 --> 00:28:37.140 stage company, period. Exactly. The thinking 00:28:37.140 --> 00:28:40.099 was that investing in later stages created weird 00:28:40.099 --> 00:28:42.579 signaling risks and distracted from their core 00:28:42.579 --> 00:28:44.980 mission, which is helping companies at the very, 00:28:45.039 --> 00:28:47.220 very beginning. What else? The other big change 00:28:47.220 --> 00:28:50.000 was the cohort sizing. In the summer of 2022, 00:28:50.380 --> 00:28:52.619 they deliberately reduced the intake size by 00:28:52.619 --> 00:28:56.390 40 percent. They went from 414 companies in the 00:28:56.390 --> 00:28:59.930 prior batch down to about 250. That is a massive 00:28:59.930 --> 00:29:02.450 cut. Why would they do that? The source cites 00:29:02.450 --> 00:29:04.529 the broader economic downturn as one reason, 00:29:04.650 --> 00:29:07.269 but it also strongly suggests a desire to focus 00:29:07.269 --> 00:29:10.509 on quality over quantity. If you have over 400 00:29:10.509 --> 00:29:13.450 companies in a batch, it's just physically impossible 00:29:13.450 --> 00:29:16.269 to give them all personal attention. 250 is still 00:29:16.269 --> 00:29:18.390 a huge number, but it's more manageable. More 00:29:18.390 --> 00:29:21.430 hands -on. A bit more, yes. And then there's 00:29:21.430 --> 00:29:24.519 the physical move. For years and years, YC was 00:29:24.519 --> 00:29:26.839 synonymous with Mountain View down in the South 00:29:26.839 --> 00:29:29.259 Bay. Right. The heart of suburban Silicon Valley. 00:29:29.400 --> 00:29:32.880 But in the spring of 2023, they moved their headquarters 00:29:32.880 --> 00:29:35.500 to San Francisco. Up by the waterfront, right? 00:29:35.579 --> 00:29:38.339 The Pier 1 area. Yes. And the source notes they 00:29:38.339 --> 00:29:40.000 had actually been considering this move since 00:29:40.000 --> 00:29:44.019 2019. It's a symbolic shift. The center of gravity 00:29:44.019 --> 00:29:46.259 for a lot of young founders had shifted from 00:29:46.259 --> 00:29:48.680 the suburban office parks of Silicon Valley to 00:29:48.680 --> 00:29:51.799 the city itself. And YC decided to follow the 00:29:51.799 --> 00:29:54.059 founders. And we also saw Michael Seibel, who 00:29:54.059 --> 00:29:56.599 was a really central figure for a long time, 00:29:56.740 --> 00:29:59.660 step down as managing director recently in March 00:29:59.660 --> 00:30:02.500 of 2024. Yes. So there was a lot of shuffling 00:30:02.500 --> 00:30:04.680 in the upper ranks and all that seems to be pointing 00:30:04.680 --> 00:30:07.339 in the same direction. A leaner, more focused 00:30:07.339 --> 00:30:10.140 Y Combinator. Let's broaden the lens out a bit 00:30:10.140 --> 00:30:12.599 to the entire ecosystem that YC has built over 00:30:12.599 --> 00:30:14.779 these years. We listed the giants earlier, Airbnb, 00:30:15.140 --> 00:30:17.720 Stripe and so on. But the alumni list is just 00:30:17.720 --> 00:30:19.859 so incredibly diverse. It's not just consumer 00:30:19.859 --> 00:30:22.339 apps for your phone. No, it's really not. You 00:30:22.339 --> 00:30:24.759 have this whole category of what you could call 00:30:24.759 --> 00:30:27.440 tech dev tools. These are companies like GitLab, 00:30:27.720 --> 00:30:30.779 Docker, specifically the part that was container 00:30:30.779 --> 00:30:34.019 Linux Heroku, PagerDuty. These aren't household 00:30:34.019 --> 00:30:36.619 names for your average person. No, but they are 00:30:36.619 --> 00:30:38.819 the essential tools that other developers and 00:30:38.819 --> 00:30:40.819 other companies use to build their startups. 00:30:41.200 --> 00:30:43.759 YC essentially funded the infrastructure of the 00:30:43.759 --> 00:30:45.839 modern web. And then you have the big international 00:30:45.839 --> 00:30:48.519 success stories. Rappi in Latin America is a 00:30:48.519 --> 00:30:51.710 huge one. It's basically... Everything app for 00:30:51.710 --> 00:30:55.690 LATAM food delivery, groceries, payments. And 00:30:55.690 --> 00:30:57.970 a more recent one is Deal, which helps companies 00:30:57.970 --> 00:31:00.329 hire and pay people remotely all over the world. 00:31:00.529 --> 00:31:03.089 These companies prove that the YC model can produce 00:31:03.089 --> 00:31:05.789 massive unicorns serving markets way outside 00:31:05.789 --> 00:31:08.210 the U .S. And we mentioned the nonprofits earlier, 00:31:08.309 --> 00:31:10.130 but there are some really fascinating ones in 00:31:10.130 --> 00:31:12.700 there. WOTC was the very first nonprofit they 00:31:12.700 --> 00:31:15.579 funded, and it focuses on health care crowdsourcing 00:31:15.579 --> 00:31:17.920 for people in developing countries. But they 00:31:17.920 --> 00:31:20.740 also backed an organization called 80 ,000 Hours, 00:31:20.880 --> 00:31:23.380 which is a really prominent, effective altruism 00:31:23.380 --> 00:31:27.839 career advice organization. Right. It shows a 00:31:27.839 --> 00:31:30.299 willingness to support organizations that build 00:31:30.299 --> 00:31:33.440 social capital, not just financial capital, which 00:31:33.440 --> 00:31:35.380 I think is an important part of their identity. 00:31:35.559 --> 00:31:37.859 But not every program they launched survived 00:31:37.859 --> 00:31:39.960 the test of time. We talked about the continuity 00:31:39.960 --> 00:31:41.900 fund closing, but there was also something called 00:31:41.900 --> 00:31:44.430 the YC Fellowship, right? The fellowship is a 00:31:44.430 --> 00:31:46.849 really interesting case study in their willingness 00:31:46.849 --> 00:31:50.170 to experiment. It was launched in 2015 and it 00:31:50.170 --> 00:31:52.309 was aimed at companies that were, believe it 00:31:52.309 --> 00:31:55.410 or not, too early even for the main YC program. 00:31:55.769 --> 00:31:59.380 Too early for YC. Is that just like an idea on 00:31:59.380 --> 00:32:02.319 a napkin? Pretty much. A team with an idea, but 00:32:02.319 --> 00:32:04.420 maybe not even a prototype yet. And initially, 00:32:04.599 --> 00:32:06.980 the fellowship was equity -free. It was just 00:32:06.980 --> 00:32:09.539 a grant and some advice. And then they changed 00:32:09.539 --> 00:32:11.500 the model. They did. They shifted to a model 00:32:11.500 --> 00:32:15.940 where they gave $20 ,000 for 1 .5 % equity. But 00:32:15.940 --> 00:32:18.960 by 2017, they discontinued the whole program. 00:32:19.240 --> 00:32:21.180 It seems like they just struggled to find the 00:32:21.180 --> 00:32:24.400 right economic model for that ultra, ultra early 00:32:24.400 --> 00:32:26.819 stage. Or perhaps they just realized that the 00:32:26.819 --> 00:32:28.980 main program was elastic enough to handle those 00:32:28.980 --> 00:32:31.460 companies if the idea and the team were good 00:32:31.460 --> 00:32:33.420 enough. You know, why fracture the brand with 00:32:33.420 --> 00:32:36.480 the YC Lite version? That makes sense. So as 00:32:36.480 --> 00:32:38.759 we look back at this entire 20 -year history, 00:32:38.900 --> 00:32:42.539 from a small summer program in Cambridge to this 00:32:42.539 --> 00:32:46.859 $600 billion value creation engine, what's the 00:32:46.859 --> 00:32:49.140 big takeaway for you? I think the story of Y 00:32:49.140 --> 00:32:52.019 Combinator is really a story of them applying 00:32:52.019 --> 00:32:54.660 the startup mindset to venture capital itself. 00:32:55.000 --> 00:32:57.740 They didn't just find a formula in 2005 and stick 00:32:57.740 --> 00:33:00.000 with it. They iterated constantly. They pivot 00:33:00.000 --> 00:33:01.960 just like the startups they fund. Exactly. They 00:33:01.960 --> 00:33:04.460 tried running programs in two locations, realized 00:33:04.460 --> 00:33:06.839 it was inefficient, and they consolidated. They 00:33:06.839 --> 00:33:08.819 tried expanding to China, realized it wasn't 00:33:08.819 --> 00:33:11.410 the right fit, and they pulled back. They tried 00:33:11.410 --> 00:33:13.109 doing late stage investing with a continuity 00:33:13.109 --> 00:33:16.529 fund, then decided to refocus on their core competency, 00:33:16.849 --> 00:33:19.329 which is the early stage. They're constantly 00:33:19.329 --> 00:33:22.059 searching for their own product market fit. even 00:33:22.059 --> 00:33:24.220 as the market around them changes. That's a perfect 00:33:24.220 --> 00:33:26.460 way to put it. And right now, under Gary Tan, 00:33:26.779 --> 00:33:29.880 that fit seems to be San Francisco -based, intensely 00:33:29.880 --> 00:33:32.220 focused on the early stage, and much leaner than 00:33:32.220 --> 00:33:35.180 the sprawling, almost imperial ambition of the 00:33:35.180 --> 00:33:38.019 Altman years. It's a fascinating evolution. And 00:33:38.019 --> 00:33:40.099 it really leaves us with a final, provocative 00:33:40.099 --> 00:33:43.460 thought to chew on. YC Research incubated OpenAI, 00:33:43.740 --> 00:33:46.019 which is arguably the most important technology 00:33:46.019 --> 00:33:49.099 company this decade. But YC spun that research 00:33:49.099 --> 00:33:51.720 arm off. They also shut down their growth fund. 00:33:51.859 --> 00:33:54.319 They are consciously and deliberately doubling 00:33:54.319 --> 00:33:57.680 down on the two founders in a garage phase of 00:33:57.680 --> 00:33:59.980 innovation. Which raises a really big question. 00:34:00.240 --> 00:34:03.460 Are we about to see a new golden age of raw early 00:34:03.460 --> 00:34:06.660 stage innovation? By clearing away all the distractions 00:34:06.660 --> 00:34:09.539 of late stage fund management and global franchising, 00:34:09.659 --> 00:34:12.480 is YC preparing for the next massive wave of 00:34:12.480 --> 00:34:14.340 disruption that looks entirely different from 00:34:14.340 --> 00:34:16.559 the mobile and sauce wave of the last decade? 00:34:16.760 --> 00:34:20.760 Or here's the flip side. Has the startup industrial 00:34:20.760 --> 00:34:24.559 complex peaked? Can you really just keep manufacturing 00:34:24.559 --> 00:34:28.219 unicorns on a schedule year after year? Or was 00:34:28.219 --> 00:34:31.809 the last decade an anomaly? fueled by 0 % interest 00:34:31.809 --> 00:34:34.630 rates and a unique set of technological conditions 00:34:34.630 --> 00:34:37.309 that won't be repeated. That is the $600 billion 00:34:37.309 --> 00:34:39.349 question, isn't it? Well, if you're out there 00:34:39.349 --> 00:34:40.989 listening and you think you have the answer, 00:34:41.090 --> 00:34:43.250 or at least a billion -dollar idea, you might 00:34:43.250 --> 00:34:44.989 want to go look up that standard deal we talked 00:34:44.989 --> 00:34:46.929 about. Half a million dollars is waiting. Just 00:34:46.929 --> 00:34:49.610 make sure you read and understand that MFN clause 00:34:49.610 --> 00:34:52.690 very carefully. Always read the fine print. Thanks 00:34:52.690 --> 00:34:55.190 for listening to this deep dive into Y Combinator. 00:34:55.230 --> 00:34:56.809 We'll catch you on the next one. Goodbye, everyone.