The best AI founders in the world are moving here

Why was San Francisco so definitively the center of the tech industry? Why did it all like agglomerate here? San Francisco is the place in the world where you can manufacture luck. Within a month of us moving in, they launched Twitter. I was like, "Wow, this place is incredible!" Everywhere here are people starting companies—really ambitious companies. And then COVID hit. In the last year, some things have happened that have turned things around; they feel the energy.

I actually think the biggest factor is that they have all the building blocks to make San Francisco into the best city in the world. Welcome back to another episode of the Light Cone. Today we're talking about San Francisco. It was dead, but now, like Lazarus, it is back. Not only that, we've got a new thing to talk about, which is Cerebral Valley. People from all around the world are coming to move to San Francisco, to just a few neighborhoods. We're in one of them, in the Dog Patch, to build the future— to build the future of AI.

Why is that? What's going on? I think it's worth just charting the course of San Francisco over the last few years. As a starting point, when we all got to San Francisco in 2006, it was coming out of its own Doom Loop—like Doom Loop 1.0 from the dot-com crash in the late '90s. San Francisco was full to the brim again, and then the dot bubble burst. When we got there, it kind of felt like a ghost town; there was a lot of vacancy, rents had crashed.

What brought San Francisco out of the Doom Loop was the Web 2.0 boom. All these YC companies like Stripe, Airbnb, and Dropbox moved to San Francisco, and they started hiring employees who moved into apartments. The tech economy just dragged San Francisco back out of the Doom Loop.

This is a very specific point: the Silicon Valley geography. I remember when we were moving here in 2007, there was a real negative connotation about choosing to base your company in San Francisco instead of, like, Palo Alto—yeah, Palo Alto, Mountain View, San Jose, even anything south. I think it linked to the dot-com era, where the perception, the belief was that during the dot era, all of the opportunists had come and moved into San Francisco because they wanted to live in a cool city and have cool things to do instead of working.

It was a real conventional wisdom that post that era, if you were serious about starting a company, you'd choose to be based in the peninsula. If you chose to be in San Francisco, you were like actively choosing to not be serious about your company. Investors really paid attention to this.

Yeah, and we put YC itself in the South Bay. Should we talk about where YC has been over the years? Yeah, Y Combinator got started in Mountain View on one of the most legendary streets at this point, called Pioneer Way. As Paul said in 2005, Mountain View was the place to be; Google had based its headquarters there. And people don't look at Google the same way anymore, but in 2005, Google was the undisputed king of innovation. All the smartest people wanted to be affiliated with Google, and so Mountain View became this mecca for smart people; it's the place where all the serious technologists were conglomerating.

Another major factor was just that in that era, people starting companies tended to be slightly older, right? Even like the Google founders were PhDs, like late 20s—maybe it's not old but not like 21 or 22 just out of college. I think even late 20s, early 30s, and there was a lot of it from Palo Alto because of Stanford, right? Yeah, and people wanted to be in the suburbs, basically. And the investors wanted to be in the suburbs, too; they were all older. So there was just this natural pool, and then the first sort of young founder, Mark Zuckerberg, chose to be based in Palo Alto as well, right?

So like the serious up-and-coming startups were in Palo Alto. Stripe was also initially in Palo Alto, but then I think like YC actually, not because Paul wanted it to, but it ended up contributing to the move to San Francisco dramatically. Like, we probably pulled it forward by years.

Yeah, you had a bunch of young people who wanted to not be in the suburbs, and they actually all ended up being in like the same building, right? Like you guys were living in Scraper, and I lived in the same building. This building in the North Beach area of San Francisco called Crystal Towers came to be known as the Y Scraper because there were so many Y Combinator companies in it. We actually wanted to be living in the building, but we were rejected.

Really? Yeah, we were rejected because we came to submit our application. They told us there were lots of spaces, and then they told us to go around the corner to the leasing office to get it. As I was walking out, I bumped into Justin Khan, who at the time was working at Justin.tv, and so he walked around everywhere with a camera on his head, which it turns out the building did not like.

So they saw me talking to Justin, and they must have called the leasing office to say, "Reject their application." So by the time I got around, they said, "There's no room left." What? Like one minute ago, you said there was plenty of room!

I was like, "Um, um..." But so we ended up living vicariously because we tried to come over to the building as much as we could for dinner and lunches, and just every time we would learn so much. Like you had the first of Dropbox in there, Scraper was in there, Weebly.

There's actually a very important point here, I think, which is for people who are builders, it's actually important to be around other people who are like that. So, anyone who has built something that you respect, that you think is really good and that you want to build things similar to, try to be near them. Actually, just being in a community with other people like that greatly increases your chance of success.

I mean, I think that was true for us. I mean, that's why we, for Posterous, we applied to Y Combinator. We looked at Steve at Reddit or we looked at Adrew from Dropbox or James from Heroku, and we said those were the kinds of things that we wanted to build. I think, you know, today that sort of continues. Like, you want to be near people like that.

From my personal experience, I grew up; I was born and grew up in England. YC reading PM essays, and hearing about YC and getting accepted to YC in 2007 is the reason that I moved. But I initially only planned on moving for three months. Like, that's how long we could stay on a tourist waiver. But as soon as we got here, it was just like a complete world of difference.

I went from being around all of my peers working in banking consultancy, just thinking I was unemployed or too lazy to get a real job, to being in San Francisco and the Bay Area, especially amongst the YC founders. It was like, "Hey! Everyone's doing a startup! Everyone's in the same boat; we can all support each other!" It was just like in the water.

It was my first time experiencing being in an environment that felt like nourishing versus draining for starting companies, and you could feel that really intensely. Just being in San Francisco.

And then you found yourself in one of those legendary rooms very early on with your first startup. Yeah, it was, again, just like a total serendipity-luck factor. Evan Williams, the founder of Twitter, had happened to be speaking at a conference at my college in like 2006, and we spoke to him, and he was like, "Hey, if you guys move out to San Francisco, we have some free office space and you should just come hang out there," which is what we did.

That startup at the time was called Odio, which is a podcasting company that was not actually doing very well. You could feel it in the office. I mean, it's why they had space and free desks, right? They had more space than they needed. It was very much, people were coming like 9 to 5, and so it just wasn't like a super dynamic environment.

But within a month of us moving in there, they launched Twitter, and Twitter just like took off. Suddenly, the whole energy in the place changed. There were always interesting people coming in and out of the office, and you got really energized by being around something that was clearly working so well. It made you want to like put in more effort and feed off the energy.

What's funny is our startup was in the same office as Twitter, and at the same time, I was running a different startup. Twitter moved offices after they outgrew that space, and they moved into my office building. My startup was actually across the hall from Twitter's second office.

This might seem like a bizarre coincidence; it's actually not that unusual because all of Tech was concentrated into this very small area in like downtown and Soma. People who haven't been here don’t realize just how dense it was, where you could go from any startup to any startup in San Francisco in like 15 minutes.

Yeah, I mean, I remember a very specific example of that. Like, we left the office to get lunch, and we stepped outside and bumped into Ron Conway, the legendary angel investor. He just starts asking us some questions. We met him at a YC event, and we just told him, "We've got a bunch of problems, we really want to get blah, blah, blah as a customer." By the time we got back to our office, he'd shot off these intros to all of the companies for us to help us with making sales.

Like wow, this place is incredible! I mean, the crazy thing I think San Francisco is the place in the world where you can manufacture luck, and for startups to really get big, lots of things need to happen to be right. There's something special about being surrounded by all the people that kind of know how, and they're all here, and they're willing to help because starting companies is actually a very anomalous thing in other parts of the world.

It's a strange job. To your point, Harsh, starting a company could be looked at like, "Oh, you're just a bum, you're not really working," at the beginning. What is going on? But here, people really celebrate it because people understand how hard it is.

How is it for you, DMC? Because you grew up in Chile. How was the move to Silicon Valley and being around this?

So I relate a lot to the story of being a misfit. In Chile, when I grew up, I was always a weird kid. I stayed indoors because also it was very difficult being the only Asian kid; there were no other Asian kids around. I really loved reading books, and when I got a computer, I really liked building a lot of these RPG games back then, emulators. I got a lot into that, and the Internet was like my bubble, but I didn't know anyone like that.

What was cool is I always admired all the builders, and my dream was that at some point I would move to be in Silicon Valley and be there. What was special that I didn't know, and there was this pull, is that when I got and moved to Silicon Valley, I started working at Intel. It was my first job. It's the first time I felt at home; there were a lot of people like me who really liked to nerd out and really to go deep on subjects and building, and it wasn't judged.

Because I was judged back then; I had to sort of hide and not talk about my interests when I go in that super deep, and that was special, where it was celebrated and you can build. There are also the experts in different areas, and everyone is just as deep as you with building and it's just as exciting and an optimist instead of being judged on why you like this weird thing.

One of the funniest things that I remember from reading Paul Graham's essays, and I think it was one of the reasons why a bunch of us found his essays, was he would always describe what it was like in high school to be a nerd. I think that might be a common experience for us, where you sort of grow up in the default society.

You realize that the default society is a little bit foreign; it has different values than you. You know, it sort of values appearance more than substance—the right way to do it. That kind of keeps going even after college. Let's say you, for young founders, your friends who may not be founders end up getting whatever the standard job—being an investment banker or working for consulting, or whatever, or the big tech job.

They end up having this standard life where it's like, "Okay, they get the promotion, they get the nice vacations, they get the time off," and a lot of their meaning is more outside of work. When you're friends with, nothing against that by the way, but when your friends are like that, and you're doing your startup, and you're working 24/7 and working very hard and not really having a lot of money to take the fancy vacations, you kind of get a little bit judged, implicitly—not directly.

And that can be a bit demotivating. I think what's special about being here is you surround yourself with people who are in that same mode, in that same vibe—those who are working hard and chasing their dreams to build something. It's not looked down upon that you're not taking a fancy vacation or going to fancy dinners; it's actually, on the other hand, celebrated that you're working hard and really taking the lead.

Got such a deep point, D. I think that is actually the true reason why startups and startup founders conglomerate in Silicon Valley. Some people think it's transactional reasons, like, "Oh, this is where the investors are," so you have to be close to the investors, or "this is where the employees are," so you have to be in the place where you can hire employees.

Those, I actually think, are probably more minor factors. I actually think the biggest factor is that people become the average of the people who they surround themselves with. When you're in a city that doesn't value this kind of thing, it's really hard to stay motivated. It's really hard to be as ambitious.

When you move to Silicon Valley and you're surrounded by people, like everywhere here, people starting companies—really ambitious companies—people who are really trying to change the world in a very serious, dedicated way, it just wears off on you and it makes you motivated to want to do the same thing.

It's also like there's a particular sense of long-term ambition here where startups just take so long to actually work out and make any difference, and everyone who starts out here is on that journey. I think generally people are realistic and want to make a difference, but you're also just very incentivized to think long term and not pay attention to short-term status games.

You care; you kind of care about like your relationships over a long period of time. If you contrast that to, say, a finance Hub like New York, there's a huge amount of ambition there, but it tends to be more like, "How much money are you making right now?"

One thing we all see is people come to San Francisco or YC with maybe their ambition level, like they think is at like, you know, a two out of five scale, but they grow from all the other people. Especially if the company starts doing well, their ambition changes, right? I think that's a very unique thing here where if you're working on technology, there's almost no bound to how far your ideas can go; it's just how long do you want to keep pushing them and working on them for?

If you're around other people who are also interested enough in what they're doing to keep doing it, versus trying to hit some end state of success, I think that's how you end up with these companies where you have Larry Ellison working on Oracle for like decades.

Like anything that happens anywhere else, this is actually a place that generates social movement. We believe X; I believe that AI and large language models might recreate every type of enterprise software, and that's happening right before our eyes right now.

Or we believe that electrification and decarbonization will remake how we use energy in the world; like that's sort of happening before our eyes.

Then being around people who have those ideas—like not all of them are going to be right. Even having a culture where being wrong is okay—that's very rare, and I think that's, you know, having a truly inclusive place where it’s, "What do you believe, and what do you believe that nobody else believes yet?"

That's the very unique reason why I think San Francisco is the most special place in the world. Right up until COVID, it was clear that San Francisco was the center of the startup and technology world.

You’re completely right; every neighborhood was becoming busier. Rents and property prices kept getting more and more expensive. It was very clear the center of everything, and then COVID hit.

Yeah, yeah. Like, I think that was probably the thing—the backlash. Office vacancy was like 0.5%. New apartment buildings, especially in the area where the Salesforce Tower was built, I always felt that that felt like another turning point, like a huge tower.

Like that whole area, we had a startup office based right around the corner from there, and just the new apartment buildings going up on every block, like it really felt like the center of everything.

I remember in 2019, it seemed like one of the biggest problems that YC had was that there was not enough place to put all the people who wanted to move to San Francisco. There was not enough housing; there was not enough office space. We were just bursting at the seams.

Yep. And then COVID hit. What happened in COVID? How did things kind of turn around?

So, the downturn in a matter of a few months. Everything changed in a matter of a few months. I think right up until COVID, everything we said is very true. It was clearly the center of the tech world; everything—people were moving. But there were clearly problems with the city like beneath the surface: homelessness was increasing, crime was increasing. There were all these tensions that people looked past because like you had to be here.

I felt like COVID hit, remote work became necessary, and it was a chance for a lot of people to finally move out and experience living in another city where they didn't have as many of these problems as San Francisco did.

I would meet people who would talk about San Francisco as though—Jared puts it this way—like Gotham City. If you were on Twitter or if you were just bumping into casual people, they were like, "Oh, like it was really like, wow, you're from San Francisco. I'm sorry you go out at night." And this was basically all of, I think, 2021 and most of 2022.

Then in the last year, some things have happened that have turned things around and are starting what Gary's calling this boom loop. What happened in the past year?

Well, the clear thing is AI, right? Like the ChatGPT launching was like a complete reset and brought everyone back. I think like the broader point—in order to work on AI pre-ChatGPT 3 launch, you have to be sort of like a counterculture misfit. It was not at all a mainstream thing to be working on.

It was kind of—it was not cool to work on AI. If anything, actually, you know, it was a little bit of a dead end for some people at that point. It was a bit of one of the AI winters back then, and people were keep going because they knew something that everyone else didn't.

This is that part of being counter—I think it's a really important point because pre-that, like, and during the COVID remote work period, there were multiple hubs that had the claim to be like the next Silicon Valley.

I don’t think anyone was claiming that any one of them would be as big as Silicon Valley, but there was just like a general view that, "Hey, the future of innovation is just going to be distributed and dispersed." Some of it will be in New York, some of it will be in LA, some of it will be in London—all around the world.

But the thing that actually drove the wave happened in San Francisco. It doesn't seem coincidental. If anything, you can argue the fact that the city got so bad, and it’s still—like, it’s come back and it’s like now the center again—is just evidence of how strong the network effects are here.

Yeah, and it's not just OpenAI. Basically, all the AI ended up in San Francisco. OpenAI, Anthropic, Scale AI—basically all the big AI companies, all the big tech AI research labs—it's all in the Bay Area.

And now we're seeing like post-ChatGPT, like we're seeing San Francisco really come back. How is it different? What are we seeing that's different in sort of this iteration of San Francisco versus like the peak pre-COVID?

Well, commercial real estate is down, but I guess the wild thing is we still have some real big problems downtown and in large parts of Soma. This is sort of where they allowed developers to build housing. By default, I think a lot of people watching this who are not from San Francisco, a lot of them might end up in Soma and then realize, "Oh, this isn't a place where I want to live," and then they'd leave.

But, you know, I think the thing I want people to sort of understand is like Cerebral Valley is not Soma. Cerebral Valley is actually a play on Hayes Valley, which is in the center of the city. There’s lots of incredible food and shops, and that's where a lot of the best AI companies are choosing to be.

Right now we’re sitting in the Y Combinator offices. We signed more than 100,000 square feet here. The amazing thing now is we can have three different events going on all at the same time over here. This is a very different part of town that has a neighborhood vibe that's very safe.

So, you know, explore the neighborhoods. Figure out; don’t just do the default thing which is live in FIDI or Soma. There are so many incredible neighborhoods that might cost a little bit more, but you will get the real San Francisco experience.

As someone that moved back, I think the single thing that struck me most that's different is how much the neighborhoods have changed. Like, it really matters which neighborhood you're in now. I feel like pre-COVID, you sort of knew, like, the Tenderloin and the Civic Center were like the bad areas of San Francisco, and outside of that, every neighborhood maybe had a slightly different flavor, but it didn't actually matter too much if you were based in SOMA versus FIDI or even the Mission.

But now it's like it's very important that you pick the right neighborhood. Right? I feel like YC, now we're in like the Dog Patch, which we sort of locked into. I would say like what is probably the best neighborhood in San Francisco.

Yeah, how we talk about moving YC to San Francisco, which is something we talked about for years, but COVID happened and other things happened, and now we’re finally in a position to do that. You two actually led the search for YC's new headquarters. Do you want to talk about the places that you looked at and how we ended up here?

I mean, there were lots of neighborhoods, lots of different places. But if anything, we did the startup style, which is Dog Patch is really safe. It’s being redeveloped, and this was actually before OpenAI moved in over here.

There was a space right next to where we are right now that was the powerhouse for Pier 70, so it literally housed the engines that generated the power for where we are, the shipyards that built a third of the US Navy during World War I and II. So it’s sort of like this incredible symbolic significance of so much power being generated in that house.

So, that's actually our main event area, and then we just took over the office next door. Now we have a really sweet YouTube studio right here, and across the street, there are also two other YC companies, Astranis and Gusto.

Right, the Dog Patch is great. That's where our center of gravity is. But what's some of the pushback you all get from founders about moving to San Francisco? What are things they're worried about, and what's the advice you give them? Especially international founders.

There’s a lot of the consumption of news from Twitter, and there are a lot of things about being not safe and being scared of coming here, but when they actually move to the right neighborhood, they're pleasantly surprised, which is nice.

But this is not to say that the city doesn't have problems. There are definitely neighborhoods that still should be avoided, which we actively advise them where to be actually. Which neighborhoods? Where should people be based?

I mean, personally, I love, you know, Mission Dolores, Noe Valley. These are great areas; they’re a little bit more expensive. I think, you know, Glen Park is actually an incredible neighborhood that's sort of hidden. Bernal Heights is right over the hill from where we’re at in the Dog Patch.

And Mission Bay is brand new, and it's very close to YC and OpenAI. So, I actually think that during the batch, we actually tell people, "Can you be within a mile of Y Combinator in the Dog Patch?" The reason why is you know you're going to be working 80, 90 hours a week, like writing code or talking to users.

On those Friday nights, you're going to come out to actually this office, and we have a launch event where people demo their software and grab a beer. Afterward, they go out to the bar or grab pizza.

Then, you know, Friday and Saturday nights, that's when you can let your hair down a little bit and actually see the other people in your batch, and then those are the relationships that you sort of take with you for frankly the rest of your life.

I mean, one of the more surprising things that I’m still surprised by is, you know, when I got married after doing YC, two whole tables were my Stanford friends, but two whole tables were actually my Y Combinator friends. Paul Graham even came, and he even wore pants!

There’s actually a founder map that lays out where all the founders of the current batch are staying, and the center of gravity is around our office here in Dog Patch. That's not random reason; that's by design.

Maybe we can talk about why that is. I mean, I remember in my YC batch, we actually did the manual version of this, making a collaborative Google map in summer of 2008 while it was still in Boston.

The funniest thing was that’s how I got to know my co-founder, Brett Gibson. Him and his co-founder were working on a different startup called Slinkset, and we actually would just grab beer, pizza, and beers every weekend. When they couldn't raise money, we ended up sort of merging our companies.

Brett Gibson, who ended up being my co-founder for Posterous, ended up being my co-founder for Post Haven and also worked with us at YC, the software team. Now he runs Initialized, and it all started by just him happening to live next to me in Boston when I was going through YC.

Where you live, who you hang out with—I mean, it's pretty wild that these things end up mattering. San Francisco is the best place in the world to build that network that feels— I mean, you’re not building a network; you’re actually just making friends for life. That’s like the real version of it.

Then you mentioned Cerebral Valley. It did actually start as Hayes Valley. You could play on that, but I actually think Cerebral Valley is this neighborhood. It actually is this—it's kind of weird that Cerebral Valley became known as Hayes Valley.

I think the real argument for that is just that there is this one hacker house in Hayes Valley, but it was literally just a house. Again, what we’re talking about during the period that that hacker house started, San Francisco was pretty dead.

Yeah, it was so dead that like the hacker house was probably like the most happening place. You could become the center of San Francisco by literally having the one house, like 10 hackers in it working on like AI stuff.

It’s not true today. OpenAI has this huge lease right around the corner from here in Mission Bay, which is next to the Dog Patch. Right here, the startups are going to start being based here. This seems like it's the real, real Cerebral Valley right here.

This is the real Cerebral Valley, and I think a lot of the companies that we're working with right now, they're going to find product market fit. They’re going to get their first million, then 10 million, then 100 million, then billion dollars a year in net revenue.

They’re going to fill the office towers in FIDI and SOMA. They're going to build these businesses that create thousands of jobs, and there’s not going to be, you know, one Salesforce Tower. There might be dozens, if not hundreds in the next 10 years.

So, we’re really talking about a real boom loop here. This is already happening. I was talking to companies from the last couple of batches, and there’s this office space in SOMA where there are a bunch of YC companies that just raised their seed round, and they’re surrounded by each other, and they feel the energy.

There's this thing about being in the right vibe. They need a cool name like Y Scraper!

Then I think we're also seeing, like, another fact of being here is like the AI specifically. If you're working on developer tools, like in a big area right now, like you want to be here.

All the best engineers who are going to be using your cutting-edge developer tool or infrastructure all based out here. If you're building AI software—you need someone, especially enterprise software—you need someone who’s going to be your champion and advocate and is going to take a bit of a bet on like a startup that’s using some, you know, cutting-edge AI model that’s not fully proven yet.

And they're all like those people tend to be here. It’s always clear that you can build companies outside of San Francisco and Silicon Valley. There are always going to be exceptions, and maybe there’ll actually be more of them over time, but I just think all else being equal, if you're starting out, like, you want to play, just maximize the odds to be in your favor.

Like you should be here despite the fact that it will always be exceptions. Why don’t you want to get more lucky here in San Francisco?

Yeah, maximize your luck. So the great thing is, I mean, we want San Francisco to be hyper-inclusive, and what that means is you don't have to pay that much for rent. You can feel safe walking down the street. There are great small businesses that are, you know, thriving, and then frankly, this is the place where you should start your startup and thrive.

Here as well, 100 years from now, what does San Francisco look like in tech?

I mean, I'm a big fan of Gene Roddenberry's ideal for Starfleet command. I mean, this is the ultimate techno-optimist dream, which is we unlock so much abundance in the world that money doesn't exist anymore, and people can search the stars to try to find and create their own meaning.

I think that there's no mistake like Starfleet command from Star Trek: The Next Generation and Star Trek, you know, the universe broadly, is right there in the Presidio in San Francisco. I think that if you take the agglomeration effects of the smartest people in the world, the best builders in the world, and you put them in one place, and they create this scene, like this, you know, set of people who all run in the same direction, create these companies that matter.

Then people stay here, and you know, they make the schools awesome, they build housing; they actually, you know, invest into arts. We actually make the city the kind of the best city in the world.

Like, we have all the building blocks to make San Francisco into the best city in the world where you know, give us your misfits. Give us your nerds. Give us your autists. Give us the people who, you know, just wouldn't fit anywhere else.

Like they have this capability with their hands and with their brains to create something that has never existed before, and we put them all in one city, and they go and create software and hardware and technology and biotech and climate tech that will touch billions of people.

All of that wealth comes back into this one city, and we make the city more and more awesome. That's what San Francisco is; that is what San Francisco could be. We will manifest this.

The boom loop is happening, guys! So maybe that's a great place to end. That's it for this week of the Light Cone. We'll see you guys next time.