AI Techies!

Everything you didn't realize you wanted to know about the denizens of Cerebral Valley

Photo by Lan Dao

“I think about a world to come/ Where the books were found by the golden ones” — David Bowie

This is the third in a long-running series about subcultures. The previous posts are about tankies, weebs, and zines. Today’s post is about AI techies.

The word “techies” is pretty broad and amorphous. In the broader sense it could mean anyone who works in the tech industry, or even just anyone who studied a technical subject in school. In San Francisco, it’s often used in a much narrower sense, to refer to the wave of IT workers who moved into the city in increasing numbers in the 2010s. But in the last few years, I’ve increasingly heard the term “techies” applied to an even narrower set of people — a relatively small subculture of youngish people in San Francisco and the peninsula to the south, who often live in group houses and who are loosely clustered around the AI industry.

In January, Bloomberg Beta’s Amber Yang coined a term for the area in central SF where many of these young people are moving: “Cerebral Valley”.

Amber Yang @theamberyang

ppl in sf are now nicknaming Hayes Valley "Cerebral Valley" because all the AI communities and hacker houses are there AI seems to be the one thing in startup world that demands in-person rn so much so that if you're not in sf, you're at a disadvantage

Many pointed out that “Bayes Valley” would have been a more appropriate name, given that the AI industry is all about statistics…but perhaps that would have sounded less cerebral. Liz Lindqwister of the San Francisco Standard later wrote an article politely pointing out the fact that the group houses where SF’s new techie crowd congregates are actually located in nearby Alamo Square and the Lower Haight, not in Hayes Valley proper (SF micro-neighborhood nomenclature can be very confusing). In any case, here’s the area we are roughly talking about:

In this small area, you can find “hacker houses” where the new crop of techies lives, “third spaces” where they co-work and hang out, and cafes and restaurants and bars that they frequent. It’s sort of an urban economist’s dream — a local cluster where smart, creative people working in the same industry can rub elbows, exchange knowledge both explicit and tacit, form companies together, establish job-finding and hiring networks, and invest money in each other’s enterprises.

But it’s also a scene where people have a lot of fun. Last night I went to a cyberpunk-themed costume party “rager” at a group house, where if you got tired of drinking and dancing, you could go in a room called the “Church of GPT” (pictured above) and talk to a British-accented version of GPT that would pretend to be a sarcastic god, while projecting creepy AI-generated art on the walls.

I’ve been hanging out with people in the tech industry for about two decades now, and it’s interesting to watch how the culture has changed. So I thought I’d offer some observations on the most recent iteration.

Everything flows from the AI boom

The “AI techie” subculture I’m describing is not just people who work on AI. There are a fair number of people from other parts of the tech industry, especially young founders and junior VCs. And, as is usually the case with a tech boom, there are some amateur artists and philosophers hanging around (the latter being concentrated in the Effective Altruism area), soaking up the glory of the latest New Thing. Even the occasional econ blogger will show up to the parties.

But AI is really the core, and there’s good reason for that. With the end of the Second Tech Boom, much of the excitement — and money — has gone out of the consumer internet space, especially at the VC level. There was a weird little interregnum in 2020 and 2021 when suddenly all the money was in crypto, and the AI folks — who had thought it was their time — were sort of hanging around watching all these weirdo interlopers show up throwing around stacks of cash, and wondering if they had gone into the wrong thing.

Then crypto crashed, the weird people who had been hanging around town mostly vanished, and ChatGPT and AI art came out and wowed the whole world. The natural order of things righted itself, and AI continued its serene, implacable march to supremacy.

Science Is Strategic @scienceisstrat1

Artificial intelligence is eating venture capital AI startups attracted 21% of all VC investments in 2020, up from just 3% in 2012 Below is a 🧵 on how the deep learning revolution has drawn in billions from Silicon Valley & beyond Cc: @Noahpinion @erikbryn @amcafee @scmallaby

A VC friend recently told me that “100% of my deal flow is in generative AI”. That degree of dominance won’t last forever, and every new tech has its share of hype and irrational exuberance, but AI is highly unlikely to be a crypto-like bubble. For now, AI is it.

Money doesn’t just magically create subcultures, of course. But it can really help! The dreams of young people and their need for companionship are what create subcultures, and when it comes to tech, those dreams require money. And with SF rents being what they are, they require more money than usual. And if you want to be an amateur artist or philosopher and hang around cool young people and help spin their visions of the future, it helps a lot if they have money, or know people with money, to be your patron.

Because although money enables the AI techie subculture, fundamentally it’s about dreams. It’s about feeling like you’re part of something new and special and amazing that no one else is a part of — like you’re at the center of the world, where the future is being created, and that you and your chosen band of companions are the people who get to create it. That’s the dream that fueled the hippie counterculture in 1960s San Francisco, as Hunter S. Thompson wrote:

[N]o explanation, no mix of words or music or memories can touch that sense of knowing that you were there and alive in that corner of time and the world…There was a fantastic universal sense that whatever we were doing was right, that we were winning…We had all the momentum; we were riding the crest of a high and beautiful wave…

Of course, the AI techie subculture is far, far smaller than the hippie counterculture was — perhaps a few thousand people, even by the most expansive definition. And the sense of change is more technological and less cultural. But the romantic feeling of being in the center of the Big New Thing is recognizably similar.

There is a whole package of fairly standard beliefs that stem from — and help to intensify — this romantic feeling of momentum and importance. These include, but are not limited to:

• the belief that the AI boom is just getting started and will supercharge the economy

• the conviction that AI will replace most human endeavors, including most jobs

• worries about malign artificial general intelligence (AGI)

• a believe in the crucial importance of AI alignment (designing AI to be human-friendly)

Any of these ideas may be largely true or largely false, but the purpose they serve has little to do with their truth or falsity; they help motivate the people in and around the AI world, and to give a sense of purpose, uniqueness, and specialness.

Now it is worth noting that these beliefs can occasionally lead to strange social outcomes in the AI techie subculture. If AI takes all of the human jobs except for AI researchers and engineers (who will of course be fabulously rich), then perhaps you might think it’s very important to date and eventually marry other AI techies if you want your kids to have an economic future. That attitude, I suspect, is behind some of the awkward dating outcomes that people in the subculture occasionally complain about:

Ava @noampomsky

people who say there are no hot people in SF need to understand that conventional hotness is not that valued in this city; instead of hitting on models dudes here want to make sure you at least won a MIT hackathon before you ghost each other after a second date at state bird

But even though assortative mating based on the presumption of imminent human near-obsolescence is a bit silly, it’s also harmless. Most subcultures involve a lot of people dating other people in the scene. AI techies have a pattern of doing the things that normal young people do, but coming up with elaborate AI-related reasons to do them.

But what’s a little different about the AI techie subculture is that the confluence of money, youth, and freedom leads to an interesting balance between the desire for romantic bohemianism and the desire for upward economic mobility.

The bohemian/yuppie optimal control problem

“With a beautiful bombshell I throw myself into my work” — Radiohead

As the tech industry has changed, so has the culture of young people in tech. I’m far too young to have experienced the 1980s Silicon Valley culture that Annalee Saxenian talks about in Regional Advantage, but I am old enough to remember the techie culture of San Francisco in the 2000s. At that time, the SF tech scene wasn’t something you went into if you wanted to make a lot of money. In the years between the dot-com crash and the financial crisis, if you wanted an escalator to the upper class, you didn’t do a tech startup — you went to the East Coast and joined a hedge fund, a private equity firm, or an investment bank. And if you wanted to make big money in tech, you probably went down to the peninsula. The SF tech scene attracted the bohemian dreamers — people who wanted to sit around in a hoodie and jeans and make cool websites to connect the world and give the voiceless a voice, and then go to raves or East Bay warehouse parties or punk shows at night. If you read histories of this period, like Nick Bilton’s Hatching Twitter, you’ll get some of the flavor of this period.

In the Second Tech Boom — the years between the Facebook IPO and the crash of 2022 — the flavor of this bohemian tech culture persisted, even as big money poured into the internet space. The VC industry and incubator culture turned startups from passion projects into big business, while a bunch of go-hards from the Ivy League started coming out West, but SF tech culture still encouraged you to go to Burning Man and take some psychedelics and maybe call yourself “poly” instead of “single”. Most of the people in the industry were still very much on the political left. Group houses like Monument and hacker spaces like Noisebridge deliberately encouraged artists and creatives to mingle with coders and engineers. That culture is still around to some degree, of course, though rising rents push more of the bohemians out of town every year.

The new AI techie culture is a bit different. In many ways, it feels more conservative than the culture that built the internet. “More conservative” doesn’t mean “conservative” in an absolute sense, of course — these are still (mostly) people with college degrees, after all, and you’re not going to see any Trump gear or hear any rants about the dangers of “gender ideology”. People still prefer to call themselves “poly” instead of “single”, and no one is going to get on your case for doing a couple tabs of acid. But drugs and alcohol are frowned on more than they used to be, in a quiet and genteel way. There’s more rhetorical support for things like monogamy (gasp!) and childbearing (double gasp!). AI techies are a bit less likely to go to local Burner events and a bit more likely to go to Vibecamp, which I once heard described as “Burning Man for autistic people”. And there’s much less passion for social justice — at times, the concern over AGI almost feels like a way of tuning out the cultural battles of the 2010s. I hear a lot of disgust with SF’s homelessness situation, but not as much sympathy for the homeless themselves.

This might just be the result of the waning of wokeness after the turbulent ‘10s. It might also be because techies nowadays are more likely to be children of immigrants from more conservative regions of the globe — mostly from East and South Asia and East Europe, but also some from the Middle East and West Africa. It also might be partially due to the nature of the technology being created — the internet was inherently a utopian, humanistic project, whose fundamental premise was that getting more people to talk to each other was good. AI, on the other hand, is generally framed as a way for companies to replace human beings (though I heavily doubt this will be its actual effect).

But I think that a big difference is just money. As the startup world has become a big, institutionalized business, money has become more and more available to a younger and younger set of people (the Thiel Fellows being the ultimate example of this). Now, after the tech crash, VC money is crowding even more into the AI space in order to use its remaining “dry powder”. If you’re a young, smart person in the AI field, chances are someone is going to try to hurl a big chunk of money at your head. And even if you fail to grasp the top rung of the golden ladder, you’re probably still likely to be able to land a highly paid Big Tech job, despite all the layoffs.

But climbing that golden ladder requires you to do more than just sit around in your pajamas and write some code. You need to network. You need to meet VCs who might fund you (or, if you’re a VC, meet founders you might fund). You need to make friends with engineers you might hire, or founders who might hire you, or people who might give you ideas about how to solve technical problems. You need what Mark Granovetter calls “the strength of weak ties”.

There is a time-honored name for young urban professionals who dedicate their time to climbing the economic ladder: “yuppies”. There is nothing wrong with being a yuppie; economic success is a perfectly fine goal. The world has room for all kinds — slackers and go-hards, builders and artists.

But today’s young people, especially in San Francisco, are constantly reminded that there’s more to life than the golden ladder. You only get one youth, and if you spend all of it chasing dollars, you might wake up at age 45 and realize that you never had a star-crossed romance, or danced to your friend’s DJ set til morning, or dated three people at once, or smoked something that made you see aliens, or sat around with your band of trusted companions talking about radical ideas that might decide the fate of the world.

People used to working in AI are nothing if not brilliant and relentless optimizers. And so they have, very logically and skillfully, set about crafting a lifestyle and a social scene that lets them solve the multi-armed bandit problem of finding a balance between yuppie social climbing and bohemian youth. The hacker houses and the communal spaces and the parties and retreats are dual-use products — to some degree at least, they let you live the romance of youth while also building your professional network.

That’s not a perfect solution. Life is a constrained optimization problem, not an unconstrained one — without living two full lifetimes, you don’t get to be both entirely bohemian and entirely yuppie. No matter how laid-back you are, there will always be something a little distancing about knowing that the person you meet at a party might be your lifelong friend or the person who writes your company its next check.

And AI techies’ need to network can cut them off from the wider community around them. Whereas the techies of a previous age made constant efforts to stay close to artists and musicians, or to SF’s traditional “queerdo” culture, or even to the local working class, I don’t see much of this among today’s AI techie crowd. At times it can feel almost like the AI techies move through a different San Francisco that’s out of phase with the traditional one — a parallel city where non-tech people are ghostly wisps of smoke. I heard a story of how a hacker house tried to limit its membership to either AI researchers (not engineers!) or 2nd-time startup founders with a successful exit. (Naturally this push failed, given the small number of people who both fit this description and are young enough to want to live in a group house).

That’s pretty elitist. But it makes a certain sense, if you’re optimizing your bohemian urban youth in a way that gives you the greatest possible shot at exiting into a wealthy middle age. An hour spent with slackers or punks is an hour that could have been spent with someone who might be a valuable part of your network, or an hour having a peak experience on top of a mountain, or an hour spent building an artificial god. We have a limited number of hours on this sphere.

But at the same time, I feel like something is lost by this approach. Optimization within known constraints cuts us off from the unknown — the influx of ideas from outside our echo chamber, the life experience that expands our perspective in ways we never could have anticipated, the friendship that opens us to a more universal understanding of humanity. And the failure to mingle with other social classes makes it hard to get involved in local politics and community activity in productive ways.

(I don’t want to generalize unnecessarily, here. Some of the AI techies understand this perfectly well, and make efforts to get out of the bubble. But more should do so. It would be great to see a few normies at the hacker house parties once in a while.)

In any case, though, the AI techies are more fun than any other bunch of similarly ambitious yuppies I’ve come across. No, I don’t actually believe they’re building a god, nor do I think they’re going to be the only people with gainful employment in 20 years. But I like them anyway.

---