On human behavior

my 2023 letter of things I've learned about people in a crazy year of technology

Dec 31, 2023

It’s been a few years since I’ve written an end of year letter, but this year was an outlier one in so many ways that I’d be remiss not to document what I’ve learned about myself and the people around me.

A bit about me… I’ve been investing in seed stage AI companies for two years now at Bloomberg Beta. I started my job as an investor while I was finishing my master’s in computer science from Stanford… a few years after the release of the transformer and a few months before the LLM Cambrian explosion in the summer of 2022. I’m in my mid-20s. I live in San Francisco. When people ask me what I do in SF, I sheepishly say “I invest in AI companies” and brace myself. It’s not quite the same as saying “I work in crypto” in 2021, but you can imagine how someone in the bubble might feel a bit self-conscious. When I leave town and people ask me, I get quite a different reaction… “Oh you invest in things like ChatGPT?”, “Should we be scared?”, “What’s it like in SF right now?”

It’s undeniable that it’s a special time in SF. My friends who built startups in the city during the early days of the internet say that the energy today is the same as it was back then… something about a post-Covid era driving people starved of meaningful community together along with the type of intellectually curious people who become AI accelerationists that gives the city a centralized hopefulness. But with any gold rush of opportunity— the ability to make an outsized amount of money and to define a pivotal moment of technology— people’s true colors show quickly and quite obviously.

A few weeks ago, my friend Nick and I were talking about how we need to have “a Joan Didion of SF”… exacting and honest new-journalism descriptions of what it’s like to be in a place synonymous to Hollywood in the ‘70s with all the ambition, idealistic yearning, and greed. Here’s what I’ve observed and learned about people this year…

At the start, the incumbents can be dismissive… and then quick to claim precedence

One of the most common and naïve questions I’ve gotten from people this year is: “Do you think AI is a bubble?” The earliest AI research started in 1956 with deep learning defining the early 2000s. Machine learning researchers know that intelligent LLMs have been coming for decades but are reluctant to admit that the year cultural historians will likely call the next inflection point in the history of the internet after web, mobile, and social media is 2022… when ChatGPT, Midjourney, and Stable Diffusion were popularized. After all, technology is measured by its impact on society at scale and its ability to capture the imagination of the consumer.

When I was in undergrad, the hottest and “most prestigious” concentration in the computer science major was artificial intelligence… what that really means is “glorified data scientist” as we learned to be proficient in various Python libraries to train small models from scratch and figure out the right ways to trial and error different model parameters. Today, you don’t need a specialization in artificial intelligence to build an app that uses LLMs… you can unlock a lot of power from a simple API call, and anecdotally, that makes a lot of early machine learning pioneers squirm after they spent years cutting their teeth on linear algebra to build complicated models.

The same goes for builders and investors who were around during the early days of the internet. We live in a very different world today driven by social dynamics that didn’t exist during the dotcom era where having precedence doesn’t necessarily increase odds of success. As we’re gaining our footing in this new world of AI, it’s only wise to realize that power dynamics can shift overnight. Dirac famously claimed that “even second-rate physicists could make first-rate discoveries” in the 1920s, which I think is analogous to AI today.

Surrounding yourself with people allegiant to their cause will prove for a better social experience

People who know me very well think that I’m someone who makes a quick judgement about someone or something and then becomes very confident about my opinion. Depending on the person, they either find that endearing or annoying. I trust my gut on most of my first impressions of people, and throughout the year, I’ve come to admire people who are extremely disciplined and dedicated to the ideology of what they want to work on, and SF is the city with the highest concentration of people like this that I’ve ever been to.

People who move here typically have a reason to… it’s not for maximizing entertainment or finding themselves. The culture here doesn’t really foster competing over gimmicks that can only be won over with money like getting a restaurant reservation or impressing friends with clothes and other status symbols, so I’ve seen that it’s easier for people to stay allegiant to whatever they came here for in the first place.

I’m not saying that SF is devoid of people chasing status though… it’s just different. I think one of the most dangerous delusions I’ve seen is people associating proximity to power with real power. The more positive spin on “everyone in SF works in tech” is that most people building and investing are within two degrees of connection away from each other and that the city is not that socially hierarchical. People who namedrop and make a big deal about being at the same party as X tech celebrity are typically worth avoiding as you realize that it’s not statistically insignificant.

Always question new beliefs you’re fed

Two things about me: (1) I hate gossiping about other people’s personal lives, (2) I immediately zone out of superficial conversations… and I’ve come to judge cities by the types of conversations I overhear people having at coffee shops or at dinner dates. When you start paying attention to stuff like this, you realize that there are very few conversations that pass the Bechdel and reverse Bechdel test between groups of women and men outside of SF. That’s not to say that I don’t enjoy getting to know a person… I just think it takes a lot of deeper digging to understand their mental models, see their quirks, and understand how they got there from the way they were raised.

I’ve studied a lot of philosophy, and when I was in school, my favorite memories are staying up super late and chatting with my roommate about Sartre, Camus, Heidegger… I tend to become really close friends with people who’ve thought about the same types of questions as I have. When I first moved to SF, I found these types of conversations all around me naturally, and I told one of my closest friends here that “SF feels like an extended college campus… a place where you don’t have to give up your idealistic curiosities for reality just yet.”

But there’s such a thing as overdoing it. When you live in a competitive environment where everyone’s competing on similar planes of status like intellect, impact, and wealth, you see people trying their hardest to differentiate from the crowd. I’ve heard far too many people in the valley cite Peter Thiel for his discipline and philosophical knack and then completely buy into his set of libertarian beliefs. I agree and disagree with a lot of what he writes, but I think it’s dangerous for newcomers to the valley to wholeheartedly accept Thiel as a status symbol to signal they are intellectual without first reading Girard and Strauss and knowing what they’re buying into. It’s incredibly ironic because it’s mimetic.

Grifters tend to think small and prioritize fun

Crypto attracted all sorts of bad actors because there was the promise of becoming ridiculously rich almost overnight. That’s not the case in the world of AI (although there are the exceptions like Lensa and other consumer apps) because you still have to go through the motions of building a company like product thinking and go to market strategy. I think grifters in AI today are people who tend to build in niche and small markets. In a world abundant with longterm ideas today, it’s easy to see what motivates people who choose to work on short sighted visions.

✨ personal updates

I live in San Francisco but finally moved away from a downtown skyrise, which has done wonders for my mental health! I hosted five 12-person dinners this year… typically centered around a theme or a question with Chatham house rules. I’m planning on hosting a lot more next year, so find me on Twitter (@theamberyang) if you want to join.
I spend pretty much all my time thinking, writing, or talking about AI these days as an investor at Bloomberg Beta. And I was named to Business Insider’s Rising Stars of Venture Capital list this year!
I unwittingly coined the term “Cerebral Valley” this year, which kickstarted a lot of the AI founder’s community in SF… and talked to folks at the Washington Post, The Economist, WSJ, and the SF Standard about it.
In the second half of the year, I started running 3-4 miles daily and picked up pilates too.
For the first time in my life, I experienced deaths of loved ones… two of my grandparents and learned a lot about myself and my values.
I took my first international trips post-Covid this year!
The best books I read this year: Klara and the Sun by Kazuo Ishiguro and The Origin of Consciousness in the Breakdown of the Bicameral Mind by Julian Jaynes
The best essay I read this year: “The Straussian Moment” by Peter Thiel

amberyang