What America has meant to me

Every time I have given up on America, it has been a mistake. And yet, America has never, ever given up on me.

I write this on the occasion of America's 250th birthday, sitting alone to do some reflection as others gather at Crissy Field to watch the fireworks in the fog, or in various house parties on this this cold (42F, 6C) summer night. We have just come off the week of the third AI Engineer World's Fair (Jun 29-Jul 2), at 7000 registrants, the biggest AIE yet and our first in Moscone West.

Pretraining: My First America Run (2007-2010)

19 years ago, I moved to America for the first time, having being given access to Facebook back when it was still just Ivy-League only:

I wish I had documented so much more of my entrance into the US than I already do. Getting my social security number. Getting the very first iPhone (also a 2007 arrival). Asking for an "ATM" and learning that here in America we call it a "cash machine" (I later learned that this was probably an East Coast/Philly thing). Pretending to have read Ben Franklin's autobiography. Finding and losing role models. Taking my first government studies class with a congresswoman and then losing respect for government/nonprofit work in Lilongwe. That whole business in Cuba we don't want to talk about. Pledging a fraternity. Joining a gimmicky but fun acapella group. Discovering Wicked, and In the Heights to see an early pre-Hamilton Lin Manuel Miranda.

My new American classmates accepted me, though I was too strange and asian to ever truly be "in-group" - voting me for leadership positions and inviting me in to small house parties.

I started a real money investment fund for Huntsman (the biggest on campus), and constantly butted heads with our administrator, doing a bullshit study abroad in post-Olympics Beijing, all while casting longing eyes at the DARPA Challenge carsbeing built by the M&T program I really ought to have joined (if i had not been so dumb as to rule myself out of even trying to apply for Stanford). Self-teaching 3 years of Calculus in one summer so as to speed run a Math minor just because I thought it might be fun. Falling in love with statistical modeling with J. Michael Steele. Running a small seminar with G Richard Shell and having his personality test diagnose my interest in the literature of success and influence:

Getting an Asian Fail in Marketing 101 because it was far too overly tuned to American culture and I didnt fundamentally agree with/know how to apply the subjective frameworks taught (20 years later it is still hilarious that others turn to me for advice on my worst subject in business school).

Going super hard on Finance only to find that it is a house of cards built on false precision, as the world collapsed around us (look up the Li gaussian copula for the most egregious of these). becoming an RA for the 2010 huntsman class, getting kicked off campus for dating Z, and getting to know early stars like Jon "Every Vowel" Youshaei, and Sarah Guo, all of whom succeeded by betting hard on tech where I still bore unjustified conservatism from the 2001 tech crash (mind you, even then Google was already a great brand, but Twitter was mostly for posting status updates, and Netflix was still mail-by-DVD).

HUGE mistake: I graduated early and went back to serve in the Singapore central bank - probably still today a top 3 mistake I have ever made in my life, as it ended my relationship, destroyed my personal finances and set my life back 5-10 years.

Disillusionment: My Second America Run (2015-2020)

11 years ago I had just completed my CFA and Master's as a lowly sellside options trader, and G had called me up to pull me into the big league hedge fund life. This was my first move to San Francisco - where I did everything wrong - living in the Tenderloin, doing Finance in a Tech town, having no friends for 2 years (at my lowest point, paying a dating coach for a really terrible weekend class). I was jetsetting all over the world in business, but at the cost of zero life. I worked hard but rarely had original ideas, serving as "analyst" rather than investment principal, and missing the big AMZN run and getting squeezed on a correct but bumpy thesis shorting the Taiwan LCD complex, getting paid to be wrong but lucky shorting Toshiba-Westinghouse, and getting SKX consumer theses flat out wrong, and finally having our boss quit on us after a middling couple years, made me finally lose any love I had for working in Finance and start working on my Tech pivot. Sentieo/Alap welcomed me with open arms and I moved to New York to work on FinTech, but ultimately even that fun startup period was short lived with the poor India-US engineering culture and I eventually quit to learn to code proper at age 30, stretching out the limits of my visa before finally getting hired at Two Sigma for my first job as a software engineer.

Two Sigma was a high paid but bad job with an ineffective boss, and my frustration with this led to participating actively in NYC tech meetups and then my first real big break with React Rally, ultimately leading me to realize the value of Learning in Public which quickly led to Netlify, the first semi-successful startup experience I'd actually had. Netlify enjoyed a brief golden age running up to its first 1m developers, but rotted quickly due to gross mismanagement and poor strategy and I joined AWS just as Covid started to shut down the world, and I fled the land of the Moviepass, back to Singapore where we had seen the SARS coronavius 16 years prior and MERS 8 years ago.

Another mistake: Covid wasn't as bad as feared, and actually the people who stayed in NYC and in SF would have the best years of their personal and professional lives (much much more so in SF - I should not have given up on SF, but I wasn't a developer at the time so I didn't know). But of course the worst case scenario could have been death so I don't regret this as much.

Missing Generative AI

In "The Roaring 2020s", written in Nov 2020, I wrote:

• Generative AI: Generative tech will continue to improve at an aggressive rate. Our expectations have adjusted significantly from 2019 when things like these were "too dangerous to release". Generative tech will pass limited Turing tests, but mostly as a society we will be on high alert for fake news. I am optimistic about the potential for Generative AI to serve as a "bicycle for the mind" rather than a replacement or threat to humankind.
• I do not think we reach any sort of "Singularity" by 2030 because AI is still shockingly inefficient compared to humans.

Of course this was directionally correct, but far too measured. It is interesting to see how I classified this as a "smaller trend" vs, say, "Plant-Based Meat".

Re-invention: My Third America Run (2021-?)

I had given up on America again, but it kept calling me back.

My Netlify disillusionment and continuing dissatisfaction with the state of serverless even at AWS led me to write about Reconstructing the Monolith, which I never even had the confidence to make into a blogpost but triggered enough discussion that it eventually led to two(!) serious job offers for Temporal, which finally got me to move to Seattle as Covid ended. Seattle was horrible for a single male (I ended up moving out to Miami to work remotely), and Temporal also had management issues, but the thesis was correct enough and Temporal is by all accounts doing enormously well today after cleaning house.

I remember walking a Miami beachside road when Stable Diffusion launched, and I remember thinking that nothing else I had worked on even remotely mattered anymore. Because of impostor syndrome from not having background in AI (other than my George Tech OMSCS Online Masters ML courses), I established the "lspace" blog (borrowing from Discworld) and thanks to my prior 7? years of blogging it had some immediate readership and developed into Latent Space. I was on holiday in Singapore when ChatGPT was released, and I basically knew I had to drop everything to pursue AI, with no SF base, no network, and no lab plan. This was the most critical 3 months in the history of Latent Space - in December I published my first guest post with what became Codeium/Windsurf/Cognition, in Jan 2023 we recorded the first LS podcast (with Andrej picking this up in May), and in Feb 2023 we held the first Demo Day that led to the insight that became the Rise of the AI Engineer, leading to everything I am known for today.

I think the lesson here is that when I found a reason to come back, America welcomed me back with open arms, and supported/encouraged my amateurish exploration learning in public until I became less and less of an amateur.

I now think this is my last run in America. I will probably retire to Singapore where my beloved family and lifelong friends will always be, but I have at this point spent about half my life in America and certainly all my working and important professional years here, hopefully giving her as much or more as she has given me.

A parting thought

As someone who has a foot in both Asian and American worlds, increasingly I see America as the place that rewards first principles thinkers and builders; that asks "why not?" instead of telling you why you cannot; that imperfectly supports optimism, reinvention and perseverance in spite of cynicism. I don't feel fully integrated and probably will never feel fully American, but I feel accepted and welcomed, sometimes more so than in my home country, and THAT feeling is something as worth celebrating on July 4 2026 as it was on July 4 1776.