Life Update Summer 2021

Recently, I left my job at Tecton after working there for 2 years (from when I was 19 to 21.)

For those who don’t know, Tecton is building data tooling for ML. The product we were building addresses (in my opinion) the most important and painful problems with getting ML actually deployed and in front of users.

I joined Tecton August 2019, 9 months after the company was founded. I was the 5th employee; I joined as a general "intern" and then eventually became the PM of our open-source project, Feast.

First of all, working at Tecton was awesome. I have a ton of gratitude for the team, they took a chance on me and spent significant resources training me up (Mike, Kevin, Willem, Matt, etc.) I learned an absolute ton from the people I worked with about so many things. Tecton is gonna kill it, thanks for having me.

So, why did I quit? There were a couple of reasons, but the most important was that I realized I wanted to work on climate instead of ML/data tooling.

I am the type of person who works hard. I spend evenings and weekends wrapped in the problem I’m solving. I do this because I find it fun and because I enjoy it. Outside of my enjoyment, my work creates output, though. When I’m 60 and look back on my life’s output, I think I’d be upset if I spent it advancing data tooling, when I could have been advancing some domain that is important to humanity.

I have the personal opinion that climate is the most pressing problem for our generation to solve. I feel a pretty strong loss aversion towards the earth and I really do believe that my incremental effort can make a big difference.

I don't think we all need to quit our jobs and work on climate. There are a lot of problems worth working on. I think that if your work is creating any sort of value for the world at all, that's amazing! But I know that due to my personal feelings I'll have the best long-term satisfaction (and likely create the largest outcomes) by aligning myself with the problem that I care about.

So, what’s next for me in the climate space? Saying you're focusing on "climate" is somewhat of a meaningless statement, because climate isn't really even a problem, it’s more a symptom of a bunch of sub-problems.

Well, I don't have a specific focus yet. I was trying to figure out what to work on next while working at Tecton, but I really liked my job and I'm really bad at context switching. So I never made much progress. One of the first things I'm doing is figuring out my specific focus.

I do know I want to stay away from incremental solutions and small problems. My POV is that solving problems with a strong economic impact is the best way to achieve scalable impact, so I'll probably limit my scope to these problems as well. I’ll keep you posted on any specific problem I start working on.

I do not have 100% confidence that I’ll enjoy working on climate. For instance, the work might be radically different from what I’ve done before and I might hate the day-to-day. But my intuition is that this won’t happen, because there are a ton of domains I’m fascinated by, like energy and materials science. And I think there are roles/positions that are pretty close to what I’ve been doing.

I worked pretty hard while I was at Tecton, and I put off doing a bunch of exploratory-stuff that I wanted to do while I was young. In the interim before I jump into my next thing, I'll be catching up on:

I've heard from other people who have taken open-ended breaks that having limitations and goals can be really helpful. In general, I'm trying to be intentional about setting constraints and deadlines for the work I'm doing during this off-period. Pls DM me if you are interested in any of the ideas here or just want to chat! I am particularly interested in talking if you’re interested in climate as a founder, engineer, or scientist: I’m sure we can be friends and exchange interesting ideas. Twitter.

Addendum: was it hard to quit?

When I've told some people about me quitting, some people have been surprised and some people thought I was going crazy. I guessed this looked like "career suicide" to those people. Well, making this change was actually hard for me too, because of the following factors:

  1. I loved my work, both the day-to-day and the amount I was learning. I also really, really liked the team.
  2. I was intellectually interested in the problems we were solving.
  3. Things were going amazing at Tecton/Feast. Usage on Feast was growing by XX% per month including amazing customers like Robinhood, Salesforce & IBM. Shopify and Twitter also became contributors (mind blowing to me!)

There are also the following other factors that were actually less important to me, but led people to become surprised with my decision:

  1. The ML/data market is growing really fast. You can start a very valuable company very quickly within the space.
  2. I have domain knowledge that came as a result of unique experience. Since the market is so early and I was on the bleeding edge, I've become somewhat of an "expert" in various domains around ML/data.
  3. Climate and applied ML have a somewhat low cosine similarity. Maybe I can start a company at the intersection of the spaces, but I might have to build competency in a whole new domain. That seems risky!

So why did I want to change domains if all of these things are true? Well the most important thing, I think, was that all of these great things happened pretty much accidentally to me. I was only intentional about optimizing for two things:

  1. Working on things that were interesting to me.
  2. Working at places that'd I grow a lot at (hence, with great people.)

All of this other stuff just happened to precipitate out of those two optimizations. So I'm confident great things will continue to happen to me if I continue to follow that compass.

And sure, I'm losing some domain knowledge I had. But my interest is making sure that I impact climate in a major way over the next 20 years. Spending 2-3 years learning a new domain is actually totally fine if you zoom out and look at the whole journey.

I learned a bunch of generalizable skills in my role, like building product, understanding really complicated stuff, enabling technical people to do great work, and making shit run well. I'm not losing any of that.

I also somehow end up picking up domain knowledge really quickly. I think it's based on the places I find myself working (pre-PMF companies on the bleeding edge.) So I have confidence that I’ll pick up a new domain quickly.

On the "career suicide" point, my rationale is pretty simple actually. I don't care about my career at all. I don't want to be a VC or lecture at Stanford or have my name on some hospital. If I worked on five failed climate startups in a row, I’d be totally happy with that, as long as they were “good failures.” So I don't care about killing my career, it never mattered to me. ∎