In 2013, I started a company that made cloud-based software for training and deploying machine learning models (Ersatz Labs). I started it with profits from an IT consulting company I'd started in 2010. We gained some early traction and raised some angel money but ultimately failed to take it further and shut down in 2016.
A failed startup is a common enough story. Conventional wisdom tells potential founders not to worry about failure because they can "fail up" when things go wrong. In my experience, this was not the case.
When we finally shut down, I was extremely burned out. I had done everything I could to find a better outcome, but it had become clear to me that the product was not the right fit for the market and growth would be too slow to support a business. We could have pivoted, but organizationally things had started going south as soon as I raised angel money, brought on a new co-founder, and increased the size of the team. Stories for another day, but in the end it was better to just walk away.
When we shut down, I did receive some employment inquiries, mostly at companies I had ideological issues around. Even aside from ideological issues, I was in no frame of mind to start a new high-pressure position right out of the gate. I needed to take a break and get over my burnout.
In hindsight, I should have gone to a small, cheap island nation and swam with dolphins. Instead, I thought it would be fun to make a computer game. Soon enough, this became its own obsession and I simply replaced the stress of a startup with the stress of making a game against a ticking clock of diminishing savings.
After a few months, I really did need to make some money so I called around to some of the contacts I had met in my travels and managed to find some consulting work. After some months of that, and with enough runway to go for another year, I went back to the game and finished it. Unfortunately, it wasn't as finished as it should have been and the launch was a failure that left me in a pretty dejected state yet again. I worked on it a while more but it was always an uphill struggle and I really ran out of steam.
And so it was at this point that I decided I'd had enough of making and launching products. Starting Startups. Maybe I just wasn't very good at it. Fine.
So I wrote a resume for the first time in a while. I went on HN Who's Hiring, looked for people hiring data scientists or python web developers (pretty much everyone) and sent out resumes. My response rate was something like 60%, which seemed pretty good to me. The initial phone screens mostly went fine too. I approached it like any project and asked questions about the organization and what I would be working on, explained where I could help, and linked it back to previous projects where I had done something similar.
But then I went to the in-person interviews. Being a self taught software developer who has been programming since I was 12, I had no idea how to reverse a binary tree. Or whatever I was supposed to do with some of the probability related word problems I got. Or certain questions about complicated SQL queries (I use an ORM, rarely do I need to write a query by hand). I had been working as a consultant and/or starting companies for almost 10 years and found the b2b sales process is literally nothing like the hiring process for SV companies. In one we talked about project goals and solutions, in the other we talked about hypothetical word problems that I have no doubt are popular in CS programs.
And so I got a lot of passes. I asked for feedback but the feedback was always pretty meh. One company said "uses Windows so we don't think he'll be able to integrate into our OSX based workflow". Ha!
In any case, after about 20 of these in-person interviews going poorly, I really disconnected from the whole process and found myself deeply frustrated. Not only were these programming word problems ridiculous, but it seemed like everyone I interviewed with was a caricature of narcissism and delusions of grandeur. A surprising number of people talked about IPOs in order to justify lower cash pay but SPOILER ALERT very few of them ever IPOed and several of those companies don't exist anymore. A large portion could ask hypotheticals but couldn't show me a list of issues or actually explain what I'd be working on.
Then some acquaintances offered me a machine learning related job which sounded pretty great, even though the pay was lower. Unfortunately, it turned out to be a sweat shop, their whole personality changed on day one, and they fired me after a few months because they found they couldn't bully me into working longer hours for lower pay.
So at that point I really did say fuck it and just left the country to travel around and work as a digital nomad. I've been doing that for a couple of years now and I work an average of 2.5 days per week on machine learning or web projects and have rebuilt my savings and my whole life. This year with COVID has admittedly not been great, but just about every aspect of my life is a million times better than it was when I lived in San Francisco. The things I miss most are In-n-Out and Mexican food.
What's the lesson here? In my case, I suppose I learned what works for me is to scratch itches and follow threads, keep my cost of living low, and don't work too much. In the larger picture, I would comment that startups and tech companies claim to have a new kind of company culture, but they are (mostly) just the new factory with tech workers as the new factory workers. Rather than work at a startup, I much prefer being good at what I do and doing it on a temporary contract basis with clear boundaries and compensation. A job at a typical startup has no boundaries and feels like a cult. So I guess being unemployable works out fine for now, and maybe someday I'll get around to making something again.
I'm going to be blogging more about crazy things I saw working in Silicon Valley between 2009 and 2018 so if you find that kind of thing interesting, please follow me on twitter.