On Startups

The last two years have passed by in a quick blur. Moving to a 70-person company from a 70 thousand person company, I knew I had plenty of things to learn, but I also had a lot to unlearn. Here goes:

Responsibility

If you're interested in doing something, raise your hand and you can do it. In fact, sometimes, you can just do it. If you don't, you're left wondering what could've been.

This might seem like common sense, but larger companies operate on hierarchies and in fiefdoms. At a startup, the distance from idea to reality is shorter and has less friction. There are less toes to step on, less projects fighting for headcount/funding, and often no established procedures for things.

If you observe something going wrong, mention it. There are typically no established procedures, so often you're the first person to see something going wrong. And the only one who will... until it gets too big to ignore.

You can simply choose to do things

Expertise

If you volunteer to try a new piece of technology, you become the go-to person for that technology.

I volunteered for the team tasked with building a Content Management System on a short deadline. Why? I was frustrated with how subsurface content was managed on my team at Shell and I had some ideas that I never got to implement. It was also a chance to work on business software, as opposed to consumer applications. People eventually got pulled off the project and I stuck around. By the end, I had spent the most time on it and knew the most about it. I became the go-to person for any content management project in the company.

I got to spend time refactoring one of our most popular products because I spent so much time on it and would point out the problems and talk about how it could be better. In hindsight, I didn't really know what I was doing and maybe should've had someone senior guide me through it. But I had to figure it out as I went because my reputation was on the line.

I currently work on a research team exploring new game concepts using Generative AI. I have no AI credentials. But I was interested, experimented with, talked about, and was willing to do AI-related things. My initial perception was that the company would hire an AI expert at some point. But... Where would they find one? If they found one, why would they come here?

Share what little you know as soon as you know it. If you're wrong someone will correct you. Otherwise, keep learning till you know you're not wrong.

Vibes Matter

Startups are an avenue to do focused work with outsized impact. How so? Everyone has some level of ownership in the company and the teams are very small.

Misery loves company and bad feelings can poison the atmosphere faster than you think. At a large company people often feel replaceable and, sometimes, a whole team leaving can be remedied. A few departures at a startup could mean the end.

Work with competent people that you like. Treat one another kindly and ethically.

Clock Speed Matters

I've worked on projects where check-ins happened quarterly, monthly, bi-weekly, and weekly. Less is more. Why?

• You can stretch a task to fill the amount of time you have to complete it
• Burnout is more a reflection of lack of accomplishment than overwork
• Perfect is the enemy of done
• Small wins matter

There is nothing worse than spending an elaborate amount of time planning or working on something only for it to get canceled because priorities changed while you were heads down.

Strive for the most amount of completeness in the least amount of time. Then seek feedback on your execution.

Touch Grass Often

You build the best intuitions around products you're building when you have the most contact with reality. The best product you can build is one you use often. The next best is when you get a chance to watch people use it. If you can't do either, seek customer feedback. Otherwise, rely on A/B tests (and hope your model of the world is correct).

Compounding is Hard to Perceive

If you start going to the gym consistently, you probably stop noticing big changes after the first few months. Your eyes get used to what you're seeing. Seeing someone's reaction one year after is the best reaction outside of a before-after photo. Good habits are the real gains.

The company grew more than 4x in the 2 years since I joined, but it doesn't feel like it. The head count is about the same. The office still looks the same.

Find something you care about and focus on it for a long time.

All the fun happens IRL

COVID showed us we could operate a lot of things remotely. But if you're going to spend most of your waking hours working, it's better to be in-person with the people you're working with.

I drilled wells in the Gulf of Mexico remotely. I learned how to code remotely. I spent my first 6 months at Volley remote.

But I moved to San Francisco to be in-person and there is a world of difference.

• It is easier to ask for and receive help.
• It is easier to build trust.
• And even when you think you're slacking off, you can get unblocked with your work in passing.

SF is the best place for tech. "Go to the Hollywood of your industry."

On San Francisco

I was quite confused the first few times I visited San Francisco.

• The city has outsized financial and cultural impact, but it only has around 800 thousand residents.
• People come here to build companies, but the city is averse to building housing.
• There is incredible wealth being created here every day, but there is jarring inequality visible as you walk around.
• People come here to change the world, but the city looks like it's been neglected.

Lessons so far?

Twitter SF is not Real Life

I got in a robotaxi the second weekend I arrived here and that's become my primary means of transport. Most of the rest of the world still sees them as a problem that will get solved some time in the future.

SF IRL/Twitter tech hype bubble is contagious, and can be toxic if not taken in doses. I remember getting overwhelmed at events in the weeks following the release of GPT-4. I had to cut back on meeting people because it seemed like that's the only thing anyone could talk about. It's important, don't get me wrong. But AI is not normal dinner table conversation outside of SF.

People

It's quite easy to make friends because all the transplants are kind of self-selecting. The high cost of living and challenges (see above) filter for people who have similar ambitions and interests.

All the experts are here and are very accessible. They spend all their free time on hobbies that end up having an outsize impact on the world.

Don't Neglect Your Environment

This city would be a lot different (better) if a fraction the attention paid to tech was paid to the city's governance.

The city would be cleaner, less people would die of overdoses on the street, and petty crimes would drop. There would be more housing built, which would prevent housing costs from rising due to the enforced scarcity. Some people think this is a bad thing. I don't. This place has one of the best climates and sceneries in the country. I think more people should get to enjoy it.

Political involvement in increasing. Primarily as backlash to the excesses (read: incompetence and corruption) in the city government. Hopefully, this is not temporary.

Going Forward

1. I should continue experimenting with stuff. This is the best way to develop intuitions and increase opportunities.
2. I definitely should write and share more. I'm still struggling for momentum here, but this post is another rep. Maybe more bite-sized commentary or a new topic a week. We'll see.
3. Volunteer / say yes to more.