Created: July 10, 2020
Modified: July 10, 2020
Modified: July 10, 2020
surprises in having a job
This page is from my personal notes, and has not been specifically reviewed for public consumption. It might be incomplete, wrong, outdated, or stupid. Caveat lector.- What have I learned in 2.5 years at Google?
- What did I not realize?
- The model of research.
- How low the expectations are.
- How fake it felt to say 'thank you' for everything, especially at a point when I wasn't really thankful for anyone trying to help me do software engineering.
- How good it feels to work in a culture where everyone is that nice.
- What have I come to accept that I would not have agreed with before?
- I now believe in code review. I sometimes learn from the review itself, and often it leads to better code. But even if it were useless, it enforces a discipline on how you write code. For any piece, there has to be enough of a story why it exists, what role it plays, and how it works, that another person can understand. And it turns out that story is really helpful even if you're the only one working on a project. First, the existence of the review forces you to interrogate your own story. You have to tell a story to your future self about how things work.
- It's cool that I've now gotten close to the point where I am literally working on Bayesflow for a living. I'm getting paid enormously to do a version of my grad school hobby project that many people will, and do, actually use! I get to read papers and think about inference algorithms. And then I get to talk with NameRedacted about them. It's better than working on the Stan team would have been. If I saw it as my hobby project---and kept thinking about it and stepping back to articulate new directions and visions for it---I could take ownership of it like never before. the system wants you to have ownership. I'm in a space where I could learn a lot about how to do it really well. And if it's good it really can have a lot of impact.
- What does TFP need?