Learning to Make Things With a Computer

This was written in a sleep-deprived state during the first few weeks after my daughter’s premature birth.

A lot of the things you work on are new to you. It often involves the interplay between several concepts that you understand vaguely or not at all. It is frustrating and makes you feel stupid; you are not. It just means that in your eagerness to create things, you have wandered into foreign territory. Enjoy the new sights and look for subtle similarities; you may soon be able to read this landscape too.

Don’t forget that the end goal is to make things with the computer – useful things that you and other people enjoy interacting with. Right now, you are focusing on data science and machine learning because it suites your background well and scratches your itch for automation. But don’t forget to focus on the knowledge and practices that enable you to build things.

Remember to look up and assess the landscape. Do you still want to be doing data science and machine learning ten years from now? If so, what do you need to learn to stay valuable? If not, how can you use your current situation to learn the necessary skills to transition? This other thing you want to do, does it need to be your profession, or can it stay a hobby?

It’s OK to have potentially profitable hobbies without ever realizing the potential. Making things because you think they ought to exist is OK, and you don’t have to tell anyone about it either. It can just stay the thing you made and nothing more. It doesn’t make it less valuable.

Celebrate the progress you have made; it didn’t come easy. The many late hours and weekends spent studying and practicing means that you can now do things you would not have thought yourself capable of five years ago. Enjoy the things that you make, and keep on making more. Soon you will be making things that you can hardly imagine today.