Working Programmatically: An Experiment for 2023

Working Programmatically: An Experiment for 2023

Self Development

Recently, an unformed but interesting idea has been flittering about the dark recesses of my tired and over stimulated mind:

What would working programmatically look like?

This is a vague and malformed question, but I will try to give it some structure by introducing some components that are contributing to its formulation.

I want to do more Deep Work. I've been inspired by posts like this one from Nick Winter. I'm an AI Research Scientist by day, and as far as jobs go that probably takes the cake in excusing large amounts of thinking and programming time during the day.

I unfortunately also believe that meaningful science shouldn't stop at a paper and a few clicks, so I'm stuck in the daunting labyrinth of commercializing, as building products and technologies that actually impact people's lives ultimately feels like the most value generative thing to do the way society is set up. This means a lot of time working on company building, pitching, strategic capture of markets, product design, culture review and a whole load of logistics overhead to manage all that capital (human, temporal, and monetary).

So, we're at sort of an orthogonal point between a lot of these directions. Thus, the conception of the idea of working programmatically.

Surely, if I tend to use the same tools I do for development and research to handle other things in my life, I may remove some marginal cruft of inefficiency from my mind. Note that I use a loose definition of tools here, across several levels of abstraction: my computer, prosumer (read: developer minded) productivity tools, code editors for note taking AND programming, and the occasional script to try to automate needless crap I get mired in.

I know this is a badly formulated idea with tonnes of conceptual asymmetries and illogical premises, but its inherent amorphous nature allows for a few concrete benefits:

  • I look and feel like a hacker, despite getting no real positive delta in work completed.
  • I get to see vanity metrics on time tracking tools like Rescue Time, and share them for clout. Heck, my Github will finally be all green!
  • If I ever do well in my endeavor to develop humanity's tech tree and contribute to the advancement of our species, people will blur history and say "Of course Harsh built AGI, he was a computer savant! The dude used computers in ways nobody else did!" In the event this happens, this post will have been long removed, and anyone who read it encouraged to forget it.

Someday, I might better define this, providing an intellectually honest premise and corresponding lessons that others may benefit from. But I wouldn't count on it!

To 2023, the year of programmatic work! 🍻