November 16, 2022 — I dislike the term first principles thinking. It's vaguer than it needs to be. I present an alternate term: root thinking. It is shorter, more accurate, and contains a visual:
Sometimes we get something wrong near the root which limits our late stage growth. To reach new heights, we have to backtrack and build up from a different point.
All formal systems can be represented as trees1. First Principles are simply the nodes at the root.
Root thinking becomes more valuable as current growth slows
Technology grows very fast along its trendy branches. But eventually growth slows: there are always ceilings to the current path. As growth begins to slow, the ROI becomes higher for looking back for a path not taken, closer to the root, that could allow humans to reach new heights.
Root thinking isn't as valuable when growth is rapid
If everyone practiced root thinking all the time we would get nowhere. It's hard to know the limits to a current path without going down it. Perhaps we only need 1 in 100, perhaps even fewer, to research and reflect on the current path and see if we have some better choices. I haven't invested much thought yet to what the ideal ratio is, if there is even one.
Notes
1 Tree Notation is one minimal system for representing all structures as trees. ⮐
Update: 7/1/2023
On second thought, I think this idea is bad. The representation of your principles-axioms-agents-types-etc rounds to irrelevant. Infinitely better to spend time making sure you have the correct collection of first principles than worrying about representing them as a "tree" so you can have a visual. It's knowing what the principles are and how they interact that matters most, not some striving for the ideal representation syntax. This post presents a bad idea, but I'll leave it up as a reminder that sometimes I'm dead wrong.