Most First-Principles Thinking Isn't

The phrase has gotten cheap. Every podcast, every founder thread, every conference deck claims to think "from first principles." Most of it is reasoning by analogy in a better hat.

Reasoning by analogy is the default setting of the brain. You see a new thing, you find the closest old thing, you copy what worked there. It's fast, cheap, and right most of the time. It's also why nine out of ten startups in any given year look like the eighth out of ten from the year before.

First-principles thinking is what you do when the analogy is killing you. You strip the problem to what is provably true, ignore everything you inherited, and rebuild from the floor.

Musk's example is the rocket. A rocket "costs" sixty-five million dollars because rockets cost sixty-five million dollars. That's the analogy. The first-principles version is that a rocket is aluminum, copper, carbon fiber, and titanium. The materials cost about two percent of the sticker. The rest is convention, supply chain inertia, and whoever has been getting paid for thirty years. SpaceX is the company that took that arithmetic seriously and then refused to be talked out of it.

The reason most people don't actually do this is that it is slow, painful, and socially expensive. You have to make a list of every assumption sitting in your head and ask, for each one, who first told you it was true. Most of them came from a teacher, a manager, a tweet you half-remember. Some came from a textbook last updated when you were in primary school. The list will be embarrassingly long.

Then you have to keep what survives and throw out the rest. That's the part nobody likes. Every assumption you delete is somebody else's life work, and they will argue with you. You'll spend a quarter being told you are naive. If you're right, you ship something an order of magnitude cheaper. If you're wrong, you feel stupid in public. There is no third door.

The cheap version skips the list. The person says "first principles," gets credit for thinking deeply, and quietly does whatever the analogy was going to suggest anyway. You can spot it because their conclusion always lands inside the Overton window of their industry. Real first-principles thinking lands somewhere weird, like reusable rockets, or a battery factory the size of a small country, or a cancer drug priced at cost.

What I'm trying to do is smaller. Once a quarter, take one belief I act on without questioning, write it down, and ask where it came from. Sometimes it survives. Sometimes I find I've been carrying a cargo-cult version of advice somebody gave me in 2015 about a problem I no longer have.

The practice doesn't make me Elon. It just keeps me from optimizing the fiberglass mat for another year.