First-Principles Thinking
Working up from what you know is true, instead of across from what everyone does.
The behaviour, in one line: I question “that’s just how it’s done” and reason problems up from the facts I’m actually sure of.
A battery costs what batteries cost, until someone asks what the metals inside it cost. The second number is a floor. The first is a habit. First-principles thinking is the move from the first number to the second.
Most thinking is done by analogy: you do it this way because this is how it is done. That is fast, and it carries every assumption of whoever did it first. First-principles thinking is the slower move — take the problem down to the things you are actually sure of, then build back up. What you get is the set of answers convention was hiding, and now and then the discovery that the constraint everyone works around is not real. Analogy is now free. A model will hand you the standard answer in seconds, and the standard answer is the average of everything done before. Working from first principles is the reliable way to arrive somewhere that average does not go, and the only way to notice when a constraint the whole field respects — cost, speed, what one person can do in a day — has quietly stopped being true.
| What it is | What it isn’t |
|---|---|
| Reasoning up from what is true | Reasoning across from what is done |
| “Required by what?” | “That is how it is done” |
| Rebuilding from the base | Editing the existing answer |
| Slow, and occasionally new | Fast, and never new |
What it looks like. A quote comes back at four times your budget. Instead of negotiating, you ask what the thing is made of and what those parts cost. Two of the four line items turn out to be the supplier’s process rather than anything you needed.
- The constraint everyone works around. Every team has one, and it usually predates everyone in the room. Nobody removes it because nobody has checked whether it is still there.
- Pricing, estimates and “impossible”. All three are normally built by analogy to the last one. A task that took a week now takes an afternoon, and the estimate still says a week. The floor is somewhere else entirely.
- A field whose assumptions just changed. When the ground moves, reasoning by analogy still points at the old ground. Reasoning from base facts is the only thing that repoints.
- Sort the problem into what you know, what you are assuming, and what you inherited. The third pile is usually the largest.
- Meet every “we have to” with a question: required by what? A law of physics and a habit of the industry look identical until you ask.
- Rebuild the answer from what survived, rather than editing the answer that already exists.
- Then question your own reconstruction the same way. The first rebuild usually smuggles something in.
- Break it into parts
Three parts: sorting fact from assumption, interrogating every “we have to”, and rebuilding upward from what survives.
- The part that carries it
The “we have to” question. Most inherited constraints turn out to be conventions, and one question is enough to tell them apart.
- The order to learn it in
- Practise sorting a problem into known, assumed, and merely inherited.
- Then challenge each “we have to” with what law or fact actually requires it.
- Only then rebuild from the survivors, instead of editing the answer that already exists.
- Put something at stake
Take one “that is just how it is done” in your team, build the from-scratch alternative, and present both side by side.
- Compress it
What do I know? What am I assuming? What does the base truth allow that the convention forbids?
- The smallest useful dose
One constraint interrogated for every project decision that genuinely matters.
- Anchor it to something you know
A chemist pricing a battery by the metals inside it rather than by the sticker on it. Same object, completely different floor.
“The least initial deviation from the truth is multiplied later a thousandfold.”
— Aristotle, On the Heavens
The interactive version is in the course. The diagram, the simulation and the four-question check for First-Principles Thinking run in the free course, where your progress across all 24 skills is kept.
Open First-Principles Thinking in the course →
Not sure where you stand? Take the twelve-minute self-assessment, or see the 30-day course that turns these skills into a roadmap.