# First Principles Thinking: The Discipline of Reasoning from Bedrock

In 2002, Elon Musk wanted to buy a rocket. He flew to Russia three times to negotiate with companies selling refurbished intercontinental ballistic missiles. The asking price was $18 million per rocket -- roughly in line with the broader market, where launch costs had hovered around $10,000 per kilogram to orbit for decades. The entire aerospace industry treated these numbers as fixed, the way you'd treat the speed of light. On the flight home from his third failed negotiation, Musk opened a spreadsheet and listed the raw materials in a rocket: aerospace-grade aluminum alloys, titanium, copper, carbon fiber, liquid oxygen, kerosene. He looked up their commodity prices. The total came to roughly two percent of the price of a finished rocket. The gap between two percent and one hundred percent wasn't physics. It was convention. SpaceX launched its first rocket four years later.

**First principles thinking** is the practice of deconstructing a problem down to its foundational truths -- the facts that remain after every assumption, convention, and inherited wisdom has been stripped away -- and then reasoning upward from those truths to build new solutions. The term traces to **Aristotle**, who defined a first principle as "the first basis from which a thing is known," the irreducible foundation beneath all further explanation. This is not the same as **critical thinking**, which evaluates the quality of existing arguments and evidence. First principles thinking doesn't evaluate existing answers; it abandons them entirely and reconstructs from scratch.

The distinction matters because most reasoning operates by **analogy**: you look at how something has been done before, adopt that approach, and make incremental adjustments. Analogy is efficient -- it borrows work others have already completed. But it is also a cage. When you reason by analogy, you inherit not only the solutions of your predecessors but also their unexamined assumptions, their constraints, and their blind spots. First principles thinking breaks the cage by asking a different question entirely: not "how has this been done?" but "what is actually, demonstrably true, and what can I build from there?"

## Why Decomposition Unlocks Innovation

The mechanism behind first principles thinking is rooted in what cognitive scientists call **functional fixedness** -- the tendency to see objects and ideas only in terms of their conventional use. Karl Duncker demonstrated this in his famous 1945 candle problem experiment at Clark University: participants given a box of thumbtacks, a candle, and matches struggled to attach the candle to a wall, because they saw the box as a container for tacks rather than as a potential shelf. When the tacks were presented outside the box, success rates doubled. The box hadn't changed. What changed was that participants could see it as raw material rather than a predefined component.

First principles thinking applies this same liberation at scale. When you decompose a problem into its fundamental elements -- the physics, the chemistry, the basic economics -- you free those elements from the functional fixedness imposed by existing solutions. The raw materials of a battery are not "a battery." They are cobalt, nickel, aluminum, carbon, and a steel can, each with an independent commodity price. The raw materials of a university education are not "a degree." They are knowledge, credentialing, social networks, and structured practice. Once you see the elements rather than the assembled package, you can reassemble them in configurations that the conventional blueprint would never produce. This is why first principles thinking is closely related to **lateral thinking**, though they arrive at unconventional solutions differently: first principles decomposes downward to fundamentals and rebuilds, while lateral thinking moves sideways through provocation and reframing without necessarily reaching bedrock.

## Two Examples: One Boardroom, One Kitchen

The most documented corporate example is **Tesla's battery cost problem**. When Musk entered the electric vehicle market around 2004, the prevailing wisdom held that lithium-ion battery packs cost roughly $600 per kilowatt-hour. Industry incumbents accepted this figure as given and focused on squeezing out single-digit percentage improvements through manufacturing tweaks. Musk decomposed the battery to its commodity inputs and found raw material costs of approximately $80 per kilowatt-hour. The $520 gap was composed of supply chain markups, manufacturing process conventions inherited from consumer electronics, and a procurement structure no one had redesigned for automotive scale. By reasoning from the material costs upward -- building a dedicated gigafactory, redesigning cell formats, vertically integrating supply chains -- Tesla drove battery costs below $140 per kilowatt-hour within a decade, fundamentally altering the economics of electric transportation.

At a personal scale, consider the story of **Sara Blakely**, founder of Spanx. In the late 1990s, Blakely was a door-to-door fax machine saleswoman who wanted to wear white pants without visible panty lines. The hosiery industry's assumption was that undergarments were sold through established channels, designed by fashion industry insiders, and manufactured using existing product templates. Blakely had no industry experience and no fashion credentials. She started instead from a first principle: what does the garment actually need to do? The answer was compression and smoothing, with no foot section. She cut the feet off her pantyhose, spent $5,000 navigating patent law by reading books from Barnes & Noble, and personally pitched the product to a Neiman Marcus buyer by demonstrating the before-and-after in a restroom. Spanx became a billion-dollar company. Every hosiery incumbent had access to the same materials and the same customer need. What they lacked was a willingness to discard the assumptions embedded in "how undergarments are designed and sold."

## How to Decompose: The Actual Practice

The process has a structure, but applying it feels less like following steps and more like holding your breath underwater. You are deliberately cutting yourself off from the scaffolding of existing solutions, and the initial sensation is disorientation.

Start by writing down every constraint you believe applies to your problem. Be exhaustive. Then interrogate each constraint with a single question: is this a law of nature, or is it a convention that humans invented and could therefore reinvent? Gravity is a law. "Rockets have always been expendable" is a convention. Chemistry dictates what reactions store energy. "Batteries must be purchased from existing suppliers" is a convention. The fundamental truths are your building blocks. Everything else is negotiable.

From those building blocks, reconstruct your solution without referencing existing approaches. This is the creative and difficult phase, because you are working without a template. It requires comfort with what feels like starting from nothing -- and this is where many people abandon the process and retreat to analogy. The discomfort is the signal that you're doing it right.

Finally, validate your reconstruction against reality. First principles thinking produces hypotheses, not certainties. The hypothesis still needs testing. This is where the discipline connects to **Bayesian thinking** -- you've built a model from foundational truths, but you must update that model as evidence from the real world comes in. SpaceX's first three rockets failed. The first principles reasoning was sound; the engineering execution needed iteration.

## Where This Breaks Down

First principles thinking is expensive, and misapplying it produces distinctive failure modes.

The most common is **using it where analogy would serve better**. Not every problem deserves decomposition to fundamentals. If you're choosing a restaurant for dinner, reasoning from first principles about nutrition, culinary chemistry, and agricultural supply chains is absurd. The skill is not in applying first principles universally but in recognizing when inherited assumptions are actively constraining your solution space. The concept of **opportunity cost** is the relevant check here: the hours spent decomposing a trivial problem are hours not spent on the problems where decomposition would yield genuine breakthroughs.

The second failure mode is **overconfidence in your own fundamentals**. You believe you've reached bedrock, but you've actually stopped at a deeper layer of assumption. A physicist reasoning about battery chemistry is genuinely working from first principles. An entrepreneur who read a blog post about battery chemistry and now believes they've identified the fundamental truths is reasoning from a thinner foundation than they realize. This connects to **metacognition** -- the practice of examining your own thinking process, including whether your claimed "first principles" are truly fundamental or just unfamiliar-sounding assumptions.

Third, first principles thinking can produce solutions that are **technically correct but practically unimplementable**. Knowing that rocket raw materials cost two percent of the launch price doesn't tell you how to build a factory, recruit engineers, navigate FAA regulations, or survive three consecutive explosions without running out of funding. The gap between "this should be possible in principle" and "this works in practice" has buried many first-principles thinkers who underestimated implementation complexity. **Analytical depth** -- the discipline of pushing past the first adequate explanation -- is the complement here, because it forces you to examine whether your clean first-principles model accounts for the messy layers of reality between theory and execution.

Fourth, it can become **an excuse for ignoring expertise**. The history of an industry encodes genuine hard-won knowledge, not just arbitrary convention. Someone who dismisses all existing practice as "reasoning by analogy" may be discarding solutions that are conventional precisely because they work. First principles thinking is most powerful when combined with deep domain knowledge -- when you understand *why* things are done a certain way well enough to identify which conventions are load-bearing and which are vestigial.

Finally, first principles reasoning intersects dangerously with the **sunk cost fallacy**. When your decomposition reveals that an approach you've already invested heavily in rests on flawed assumptions, the temptation to protect that investment -- to treat your sunk costs as a reason to continue -- is the exact opposite of what first principles thinking demands. Recognizing this trap requires the kind of self-awareness described in **metacognition**: you must examine not just the problem but your own emotional attachment to previous solutions.

## The Bedrock Test

The next time you encounter a problem where the standard approaches feel inadequate -- where you sense that the constraints everyone accepts might not be real constraints -- try this self-test: **"Would I still believe this constraint if I had built this system from scratch today?"** If the answer is no, you've found an assumption worth decomposing. The feeling you're looking for is a specific kind of disorientation: you've removed the scaffolding of existing solutions, and for a moment, the problem feels both simpler and more exposed than it did before. That disorientation is the space between analogy and invention. It's the space where Musk sat on a flight home from Russia with a spreadsheet, where Blakely stood in a department store restroom with a pair of modified pantyhose. The gap between what things cost and what they're made of, between how things are done and how they could be done, is almost always larger than it appears. First principles thinking is the discipline of measuring that gap -- and then building in it.

*v1.0.0*