Principles: In 2002, Elon Musk traveled to Russia to buy refurbished intercontinental ballistic missiles to use as rockets.[3] Russian sellers quoted him prices in the tens of millions of dollars. Rather than accepting these prices as fixed reality, Musk asked a different question: what are rockets actually made of?
Aerospace-grade aluminum, titanium, copper, carbon fiber. On commodity markets, those raw materials cost about two percent of the quoted launch price.
The gap between material cost and market price was not a law of physics - it was an accumulation of assumptions, inefficiencies, and conventions that nobody had bothered to challenge. SpaceX now launches rockets at a fraction of the legacy cost.
This is first principles thinking in action: decomposing a problem to its fundamental truths and rebuilding from there, rather than accepting the world as presented.
Why Analogy Thinking Dominates - and Why That Matters
Most human reasoning operates by analogy. We see a new situation, pattern-match it to something familiar, and apply whatever worked before. This is efficient, low-effort, and correct often enough to be our default mode.[4] When a company needs a marketing strategy, the instinct is to study what competitors did.
When an engineer faces a design challenge, the first move is to look at existing solutions. When a founder builds a startup, the pitch often begins with "We're the Uber of X."
"I think it's important to reason from first principles rather than by analogy. The normal way we conduct our lives is we reason by analogy. We are doing this because it's like something else that was done, or it is like what other people are doing. It's mentally easier." - Elon Musk
Analogy thinking works well for routine problems. But it carries a hidden cost: it imports the assumptions embedded in previous solutions.
If those assumptions were never questioned - if they accumulated through historical accident rather than deliberate design - then reasoning by analogy perpetuates limitations that may no longer apply.
The result is incremental improvement when fundamental rethinking could unlock entirely different possibilities.
The Core Framework: Decompose, Verify, Rebuild
First principles thinking is a structured process with three distinct phases. Understanding each phase matters because most failures happen when people skip steps or confuse assumptions for fundamentals.
Phase 1: Identify and Challenge Assumptions
Every problem arrives wrapped in assumptions. Some are explicit ("batteries are expensive"), but most are invisible - embedded so deeply in how an industry operates that nobody thinks to question them. The first step is surfacing these assumptions and distinguishing them from verified truths.
The Socratic method provides a systematic approach: Why is this the case? What evidence supports it? Is it always true? What are we assuming? What if the opposite were true? Each question peels back a layer of convention.
Consider the assumption "online education is less effective than in-person learning." Questioning reveals that the evidence actually shows lower completion rates, not lower learning effectiveness.
And the courses with high completion rates share specific design elements - community features, accountability mechanisms, practical application - that have nothing to do with the online format itself. The assumption conflates format with design quality.
Phase 2: Break Down to Fundamental Truths
Once assumptions are identified, the task is to find what remains - the constraints that cannot be violated without breaking reality. These fall into three categories:
| Category | Description | Example | Can Be Challenged? |
|---|---|---|---|
| Fundamental constraints | Laws of physics, mathematics, verified cognitive science | Human working memory holds ~7 items | No |
| Probable truths | Evidence-based but potentially revisable | Remote workers report higher productivity in surveys | With better evidence |
| Assumptions and conventions | Beliefs without strong evidence, historical practices | Companies need physical offices | Yes - aggressively |
The third category is where most opportunity lives. When Tesla examined battery costs, the fundamental constraint was material cost (~$80/kWh). The market price ($600/kWh) was the product of manufacturing conventions, supply chain structures, and low production volumes - all challengeable.
Phase 3: Reason Up from Foundations
With only verified fundamentals in hand, you construct solutions without importing constraints from existing approaches. This is the creative step, and it often produces ideas that seem obvious in retrospect but were invisible through the lens of analogy.
"The formulation of a problem is often more essential than its solution, which may be merely a matter of mathematical or experimental skill." - Albert Einstein
Amazon Web Services emerged from this process. Amazon's computing needs were growing expensive. The conventional response (analogy thinking) would have been to optimize existing infrastructure. Instead, the team asked: what is computing infrastructure fundamentally?
Computational power, storage, and networking - commoditizable resources. Why should every company build and maintain its own? The first principles answer led to cloud computing as a utility, a market now worth hundreds of billions.
The Five Whys and Component Breakdown
Two practical techniques make first principles thinking more accessible for everyday problems.
The Five Whys drives past surface explanations to root causes.[5] "We need to hire more customer support staff." Why? Ticket volume is too high. Why? Customers have too many questions. Why? The interface confuses them. Why? Feature additions were prioritized over usability.
Why? The team assumed more features drive growth more than experience does. The fundamental issue is an untested assumption about growth drivers, not a staffing shortage.
Component breakdown works for cost problems. When someone says "creating video content is expensive," breaking it into components reveals: camera (one-time, depreciating), lighting (one-time), editing software (monthly subscription), time (hours multiplied by rate).
Most of the "expense" is time, and equipment costs decrease per video produced. The framing of the problem changes entirely once you see the components.
Historical Roots and Philosophical Origins
The approach Musk rediscovered has roots in ancient philosophy. Aristotle, in his Posterior Analytics, described first principles (archai) as the foundational truths from which all other knowledge in a domain is derived.[1]
You cannot demonstrate first principles through further argument - they are known directly or through induction from experience.
Rene Descartes formalized this into a method in his 1637 Discourse on the Method: accept nothing as true that you cannot clearly and distinctly perceive as true; divide every problem into as many parts as possible; proceed from the simplest and most knowable things to the more complex.[2]
This is first principles thinking in philosophical form.
Richard Feynman, the Nobel Prize-winning physicist who taught at Caltech from 1951 to 1988, embodied the method in physics education.
His approach to teaching quantum mechanics - starting from fundamental observations and deriving the formalism, rather than presenting the formalism and then showing what it explains - produced some of the most accessible physics education ever created.
The Feynman Technique (if you can't explain it simply, you don't understand it well enough) is first principles thinking applied to knowledge assessment.[7]
Where First Principles Thinking Creates Leverage
First principles thinking is not equally useful everywhere. It creates the most value in specific situations:
When existing solutions are inadequate. If current approaches work well, first principles analysis is unnecessary overhead.
But when an industry operates on obviously outdated assumptions - as transportation did before Uber, or hospitality before Airbnb - the gap between fundamental requirements and existing solutions represents opportunity.
When costs seem fixed but should not be. Any time the explanation for high costs is "that's just what it costs," first principles analysis often reveals a gap between material reality and market pricing.
This is where second-order thinking compounds the advantage: reducing costs from first principles changes competitive dynamics, which changes customer expectations, which reshapes the entire market.
When entering new domains. Newcomers lack the analogies that domain experts rely on - and that is sometimes an advantage.[8] The Wright Brothers did not succeed by studying bird anatomy more carefully than their competitors. They built wind tunnels, measured aerodynamic forces directly, and engineered from physics.
Their lack of aviation analogies forced them to reason from fundamentals, which turned out to be more productive than copying nature.
*Example*: Jeff Bezos applied first principles thinking to book retail in 1994. The conventional assumption was that a bookstore could only carry as many books as its physical shelf space allowed. Bezos questioned this: what is the fundamental constraint?
Not shelf space - distribution and customer reach. The internet removed those constraints. Amazon's initial advantage was not superior book-selling; it was questioning the physical constraint that every existing competitor took for granted.
First principles analysis revealed that the constraint was an artifact of physical stores, not of book retail itself.
Case Study: How SpaceX Reconstructed Rocket Manufacturing
The SpaceX case deserves deeper examination because it illustrates all three phases of first principles thinking operating together over an extended period.
Phase 1 (identifying assumptions): The assumption that rocket components require specialized aerospace suppliers charging aerospace-grade markups. SpaceX questioned every component.
Their question was not "who makes the best rocket engine valves?" but "what does a rocket engine valve actually need to do, and what is the fundamental cost of making something that does that?" The assumption of specialized supply was revealed as a convention, not a requirement.
Phase 2 (finding fundamentals): The fundamental costs were materials (knowable from commodity prices) plus precision manufacturing (quantifiable from industrial engineering). The markup above those fundamentals was the target for reduction.
Phase 3 (rebuilding): SpaceX built vertical integration - manufacturing most components in-house - specifically because the assumption that aerospace companies should buy rather than build was itself an assumption, not a law.
Manufacturing in-house allows direct optimization from fundamentals rather than from what suppliers are willing to charge.
The result: SpaceX's Falcon 9 launch costs roughly $2,720 per kilogram to low Earth orbit, compared to the Space Shuttle's cost of approximately $54,500 per kilogram. The fundamental physics did not change. The assumptions that determined cost did.
The Failure Modes
First principles thinking has a reputation problem: it sounds like pure intellectual heroism, but practiced poorly, it produces worse results than conventional thinking. Understanding the failure modes matters as much as understanding the method.
Confusing preferences with principles. "Users want simple interfaces" is not a first principle - it is a generalization that varies by context. "Human working memory is limited to roughly seven items" (Miller, 1956) is a widely cited cognitive-science heuristic, though later research (for example Cowan, 2001) has argued the practical limit for most tasks is closer to about four chunks; either way it represents a real cognitive constraint that genuinely constrains design.
The distinction matters because building on preferences produces fragile conclusions, while building on constraints produces robust ones.
Stopping too early. Many people think they have reached first principles when they are still several layers of assumption deep. "Cars need to be heavy to be safe" sounds fundamental but is not.
The actual principle is "crash forces must be dissipated" - and crumple zones, advanced materials, and structural design can achieve this without excess weight. If your "principle" can be challenged with "but why is that true?", you have not gone deep enough.
Arrogance toward domain expertise. First principles thinking sometimes becomes an excuse to ignore people who actually know the domain. Building codes exist because buildings used to collapse and kill people. They are not arbitrary conventions - they are lessons encoded from fatal failures.
A newcomer reasoning from first principles about construction who ignores building codes is not being innovative; they are being dangerous. The best practitioners use first principles to question which conventions are load-bearing and which are vestigial.
"In theory, there is no difference between theory and practice. In practice, there is." - Yogi Berra
Analysis paralysis. First principles thinking can become an infinite regression. You can always ask "why" one more time. At some point, you need to stop decomposing and start building - then test whether your first principles reasoning actually holds in practice.
The technique is a tool for generating hypotheses, not a substitute for empirical validation.
The Expert Pattern: Strategic Application
The most effective practitioners of first principles thinking do not apply it to everything. They use it surgically, targeting the highest-leverage constraints while using conventional thinking elsewhere.
SpaceX applied first principles reasoning to rocket reusability - the single biggest cost driver in space launch. But they used conventional aerospace engineering for heat shields, guidance systems, and dozens of other subsystems where existing solutions were adequate.
The insight is not "question everything" but rather "question the thing that matters most."
This strategic application follows a pattern: map the current state, identify where the largest gap exists between what fundamentals suggest should be possible and what currently exists, apply first principles reasoning to that specific gap, and use conventional approaches everywhere else.
The leverage point is where first principles thinking should be concentrated.[6]
*Example*: Reed Hastings applied first principles to video rental in 1997. The conventional model (Blockbuster) charged late fees. Hastings questioned the fundamental: what do customers actually want from video rental? Access to titles, convenient return, affordable pricing.
Late fees addressed none of these - they were a convention that turned out to reflect Blockbuster's need to manage physical inventory turnover, not customers' actual needs.
Netflix's subscription model with no late fees emerged from questioning the late fee assumption and rebuilding from fundamentals. By 2010, Blockbuster was bankrupt.
When Not to Think from First Principles
The opportunity cost of first principles thinking is time and effort. For most decisions, the cost exceeds the value.
First principles analysis is not needed when:
- Existing solutions work adequately and the cost of improvement is low
- The time cost of decomposition exceeds the value of the potential improvement
- The domain has already been well-analyzed and its constraints are well-understood
- The decision is reversible and iteration is cheap
For the vast majority of everyday decisions, the right approach is analogy: learn from what works, follow established best practices, and adapt rather than reinvent.
The skill is recognizing the relatively rare situations where the assumptions underlying existing practice are genuinely wrong - where the first principles suggest something importantly different from what convention produces.
The companies and innovations that changed industries - SpaceX, Tesla, Amazon, PayPal - did not emerge from people who rejected all existing knowledge.
They emerged from people who identified the one or two critical assumptions holding an entire system in place and challenged precisely those assumptions while accepting everything else. The skill is not in the decomposition itself but in knowing where to aim it.
The uncomfortable truth about first principles thinking is that it requires genuine intellectual honesty. You must be willing to challenge your own assumptions as rigorously as you challenge others' - and that is far harder than it sounds.[9]
Most people who claim to think from first principles are actually reasoning from different analogies and calling it fundamental. The test is simple: can you state your principles in terms of physics, mathematics, or verified empirical findings? If not, you may still be operating on assumptions - just different ones.
Sources & Further Reading
- Aristotle. Posterior Analytics. Translated by G.R.G. Mure. Oxford University Press, 350 BCE/1930. View source
- Descartes, R. Discourse on the Method of Rightly Conducting One's Reason and of Seeking Truth in the Sciences. 1637.
- Vance, A. Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future. Ecco Press, 2015. View source
- Kahneman, D. Thinking, Fast and Slow. Farrar, Straus and Giroux, 2011. View source
- Ohno, T. Toyota Production System: Beyond Large-Scale Production. Productivity Press, 1988. View source
- Meadows, D. Thinking in Systems: A Primer. Chelsea Green Publishing, 2008. View source
- Feynman, R. The Character of Physical Law. MIT Press, 1965. View source
- Christensen, C. The Innovator's Dilemma. Harvard Business Review Press, 1997. View source
- Munger, C. Poor Charlie's Almanack: The Wit and Wisdom of Charles T. Munger. Donning Company Publishers, 2005. View source
Further Reading
- Polya, G. How to Solve It: A New Aspect of Mathematical Method. Princeton University Press, 1945. View source
Faraday, Maxwell, and the First Principles of Electromagnetism
The history of physics provides the clearest examples of first principles thinking because physics is the domain where the method has been most rigorously applied and its results most precisely verified. The development of electromagnetic theory in the nineteenth century illustrates both the power and the required expertise.
Michael Faraday had almost no formal mathematical training. Working at the Royal Institution in London, he conducted experiments with magnets and wires, discovering that changing magnetic fields produce electric currents and that electric currents produce magnetic fields.
His records of these experiments are striking for their geometric and spatial character - Faraday thought in terms of "lines of force," physical curves that he could almost see bending through space. He had no mathematical framework to express these intuitions rigorously.
James Clerk Maxwell, about forty years younger, encountered Faraday's experimental results and decided to translate them into mathematics without importing any analogies from other physical domains. He began from what Faraday's experiments demonstrated: the relationships between changing electric and magnetic fields.
Working from these relationships as his axioms - not from any prior theory of what electricity or magnetism "really were" - he derived a set of four equations.
The equations predicted that electromagnetic disturbances should propagate through space as waves traveling at a speed that fell out of the mathematics as approximately 3 x 10^8 meters per second. Maxwell recognized this was the speed of light.
He had not set out to explain light. He had worked from first principles on electricity and magnetism, and light emerged as a consequence. The derivation united optics, electricity, and magnetism - three previously separate fields - from four equations derived from Faraday's experimental observations.
This example illustrates what genuine first principles thinking requires and what it can achieve. It required Faraday's careful experiment to establish the foundational relationships. It required Maxwell's mathematical sophistication to express those relationships without corruption from prior analogies.
The result was not incremental improvement but unification of previously separate domains - the characteristic signature of successful first principles reasoning.
The Battery Cost Curve: Quantifying What Was Thought Impossible
First principles analysis produced a specific, falsifiable prediction about battery costs that consensus opinion rejected - and that has since been confirmed with unusual precision.
In 2010, the consensus forecast from energy analysts for lithium-ion battery costs was approximately $1,000 per kilowatt-hour, with projections suggesting costs might fall to $500/kWh by 2020.
The reasoning was analogical: batteries are complex chemical devices, manufacturing at scale is difficult, and cost curves for comparable products had followed gradual trajectories.
The first principles analysis asked different questions. What are the actual physical inputs to a battery cell? Lithium, cobalt, nickel, manganese, graphite, aluminum, copper, and a polymer separator. What do these materials cost on commodity markets? What does the manufacturing process require that could be standardized or automated?
This analysis, pursued systematically by Tesla's engineering team and independently by analysts at Bloomberg New Energy Finance, suggested that the conventional forecast was based on current production methods and current scale rather than on the physical and economic constraints that bound the true cost floor.
The materials themselves were not expensive. The expense was in small-scale, low-automation manufacturing spread across a fragmented supply chain.
Bloomberg's BNEF team published a first principles-grounded cost trajectory in 2012 predicting that battery costs would fall to approximately $100/kWh by 2030. The prediction was widely dismissed as too aggressive.
By 2023, the actual cost had reached approximately $139/kWh, tracking the first principles projection rather than the analogical one.
In this instance, the first-principles projection tracked actual costs more closely than an analogy-based one, illustrating how the two approaches ask different questions - "how have similar things changed?" versus "what must this cost given what it is made of?" - though a single case cannot establish that first-principles reasoning is generally more accurate than analogical reasoning.