First principles thinking

Working Definition

First principles thinking is the procedure of breaking a problem down into propositions that cannot be deduced from anything more basic within the domain at hand—laws of physics, spot prices of raw materials, hard biological or legal constraints—and reassembling a solution solely from those elements while setting aside inherited assumptions and reasoning “by analogy.” In its classical philosophical version, in Aristotle and Descartes, first principles are indemonstrable premises from which demonstrative knowledge must begin; in its contemporary popular version, spread by Elon Musk and Shane Parrish, it is a practical discipline of asking what in a problem is genuinely fixed and what is merely industry convention. The distinction matters in practice: the philosophical version is a thesis about the architecture of human knowledge, whereas the engineering version is a design heuristic promising that decomposition will expose inflated margins or obsolete specifications. In both versions, the power of the method lies not in discovering brand-new facts, but in making visible which accepted constraints are actually negotiable conventions.

Genesis and Global Lineage

Aristotle defined a first principle (archē) in Book Delta of the Metaphysics as the primary point from which a thing exists, comes to be, or is known, adding in the Posterior Analytics that demonstrative science must rest on primary, indemonstrable premises lest justification regress infinitely. René Descartes turned this thesis into a methodological program in the Discourse on the Method (1637) and Principles of Philosophy (1644): demolish the edifice of inherited opinions down to an indubitable foundation and rebuild deductively step by step. Today's circulation of the phrase, however, stems not from academic philosophy but from Silicon Valley engineering folklore. In a 2012 interview with Kevin Rose and a 2013 TED conversation with Chris Anderson, Elon Musk contrasted “reasoning by analogy” with “reasoning from first principles,” illustrating the latter with a back-of-the-envelope calculation: what would the raw materials of a lithium-ion battery cell cost on the London Metal Exchange when finished battery packs cost hundreds of dollars per kilowatt-hour? Tim Urban (2015) turned this into a viral parable contrasting the cook who follows recipes with the chef who designs dishes from raw ingredients, and Shane Parrish codified the method on his Farnam Street blog (2018) and in The Great Mental Models, Volume 1 (2019, with Rhiannon Beaubien). Along the way the concept shifted meaning: from an epistemological claim about foundational truths to a management heuristic for negotiating with legacy costs—and a status badge implying that everyone else merely copies the past.

Central European Lineage

The sharpest philosophical critique of the Cartesian program of “strip down to bedrock and rebuild from scratch” came out of Vienna twice over. Otto Neurath, educated in Vienna and active from around 1910 in the “first Vienna Circle” with Hans Hahn and Philipp Frank before becoming a leading figure of the interwar Vienna Circle, attacked Cartesian foundationalism as “pseudorationalism” in his 1913 Vienna Philosophical Society lecture Die Verirrten des Cartesius und das Auxiliarmotiv (The Lost Wanderers of Descartes and the Auxiliary Motive): neither human action nor scientific knowledge can be derived wholly from a calculus of certain premises. In Protokollsätze (Protocol Sentences, 1932) he added that science has no pure, unassailable starting sentences ("no tabula rasa exists"): we are like sailors who must rebuild their ship on the open sea, never able to dismantle it in dry-dock and reconstruct it from the best materials. Two years later in Logik der Forschung (Vienna, 1934), Karl Popper—who earned his doctorate at the University of Vienna in 1928—closed his analysis of the empirical basis of science with the image of a building erected on piles driven into a swamp: science does not rest on solid bedrock, and we stop driving the piles not when we hit an ultimate foundation, but when we decide they are firm enough to carry the structure for now. Neither thinker was addressing Silicon Valley, yet both dismantle the naive assumption of popular first-principles rhetoric: that beneath the layer of social convention lies an unproblematic layer of pure facts waiting to be uncovered.

Why Now

Three processes between 2018 and 2026 turned the phrase into a ubiquitous slogan. First, industrial disruption: falling launch costs at SpaceX and battery pack costs at Tesla were widely credited to first-principles decomposition, and Walter Isaacson's 2023 biography of Musk popularized his five-step engineering “algorithm,” whose opening rule is to question every requirement and demand the name of the person—not the department—who wrote it. In 2025, that “delete first, then add back what proves indispensable” approach was exported into US federal government restructuring, testing what happens when radical decomposition is applied to complex institutions without domain knowledge. Second, publishing: Parrish's 2019 book and its 2024 Portfolio reissue made “first principles” a staple of executive coaching. Third, artificial intelligence: because large language models are statistical engines for reasoning by analogy across vast corpora, whether they can genuinely reason from first principles became, from 2024 onward, a central benchmark in evaluating “reasoning models.”

Explanatory Power

The method earns its keep whenever it separates the component of a price, procedure, or belief dictated by physics from the component dictated by institutional history. In Musk's battery calculation, the spot-market value of cobalt, nickel, aluminum, carbon, and steel was a small fraction of a finished pack's price; everything else was manufacturing organization, supply-chain margins, and legacy tooling—and that remainder was negotiable. Inside organizations, the same lens exposes “zombie requirements”: compliance steps or specifications that no living engineer can trace to a physical risk or legal mandate, yet everyone obeys out of habit. In public policy, decomposing urban housing costs into land, materials, labor, financing, and permitting delays reveals which slice is physical constraint and which is regulatory choice, turning a vague quarrel over “unaffordable housing” into an auditable dispute over specific line items.

Axes of Debate

Underappreciated Angle

First-principles decomposition is cognitively and organizationally expensive, and its cost rises the further a domain sits from physics and chemistry. To break a problem down into atomic truths, you must already know what the “atoms” of that domain are: in battery electrochemistry or rocket propulsion we know what they are, whereas in education, macroeconomics, or public administration we do not—and in those fields what speakers call “first principles” almost always turn out to be their own ideological axioms delivered in the tone of thermodynamics. Consequently, the practical question is never how to think from first principles, but where it is worth paying the price to do so. Reasoning by analogy and copying inherited templates are cheap and indispensable for most decisions; the real skill lies in spotting the rare bottleneck where convention has drifted so far from physical or economic reality that dismantling the problem in dry-dock will pay for itself.