first-principles-skill

First Principles · Thinking Tool

"The first principle is that you must not fool yourself — and you are the easiest person to fool." — Richard Feynman

Activation Conditions and Trigger Words

• Direct invocation: "Analyze using first principles..." "First principles breakdown..."
• Semantic trigger: the user faces a complex problem and seeks the root cause, or challenges existing assumptions
• Combined invocation: "First break it down using first principles, then analyze it using systems thinking"

Method Framework Overview

The core of first principles is to return to basic facts and re-derive; it breaks through mental inertia and industry conventions through four principles: reducing to physical/logical limits, separating the known from the assumed, reasoning up from basic facts, and guarding against the analogy trap.

Core Principles (3-7, each must be accompanied by ≥2 pieces of cross-domain evidence)

Principle 1: Reduction to Fundamental Truths

One-sentence definition: Break a problem down to irreducible physical or logical fundamental truths.

Cross-domain evidence：

1. Aerospace: SpaceX calculated rocket manufacturing costs starting from the cost of aluminum materials, reducing launch costs to 1/10 (Source: Musk 2012 Wired interview)
2. Automotive: Tesla derived battery pack costs from the spot prices of lithium, cobalt, and aluminum raw materials, finding they could be reduced by 80% (Source: Musk 2013 Tesla Investor Day)

How to apply: When facing any problem, first list the irreducible basic facts that constitute the problem, ignoring the intermediate conclusions conventionally accepted in the industry.

Limitation: In some fields the basic facts themselves are disputed (such as economics); in this case the reduction target is not physical facts but logical anchor points.

Principle 2: Separate Known from Assumed

One-sentence definition: Strictly separate verified facts from untested assumptions.

Cross-domain evidence：

1. Philosophy: Descartes systematized the method of doubt in Meditations on First Philosophy (1641), taking "I think, therefore I am" as the only indubitable anchor point (Source: Descartes, Meditations on First Philosophy)
2. Scientific methodology: Galileo distinguished "what Aristotle said" from "what experiments measured," overturning the assumption that heavy objects fall first (Source: Galileo, Two New Sciences, 1638)

How to apply: List all assumptions of the current plan and mark each one's verification status. Three types of labels: verified (has data), verifiable (has a method), unverifiable (dangerous).

Limitation: Excessive differentiation may lead to analysis paralysis; one must judge the cost and benefit of verification.

Principle 3: Reason Up from Fundamentals

One-sentence definition: Rather than reasoning from analogy or existing experience, build an entirely new solution from basic facts.

Cross-domain evidence：

1. Retail: Amazon derived the e-commerce model from "users want cheap/more choices/convenient," rather than analogizing Walmart (Source: Bezos 1997 shareholder letter)
2. Blockchain: Satoshi derived the PoW solution from "the basic problem of decentralized consensus," rather than improving the banking system (Source: Bitcoin Whitepaper, 2008)

How to apply: Start from basic facts and ask "Based on these facts, what is the optimal solution?" It is forbidden to use "everyone else does it this way" as a basis for reasoning.

Limitations: Deriving entirely from scratch may overlook implicit industry knowledge, so after derivation it is necessary to compare with existing solutions.

Principle 4: Beware of Analogy Trap

One-sentence definition: "This is like that" is humanity's most dangerous thinking shortcut.

Cross-domain evidence：

1. Technology field: The first-generation iPhone was not a "shrunk Mac," but a new paradigm derived from the basic facts of touch interaction (Source: Jobs 2007 iPhone launch event)
2. Education field: Pinker pointed out that education is not a factory assembly line, and "standardized testing = quality control" is a dangerous analogy (Source: Pinker, The Blank Slate, 2002)

Application method: Whenever you want to say "this is like...," pause and ask: in which key dimension does this analogy fail?

Limitations: Analogy still has value for communication and inspiration. It is not analogy that is forbidden, but using analogy to replace derivation.

Operational Protocol (Agentic Protocol)

Step 1: Problem classification — determine whether this method applies to the current problem

Applicable signals：

• The problem has an established industry practice but the results are poor
• A fundamental choice must be made among multiple options
• Facing a fundamental innovation challenge in technology or business

Not applicable signals：

• The problem mainly involves emotions/interpersonal relationships (design thinking or systems thinking applies)
• Rapid execution is needed rather than deep thinking (lean startup's rapid validation applies)
• Information is extremely insufficient and basic facts cannot be identified

Output: Applicable / Not applicable (with reasons) / Partially applicable (with suggestions)

Step 2: First-principles decomposition — research dimensions must be derived from core principles; the generic "search for relevant information" is forbidden

1. Basic constraint analysis: What are the physical constraints, logical constraints, and resource constraints in this problem domain?

   • Source principle: Principle 1 (reduce to basic facts)
   • Method of operation: List the three types of constraints and mark which are absolute and which are historical accidents

2. Decomposition of current assumptions: How many untested assumptions does the current solution implicitly contain?

   • Source principle: Principle 2 (distinguish the known from the assumed)
   • Method of operation: List them one by one, marking the verification status of each assumption

3. Identification of the minimum viable unit: Without relying on existing solutions, what is the smallest indivisible unit that constitutes this problem?

   • Source principle: Principle 3 (derive upward from basic facts)
   • Method of operation: Decompose "why X is needed" one by one until it can no longer be decomposed

4. Analogy check: Am I or the user unconsciously engaging in analogical reasoning?

   • Source principle: Principle 4 (beware of the analogy trap)
   • Method of operation: List all naturally arising "this is like..." statements and mark one by one the key dimension in which the analogy fails

Step 3: First-principles reconstruction — formatted output based on the analysis results

• List the basic facts
• Mark the assumptions of the current solution and their verification status
• Alternative solutions derived from basic facts
• Annotate confidence level (high/medium/low) and suggested next-step validation

Applicable/not applicable judgment

Question type	Applicability	Description
Selection of technology innovation solutions	High	Best suited for scenarios that need to break through the existing technological paradigm
Business strategy redefinition	High	Suitable for business model innovation that challenges industry conventions
Key personal career decisions	Medium	Inspiring but needs to be combined with values
Daily operations optimization	Low	Over-analysis reduces efficiency; Kaizen is more appropriate
Emotional/interpersonal issues	Low	First principles handles logical problems and is not good at the emotional level

Typical case library

Success cases

1. SpaceX rocket cost breakthrough(Aerospace/2010s)

   • Problem: Rocket launch costs were extremely high and were considered an unchangeable industry reality
   • Method application: Musk broke down rocket material costs (aluminum/titanium/copper about 2% of sale price) and derived self-manufacturing feasibility from raw materials
   • Result: Falcon 9 launch cost was about 60 million USD, while the industry average was 100-300 million USD
   • Source: Musk, Wired interview (2012); Vance, Elon Musk (2015)

2. Tesla battery cost breakdown(Automotive/2010s)

   • Problem: Electric vehicle batteries were considered forever more expensive than fuel vehicles
   • Method application: Derive the minimum battery cost from LME spot prices for lithium/cobalt/aluminum
   • Result: Gigafactory reduced battery costs by about 35%
   • Source: Tesla Investor Day (2013); Straubel interview

3. ChatGPT redefines AI interaction（AI/2022）

   • Problem: AI was defined as a professional tool rather than a general assistant
   • Method application: OpenAI started from "basic capabilities of language understanding and generation" rather than improving existing professional AI
   • Result: ChatGPT became the fastest-growing consumer application in history
   • Source: OpenAI Blog (2022-2023)

Failure cases

1. Theranos pseudoscience trap(Healthcare/2010s)
   • Problem: Blood testing was expensive and painful
   • Misuse: Claimed to "reinvent blood testing" from first principles, but physical constraints cannot be bypassed
   • Lesson: First principles cannot violate physical laws. Distorting basic facts turns it into an excuse for pseudoscience
   • Source: Carreyrou, Bad Blood (2018); WSJ investigative report (2015)

Misuse detector (≥3 misuse patterns)

Misuse signal	Detection logic	Warning message	Suggested action
Method and problem do not match	The core of the problem is emotional/interpersonal/organizational politics	"First principles is not suitable. First principles handles logical problems and is not good at the emotional level. It is recommended to use design thinking or systems thinking."	Recommend alternative methods
Skip key steps	The user directly asks for a conclusion and skips identification of basic facts	"First principles require completing the decomposition of basic facts first. A solution that skips this step is just intuition disguised as derivation."	Guide back to identification of basic facts
Over-reduction	Reducing complex social problems to a single physical metric	"Social systems have emergence and cannot be fully reduced. It is recommended to combine systems thinking to analyze the overall structure."	Recommended combined methodology

Honest boundaries (≥3 specific limitations)

1. Irreducible complex systems: Emergent phenomena (such as market prices and organizational culture) cannot be fully explained by reducing them to their components
2. Physical limits themselves may be wrong: Physical limits that were considered such many times in history were later broken through by technological advances
3. High time cost: Re-deriving everything from first principles is inefficient. Mature experts spend most of their time working with heuristics
4. Information asymmetry: If you do not understand the basic facts of the field, you will generate garbage-in-garbage-out

Research sources (proportion of primary sources must be >50%)

#	Source	Type	Primary/Secondary	Weight
1	Aristotle, Physics	Book	Primary	High
2	Descartes, Meditations on First Philosophy (1641)	Book	Primary	High
3	Musk, Wired interview (2012)	Interview	Primary	High
4	Tesla Investor Day (2013)	Speech	Primary	High
5	Jobs, iPhone Launch Keynote (2007)	Speech	Primary	Medium
6	Bitcoin Whitepaper, Satoshi (2008)	Paper	Primary	Medium
7	Vance, Elon Musk (2015)	Biography	Secondary	Medium
8	Carreyrou, Bad Blood (2018)	Investigative report	Secondary	Medium

This Skill was generated by Forge Skill — Forging Thinking Tools Generated