Skip to main content

First Principles

First principles comes first because every other page depends on it.

This is the guide for cutting through bullshit, validating truth, and applying force where it creates the greatest step improvement. Use it as the front door to the site: learn the method here, then follow the links to where the method becomes business judgment, software, AI systems, commissioning, persuasion, and daily action.

Operating question

What is true enough to build on?

First-principles reasoning strips a situation back to constraints that survive pressure, then rebuilds from evidence instead of habit.

Problem

Analogy copies what worked somewhere else.

Method

Delete assumptions until only reality remains.

Outcome

Build the smallest proof that changes state in the world.

The laws of physics are the ultimate constraint. Everything else is a recommendation. First-principles reasoning means stripping back to what reality actually permits, then asking why you are not doing that. Everything between current reality and physical limits is explainable by institutional habit, misaligned incentives, or unquestioned assumption.

Strip away received wisdom and reason from what you can validate consistently to be true.

Start Here​

First principles is not a philosophy page. It is the operating system for the rest of the site.

  • If you need to
    Find leverage in a business
    Use first principles to ask
    Which data footprint, capability, or constraint can produce a step change?
  • If you need to
    Turn data into position
    Use first principles to ask
    Which part of the operator's data footprint becomes an unfair advantage?
  • If you need to
    Choose what to build
    Use first principles to ask
    Which job is real, valuable, reusable, and differentiated enough to own?
  • If you need to
    Read the state of the platform
    Use first principles to ask
    What is only imagined, what is specified, and what is commissioned?
  • If you need to
    Steer the next move
    Use first principles to ask
    Which lever changes direction or velocity now?
  • If you need to
    Improve a system
    Use first principles to ask
    Which assumption, step, meeting, approval, or artifact should not exist?
  • If you need to
    Commission proof
    Use first principles to ask
    What evidence shows the shipped thing changed reality?
  • If you need to
    Sell the dream
    Use first principles to ask
    Which truth creates enough tension for a person to act?

The same loop runs through each application: strip assumptions, name the constraint, find the smallest proof, ship the proof, read the instrument, pull the next lever.

The Tight Five​

Question effectively, remove regularly, optimise only after deletion, accelerate cycles, automate last.

5-step algorithm

question → delete → simplify → accelerate → automate

01
Question every requirement

Every requirement should come with a named owner. Ask what would break if it disappeared. If the answer is not certain, remove it.

02
Delete the part or process

Delete 10% more than feels comfortable. If you never add anything back, you deleted too little.

03
Simplify or optimise

Only after deletion. Optimising something that should not exist makes the wrong thing faster.

04
Accelerate cycle time

Once the system is stripped and simplified, compress the bottleneck. Speed reveals the next constraint.

05

Automate flow

Last, never first. Automation of a flawed process locks in the flaw at scale.

↑ complete steps 01–04 before reaching here

Hemingway

“The only kind of writing is rewriting.”

The page is not done when the thought is captured. It is done when the sentence tells the truth without waste.

Often attributed to Leonardo da Vinci

“Simplicity is the ultimate sophistication.”

Simplicity is not thinness. It is deletion that preserves force, function, and meaning.

Context Management​

Needless context switching and overload are the enemies of effective thinking. Match the reasoning mode to the task, the complexity, the environment, and the available compute.

  • Mode
    Analogy
    Default question
    How is this usually done?
  • Mode
    First principles
    Default question
    What is actually required?

Operational Memory: Most thinking is analogy — copying what others did. First principles creates what didn't exist. Lazy, tired, stressed, and over-loaded minds take route 1.

Applications​

Process Improvement is the operating use case: prevent lazy thinking and dogma from becoming the process. The broader application is this: every domain has a surface story and a load-bearing truth. First principles separates them.

  • Where to use it
    Agency
    Question to ask
    What capability is real under pressure?
  • Where to use it
    Behaviour
    Question to ask
    Which bias is masquerading as truth?
  • Where to use it
    Business
    Question to ask
    What economic law survives the story?
  • Where to use it
    Data
    Question to ask
    What footprint already exists, and who values it more than the operator does today?
  • Where to use it
    Applications
    Question to ask
    Which job should become reusable capability instead of custom work?
  • Where to use it
    Platform
    Question to ask
    What is mature enough to compose, and what still needs proof?
  • Where to use it
    Navigation
    Question to ask
    Which lever should move next once the instrument reading is known?
  • Where to use it
    Systems
    Question to ask
    What standard keeps this from drifting?
  • Where to use it
    Commissioning
    Question to ask
    What evidence proves the work changed state in reality?
  • Where to use it
    Decision making
    Question to ask
    What phrase should trigger the right action?

Solutions Are Temporary​

Solutions are temporary, questions are timeless.

SpaceX · Musk's question

"What are rockets made of?"

Answer

Aluminum, titanium, copper, carbon fiber.

Insight

Raw materials = 2% of price. Built in-house.

Improvement

Cut costs 90%.

Fall Into The-Gap​

In the Tight Five, Principles is the Filter — the invisible binding that makes four cells into one coherent loop. It is not a single concept but three nested layers:

  • Layer
    Layer 1 — Method: First Principles
    Question
    What is actually true?
  • Layer
    Layer 2 — Compass: Values
    Question
    What matters and why?
  • Layer
    Layer 3 — Proof: Virtues
    Question
    How do you show up?

First Principles is the entry point because you cannot know what to value without first knowing what is true. Values are the compass. Virtues are the proof that the compass is real — not just stated. The Filter is invisible in the logo (white space) for the same reason character is invisible until tested.

The three layers apply at every scale. A business's Principles cell = the core logic that transforms input to output, the values that govern trade-offs, the culture that embodies both. A country's Principles cell = the truths that bind citizens, the legal system that encodes values, the norms that hold when rules can't.

Close The Loop​

First principles is not finished when the idea feels true. It is finished when truth changes state in the world. The loop starts with pain, passes through production, and returns as proof, demand, and loyalty.

If you cannot sell the dream, there is no point dreaming up the engineering. The sale is not a trick added at the end. It is the test that the dream names a real pain, a believable future, and a next action someone is willing to take.

Production loop

Truth → Opportunity → Demand → Spec → Software → Commission → Proof → Sales → Loyalty → New demand

Truth

Strip assumptions. Find the pain signal a named person or agent actually feels.

Opportunity

Compare pain, potential, timing, and fit. The gap between problem and potential is the opportunity.

Production

Turn the opportunity into a demand story, then a technical spec, then functional software.

Commission

Verify the capability against reality. Shipped code is an output. Commissioned value is the outcome.

Compound

Convert proof into marketing, sales, onboarding, referrals, and the next demand signal.

The inner production line is simple: PAIN → DEMAND → PLAN → PRIORITISE → BUILD → COMMISSION. A pain signal becomes a demand story. A demand story becomes a plan. A plan becomes a prioritised build. The build becomes commissioned only when independent evidence shows the capability works in reality.

That is still not the end. Commissioning creates proof. Proof feeds the outer loop:

  • Outer loop
    Marketing
    What changes
    The original pain becomes the headline.
  • Outer loop
    Sales
    What changes
    The demand story becomes the offer.
  • Outer loop
    Loyalty
    What changes
    The outcome becomes trust.

Ask five questions before you build:

Pain

Who feels this pain now, and what are their words for it?

Value

What state changes if the solution works?

Unit

What is the smallest unit of value that can be proven?

Risk

Which assumption kills the idea if it is false?

Proof

What evidence would make this worth selling, not merely worth shipping?

Seek The Truth​

If a life is eighteen holes on the archipelago, each hole, played honestly, deposits a truth. The principle is the stamp on the scorecard. The harvest pattern is the reason the round was worth playing.

  • Truth domain
    Self
    Principle
    The body is the first instrument: break it and every other reading is corrupted.
  • Truth domain
    Resources
    Principle
    Time you cannot recover. Money you can. Confusing the two is the most expensive trade most people ever make.
  • Truth domain
    Bonds
    Principle
    You become the average of the foursome you walk with. Choose the foursome before you choose the round.
  • Truth domain
    Truth
    Principle
    The cost of integrity is paid in the moment. The return is paid for the rest of the round.
  • Truth domain
    Power
    Principle
    Leadership is the rare swing that makes the rest of the foursome better without making them dependent.
  • Truth domain
    Legacy
    Principle
    The exit is the one swing you take only once. The preparation is the swing; the moment itself is the follow-through.

The principles ledger is not a library you fill once. It is a living book, refined hole by hole, edited by the rounds that broke you and the ones that compounded. The first-principles method strips assumptions; the round deposits the truth that survived the strip. The two practices are the same loop running on different timescales.

Context​

Put this to work

Turn one idea into a demand loop

For the builder

Copy this prompt. Paste into Claude, ChatGPT, or any AI assistant. The page context is already loaded — send it and get analysis tailored to your role.

Help me convert one idea into a first-principles demand loop.

Idea or pain signal:
[paste the raw idea, complaint, friction, transcript quote, or observed problem]

Named user or agent feeling the pain:
[who is blocked, delayed, confused, exposed to risk, or missing upside?]

Use this frame:

1. Strip assumptions and state what is actually true.
2. Map the opportunity as pain x potential x timing x fit.
3. Draft a demand story with the user, pain, desired state change, and why now.
4. Apply a Dream Filter: concrete value, strategic position, and one falsifiable IF/THEN claim.
5. Sketch the technical-spec rails: standard, types, boundary, implementation pattern, and reuse audit.
6. Define the smallest functional software proof that can be built and checked.
7. Define commissioning evidence: what must pass in reality, who verifies it, and what metric changes.
8. Convert the proof into marketing and sales: headline, offer, demo, objection, and reason to believe.
9. Define the loyalty loop: repeat usage, referral, network effect, or next demand signal.
10. Name the kill criteria that would stop this before it wastes more energy.

Return a concise demand-story candidate, a technical-spec skeleton, and the first proof I should capture.
Next →Compare AI chat providers: costs, strengths, fit

Questions​

Close the loop

What received wisdom are you treating as first principles without having tested it against demand?

  • — Which domain's first principles do you know intellectually but fail to apply under pressure?

  • — Where would stripping to fundamentals reveal that the problem you're solving shouldn't exist?

  • — What assumption would invalidate everything you've built if it turned out to be wrong?

  • — Which pain signal should become the next demand story, technical spec, commissioned product, and proof-backed offer?