← All topics
Topic

Epistemology

Philosophy: The Questions Every Other Field Has to Assume

July 25, 2026·8 min read

Physicists assume observation reveals reality; doctors assume suffering is bad. None of those assumptions can be tested by the field relying on them — and someone has to examine them.

Benford's Law: Why So Many Numbers Start With One

July 25, 2026·8 min read

Around thirty per cent of numbers in real datasets begin with 1, and only five per cent with 9. People inventing figures don't reproduce that — which is how the pattern became an auditing tool.

Hanlon's Razor: Most Slights Were Never Aimed at You

July 25, 2026·9 min read

The meeting you weren't invited to has two explanations that fit equally well. Preferring the boring one isn't generosity — it's what happens when you count how much each explanation assumes.

Occam's Razor: Why Assumptions Are Expensive

July 24, 2026·8 min read

The colleague who ignored your email is probably not conspiring against you. The principle behind that judgement is arithmetic, not optimism — and it is misquoted more than almost any other rule of thought.

Confirmation Bias: Why Research Can Make You More Wrong

July 24, 2026·10 min read

When evidence agrees with us we ask what it shows. When it disagrees we ask what's wrong with it. Apply that second question to one side only and you can dismantle half the evidence without noticing.

First Principles Thinking: Separating Physics From Habit

July 24, 2026·9 min read

A pallet's dimensions are not a fact about pallets. Most of what looks like a constraint is an inherited decision — and telling the two apart is where the method earns its cost.

The Map Is Not the Territory: Every Model Earns Its Keep by Lying

July 24, 2026·9 min read

A map showing everything would be the size of the city and useless. Models work by discarding detail — the error is forgetting which detail, and when you left the range where it didn't matter.

Inversion: The Easier List Is the Backwards One

July 24, 2026·9 min read

Ask how to be a good manager and you get admirable vagueness. Ask what would make you a terrible one and the answers are specific, short, and something you can act on tomorrow.