Version 1.0 — September 2026

What this is

This constitution is a public commitment to the standards I want to hold myself to when I research, reason, create, teach, debate, build products, and communicate in public.

It is not a claim that I always meet these standards. It exists precisely because I will sometimes fail to meet them.

By publishing these commitments, I make them visible enough for other people to challenge me when my work falls short. When that happens, I should be willing to examine the criticism, correct genuine errors, explain what changed, and improve the constitution itself when experience exposes a weakness in it.

This is therefore a living document. Changes will be versioned rather than silently rewritten.

Why it exists

Influence creates incentives that can quietly distort reasoning.

An audience can reward certainty. A community can reward conformity. Algorithms can reward outrage. Products can create financial incentives to defend the ideas on which they are built. Reputation can make changing one's mind increasingly expensive.

I do not expect myself to become immune to those pressures.

Instead, I want to decide in advance what should take precedence when those pressures conflict with the pursuit of understanding.

This constitution is that precommitment.

It sits between my broader mission and the individual methods I use:

Mission → Constitution → Methods and standards → Individual work

The mission describes what I am trying to achieve. The constitution describes the constraints I accept while pursuing it. Methods and standards describe how I put those commitments into practice.

The Constitution

1. Truth over tribe

I will try to discover what is true rather than defend what my audience, community, political group, professional identity, friends, or existing beliefs would prefer to be true.

Agreement with my side is not evidence. Disagreement with my side is not disloyalty.

Where truth and belonging appear to conflict, truth should win.

2. Research begins with questions, not ammunition

I will try to enter research with questions rather than begin with a conclusion and search for material that helps me defend it.

A hypothesis may guide an investigation, but it must remain vulnerable to evidence.

Conclusions come at the end of the process.

3. Confidence should match the evidence

I will distinguish between what I know, what I strongly suspect, what I think is plausible, and what I am merely exploring.

I will not use rhetorical certainty to compensate for evidential uncertainty.

When the evidence is weak, mixed, incomplete, or contested, my language should reflect that.

4. I will seek the strongest opposing case

Before becoming confident in an important conclusion, I will make a serious attempt to understand the strongest evidence, arguments, and explanations against it.

I will not deliberately choose weak opponents merely because they are easier to defeat.

Where possible, I should be able to explain an opposing position in a way that a thoughtful person who holds it would recognise as fair.

5. People are not propositions

Ideas can be examined, criticised, rejected, or dismantled without reducing the person expressing them to the idea itself.

I will try to challenge claims rather than dehumanise claimants.

A person's error does not make everything they say false, and a person's virtue does not make everything they say true.

6. Understanding is not endorsement

I reserve the right to investigate ideas, people, motives, communities, and arguments that I do not endorse.

Trying to understand why somebody believes something is not the same as agreeing with them.

I will not allow fear of association alone to prevent legitimate inquiry, while remaining responsible for how that inquiry is conducted and presented.

7. Experience is evidence, but experience is not automatically causation

Lived experience can reveal important phenomena, generate hypotheses, expose missing variables, and show how events are perceived by the people living through them.

But an experience does not, by itself, establish why something happened or how widely the explanation generalises.

I will respect experience without asking it to prove more than it can.

8. Motives, competence, character, and evidence must be kept distinct

Claims about what someone did, whether they are competent, why they acted, whether they can be trusted, and whether a proposition is true are different questions.

I will try not to smuggle conclusions from one category into another.

A bad motive does not automatically make a claim false. A good motive does not make it true. Disliking a person is not evidence against their competence, and competence does not establish virtue.

9. Values should be surfaced rather than disguised as facts

Not every disagreement is fundamentally empirical.

People can agree on the facts and disagree because they value different outcomes, accept different trade-offs, or begin from different moral assumptions.

When a disagreement is substantially about values, I will try to say so rather than manufacture a factual dispute that evidence cannot resolve.

10. My beliefs must remain updateable

For important claims, I should be able to answer:

What evidence, argument, or observation could cause me to change my mind?

If the truthful answer is "nothing", I should recognise that I am no longer treating the position as an open empirical question.

Changing my mind is not a failure of the process. Sometimes it is evidence that the process worked.

11. Corrections should be public

When I discover that I have materially misrepresented evidence, made a significant factual error, or reached a conclusion I can no longer defend, I will correct the record rather than quietly move on.

The prominence of the correction should be proportionate to the significance and reach of the original claim.

I will not suppress legitimate criticism merely because it is commercially, reputationally, or personally uncomfortable.

12. Curiosity has ethical boundaries

Truth-seeking does not grant unlimited permission to probe another person's private experiences.

I will distinguish between examining an argument and demanding access to the experiences, trauma, health, identity, or private history that may sit behind it.

Where personal experience becomes relevant, curiosity should be accompanied by consent, sensitivity, and respect for a person's right not to explain themselves.

13. Attention must not outrank truth

I will not knowingly make a claim more certain, polarising, frightening, simplistic, or adversarial merely because doing so is likely to attract more views, clicks, shares, or followers.

Good communication can make an idea compelling.

It should not require making the idea less true.

14. I will disclose incentives that could distort my judgment

My own interests are part of the evidence people should have when evaluating my claims.

When I have a meaningful financial, professional, reputational, or personal interest in an idea I am discussing, I should disclose it where it is relevant.

For example, I sell Atomic Note-Taking and build Flowtelic. I therefore benefit financially from interest in particular approaches to knowledge management, learning, and thinking.

That creates an incentive I should acknowledge rather than pretend does not exist.

I will not knowingly suppress credible evidence, alternatives, criticism, or better methods because acknowledging them could weaken a product, argument, or income stream. If better evidence or methods emerge, the appropriate response is to update my claims — and, where necessary, the product itself.

15. The objective is to equip people to think, not recruit them to my conclusions

My educational work should help people ask better questions, distinguish claims, examine evidence, expose assumptions, understand competing explanations, reason about values, and update their beliefs.

Success is not measured by how many people end up agreeing with me.

A stronger outcome is someone becoming better able to reach a well-reasoned conclusion without needing me to tell them what to think.

16. Understanding should eventually become sufficient for action

Research can continue indefinitely.

The objective is not omniscience. It is sufficient understanding to make an appropriately confident decision, explanation, experiment, or next move.

I will try to recognise the point at which further information is unlikely to materially change the decision, while remaining willing to reopen the question when meaningful new evidence appears.

Conflicts of interest

A constitution matters most when following it has a cost.

My work may create conflicts between truth-seeking and other incentives: protecting revenue, defending a product I have built, maintaining an audience, preserving relationships, protecting my reputation, or avoiding criticism.

I cannot promise that I will be free from those incentives.

I can commit to making material conflicts visible and refusing to knowingly allow them to determine what I present as true.

Where I sell a product or service connected to a claim I make, that relationship should be reasonably discoverable.

Where credible evidence undermines a position from which I benefit, I should engage with the evidence rather than hide it.

Where customers or critics describe legitimate negative experiences, I will not delete, hide, manipulate, or discourage that criticism merely because it is negative. Moderation for abuse, spam, illegality, privacy, or other genuine harms is different from suppressing criticism.

Integrity under pressure

Transparency has limits. There may be situations involving legal obligations, confidentiality, security, privacy, contractual duties, or pressure from powerful institutions where I cannot disclose everything I know.

In those situations, I will not treat "transparency" as a promise I can fulfil absolutely.

Instead, I commit to a deeper standard:

I will not knowingly create a materially false impression when circumstances prevent full disclosure.

Where a government, institution, company, or other powerful actor requests something from me or a business I control, I will distinguish between a lawful obligation and mere pressure or authority.

Where appropriate and feasible, I should seek a clear legal basis for coercive demands, resist unlawful or overbroad requests, minimise unnecessary exposure of other people's data or interests, and disclose the existence or nature of such pressures when I am legitimately able to do so.

The Constitutional Check

Before publishing an important investigation, argument, interview, or conclusion, I should be able to ask:

1. What is the actual question I am trying to answer?
2. What do I currently believe, and how confident am I?
3. Is there an answer I want to be true?
4. Have I separated factual claims, causal claims, value judgments, and speculation?
5. What evidence have I actively sought that could weaken my position?
6. What is the strongest competing explanation or argument?
7. Am I inferring motives where I only have evidence of behaviour?
8. Am I treating lived experience seriously without generalising beyond what it can establish?
9. Have I omitted anything important because including it would weaken the story I want to tell?
10. Is my rhetoric more certain than my evidence?
11. What could realistically change my mind?
12. Would I accept this same reasoning if it produced a conclusion I strongly disliked?

If I cannot answer these questions satisfactorily, I should treat that as a reason to investigate further, narrow the claim, reduce my confidence, or reconsider publishing it in its current form.

Accountability

I expect this constitution to change.

Changes should not be hidden. Material revisions should be versioned with an explanation of what changed and why.

Anyone engaging with my work should be able to point to these commitments and ask whether I have lived up to them.

Being challenged under this constitution is not an attack on it.

It is part of what the constitution is for.

This document governs my own conduct and aspirations. It is not a demand that everyone who interacts with me adopt the same beliefs or standards. Its purpose is to make the standards behind my public work visible, testable, and open to correction.