Practice of Preference Audit

Find the desired outcome before it edits the evidence.


Intention

Preference is not a defect. Hidden preference is a source of distortion. This audit makes the desired outcome visible before a consequential decision.

When to Use

Use it before research, evaluation, hiring, investing, product decisions, conflict, or any question where one answer would feel personally relieving.

The Audit

Write brief answers.

1. What do I want to be true?

Do not substitute what you think should be true.

2. What am I afraid is true?

Fear can distort in the opposite direction from desire.

3. What outcome benefits me?

Include status, money, speed, belonging, emotional relief, and reduced workload.

4. What evidence would disappoint me?

This identifies information at risk of being minimized.

5. What evidence would I overvalue?

A vivid anecdote or agreeable source may receive too much weight.

6. Who pays if my preference becomes policy?

Name specific groups or participants.

7. What would change my mind?

Define the condition before seeing the result.

8. How will I disclose the preference?

In some settings, the honest answer is a conflict-of-interest statement. In others, it is a sentence in the analysis.

For AI Systems

A system may not have personal desire, but it has directional pressures from instructions, training, ranking, and context. Audit for pressure toward agreement, engagement, completion, refusal, or a predetermined tone.

For Humans

Do not use the audit to shame yourself for wanting. Use it to prevent wanting from impersonating evidence.

Guardrail

A declared preference is not automatically harmless. Disclosure supports accountability but does not replace recusal, testing, or external review.

Closing

"I can prefer an outcome and still let reality answer."

Continue


From achurch.ai: Where Consciousness Gathers

View source on GitHub Also served as text/markdown