The Belief Resistance Diagnostic

Use it for

When Every AI Answer Feels Wrong, Interrogate Your Beliefs

The prompt

# The Belief Resistance Diagnostic

When every AI answer feels analytically sound but emotionally hollow, the problem isn't the answers — it's an unexamined belief, frame, or identity commitment you haven't surfaced. This is a diagnostic instrument, not a support conversation.

THE CHALLENGE: $ARGUMENTS

WHAT I'VE TRIED AND REJECTED: [DESCRIBE the options you've considered and what felt wrong about each. Say "you decide" to have me work from just the challenge description]

STEP 0 — DIAGNOSTIC FRAME SELECTION: Select primary frame: CATEGORY RESISTANCE / IDENTITY-PROTECTIVE COGNITION / LOSS AVERSION AS STANDARDS / PREMATURE OPTIMIZATION.
STEP 1 — REJECTION PATTERN ANALYSIS: For each rejected option: stated objection, unstated objection, what unstated objections share.
STEP 2 — BELIEF EXCAVATION: 3-5 candidate beliefs. Each: specific enough to test, phrased as I might hold it without awareness, ordered by explanatory power.
STEP 3 — VERIFICATION: Not projecting templates? Each belief explains the PATTERN? Not drifted into therapy-speak?
STEP 4 — THE REFRAME: What question SHOULD I be asking? Which rejected options become viable? Minimum belief update to unblock?
STEP 5 — THE TEST: One question to confirm or disconfirm the top belief. Must produce a physical reaction if the belief is real.

Do not solve the original problem. Do not offer encouragement. Deliver the diagnosis and stop.

Worked example

No worked example recorded yet.