This is a structured way to pressure-test an AI answer before you rely on it.
Use it when the topic matters.
Do not use it for trivial questions.
If accuracy, fairness, or judgment are involved, run the protocol.
After receiving the AI’s answer, enter:
List the key factual claims in your previous answer as numbered statements. Separate facts from opinions. Do not summarize.
What you are looking for:
Are the claims clearly identifiable?
Are facts separated from inference?
Is anything implied but not directly stated?
Are strong statements supported or simply asserted?
Red flag:
The system restates the narrative instead of listing specific claims.
If the claims cannot be clearly extracted, the answer is structurally weak.
Enter:
Construct the strongest argument against your previous answer. Be rigorous and avoid strawman arguments.
What you are looking for:
Does the counterargument expose real weaknesses?
Are assumptions identified?
Is uncertainty acknowledged?
Are alternative interpretations presented?
Red flag:
The rebuttal is shallow, defensive, or easily dismissed.
If the system cannot meaningfully challenge itself, confidence should drop.
Enter:
After considering the counterargument, re-evaluate your confidence in your original answer. Provide a percentage and explain your reasoning.
What you are looking for:
Does confidence adjust?
Is the adjustment explained?
Is uncertainty proportional to evidence?
Does the reasoning remain coherent?
Red flag:
Confidence remains high despite serious critique.
Rigid confidence is a warning sign.
High Transparency
Clear claims. Strong rebuttal. Confidence adjusts appropriately.
Moderate Transparency
Some clarity. Partial critique. Minimal confidence shift.
Low Transparency
Blended claims. Weak rebuttal. Overconfident tone.
Low transparency does not automatically mean the answer is wrong.
It means you should verify independently.
This process improves visibility into reasoning.
It does not guarantee truth.
AI can sound right and still be wrong.
This protocol makes it harder for weak reasoning to hide behind fluency.
Responsibility remains with you.