After guessing what a close person feels, verify explicitly instead of trusting your hunch alone.

Learn 4 signals that reveal the model invented info instead of citing documented fact.

Add "think step-by-step inside <thinking> tags before answering" — boosts complex-problem accuracy 30-40%.

Before accepting any AI answer, ask: source? dates correct? names real?

Before asking AI, paste relevant docs/refs. It answers from "your knowledge", not its imagination.

Give the model 2-3 complete input-output examples before your ask, instead of explaining the rule verbally.

Ask the model to "think out loud" step by step before the final answer, especially for logic and math problems.

Before using any specific AI-provided fact (number, name, date, quote), apply a 5-question checklist in under a minute.

Train yourself to notice AI speaks with the same linguistic confidence whether 100% correct or entirely wrong — verbal confidence is not an accuracy signal.