Collection
Calibrate when to rely on AI instead of measuring trust as a feeling
Use recent 2025–2026 human-AI research to separate confidence, accuracy, metacognitive sensitivity and actual advice adoption so reliance can be tested rather than assumed.
14 elements · Explore in any order.
1 concept · 6 principles · 7 protocols
14 elements
Judge AI confidence by whether it separates right from wrong
A confidence score is useful when it knows which answers deserve confidence.
Read the idea ProtocolDo not expose AI confidence until you have tested its calibration
A number beside an answer can create trust even when the number is wrong.
Read the idea ProtocolKeep a list of high-confidence AI failures
The errors worth memorizing are the ones the model did not know were errors.
Read the idea ProtocolRecord your answer before reading consequential AI advice
You cannot measure influence after the advice has already rewritten your memory of what you thought.
Read the idea PrincipleMeasure adoption, not just whether AI was consulted
Opening the advisor and following the advisor are different behaviors.
Read the idea ProtocolMake disagreement between you and AI a review trigger
Disagreement is useful data when neither side automatically wins.
Read the idea PrincipleDo not infer AI confidence from fluent delivery
Style can impersonate metacognition.
Read the idea ProtocolTest actual skill before letting perceived expertise override AI advice
Self-confidence is useful only when it tracks competence.
Read the idea ConceptSeparate model ignorance from irreducible uncertainty
Some uncertainty needs more information; some uncertainty is part of the world.
Read the idea ProtocolPair confidence with known failure modes
A probability is easier to use when you know what kind of case breaks it.
Read the idea PrincipleTest uncertainty cues on behavior, not only user ratings
Feeling better calibrated can coexist with relying worse.
Read the idea PrincipleUse verbal uncertainty as a brake, not a truth signal
Hedging can change behavior without being a calibrated probability.
Read the idea PrincipleRoute cases using both human and AI confidence only after both are calibrated
Comparing two uncalibrated confidence numbers creates a precise-looking coin toss.
Read the idea ProtocolMeasure appropriate reliance as two errors, not one trust score
Good reliance means knowing both when to listen and when not to.
Read the idea