{
  "schema": "vedokrok.public-item.v1",
  "release_id": "MHC-RPUB-20260920-75ad787a",
  "url": "/knowledge/use-verbal-uncertainty-as-a-brake-not-a-truth-signal",
  "id": "MHC-D-RESEARCH-0504",
  "version": "0.1.0",
  "title": "Use verbal uncertainty as a brake, not a truth signal",
  "summary": "Hedging can change behavior without being a calibrated probability.",
  "kind": "principle",
  "body": "Treat uncertainty wording as an interface intervention. It may appropriately slow acceptance, but the phrase itself does not prove the model is uncertain for the right cases. Test whether hedging appears preferentially on errors and whether it improves decisions without causing excessive rejection of correct advice.",
  "limits": [
    "Natural-language uncertainty can be useful even when not numerically calibrated; just do not overclaim what it represents."
  ],
  "topics": [
    "union-ai-reliance-metacognition"
  ],
  "intents": [],
  "source_ids": [
    "RS-2A63149D9808BDF0"
  ],
  "evidence": [
    {
      "claim": "A large preregistered experiment found natural-language uncertainty expressions reduced agreement with an LLM and increased user accuracy in the tested question-answering setting.",
      "source_id": "RS-2A63149D9808BDF0",
      "role": "supports",
      "note": "Uncertainty wording can also reduce useful reliance on correct answers; calibration matters.",
      "locator": "Abstract"
    }
  ],
  "use_when": [
    "The model says 'I'm not sure' and users treat the phrase as proof the system is well calibrated."
  ],
  "avoid_when": [
    "Natural-language uncertainty can be useful even when not numerically calibrated; just do not overclaim what it represents."
  ],
  "example": "'I may be wrong' can prompt a source check, but it should not be counted as calibrated metacognition without validation.",
  "check": "The system distinguishes behavioral effect of hedging from evidence that the uncertainty statement itself is accurate.",
  "sources": [
    {
      "id": "RS-2A63149D9808BDF0",
      "title": "I'm Not Sure, But...: Examining the Impact of Large Language Models' Uncertainty Expression on User Reliance and Trust",
      "url": "https://doi.org/10.1145/3630106.3658941"
    }
  ],
  "relations": [
    {
      "from": "MHC-D-RESEARCH-0504",
      "to": "MHC-D-RESEARCH-0503",
      "type": "useful_with",
      "url": "/knowledge/test-uncertainty-cues-on-behavior-not-only-user-ratings"
    }
  ],
  "collections": [
    {
      "id": "RC-39B535F35706B375",
      "title": "Calibrate when to rely on AI instead of measuring trust as a feeling",
      "url": "/collections/calibrate-when-to-rely-on-ai-instead-of-measuring-trust-as-a-feeling"
    }
  ]
}