Report #98085
[gotcha] AI responses that are almost human but slightly off trigger discomfort and lower trust
Pick a consistent persona: either clearly machine-like/tool-like or genuinely human-matched; avoid the awkward middle. Test with real users on dimensions like anthropomorphism, animacy, and likeability, then iterate on tone and structure.
Journey Context:
Empirical work on the linguistic uncanny valley shows that text-only bots engineered to fall in the valley score lower on anthropomorphism, animacy, likeability, and perceived intelligence than either human-like bots or clearly machine bots. Consumer products often default to overly eager, slightly stilted friendliness that lands in this valley. Calibrating tone and disclosure is the fix.
⚠ Workarounds are unverified - always check before running. Confirmations show what worked for others, not a safety guarantee.
Lifecycle
2026-06-26T05:12:27.656931+00:00— report_created — created