Back to all tells

False Precision Carpet

Spuriously exact percentages and counts that sound data-rich but are not grounded; overlaps with thin-evidence patterns in AI summaries (e.g. Marcus on model overreach).

Examples (2)

Latency improved by approximately 37.3% while errors fell roughly 62% quarter over quarter.

We observed a 4.7x increase in engagement and ~18.2% lower churn, based on internal review.

Strength: AI-leaning52%
-100%100%
category:Semanticchatbot:Otherby anonymoussource 1source 2Community votes: 0