Observed arrival · 2026-10-16
Be Nice Bot puts AI rudeness on a session tape
Credibility concern recorded. The source reference remains available for verification and correction.
A browser-based, open-source tone desk labels messages to AI assistants, tracks patterns across a conversation, and rewrites rude wording while preserving the request.
- For
- AI assistant builders and policy researchers
- Worth noticing
- The page reports zero false alarms among 43 allowed messages and identifies three needless messages within six turns as its sustained-cruelty threshold.
Field notes
The visible desk offers more than a single-message filter: users can lay turns onto a session tape, adjust the conversation window and sustain threshold, and inspect a calibration view. Its interactive exercise asks visitors to sort ten messages, while the liner notes identify benice 0.1.0 as MIT-licensed and describe a held-out set written before the rules. The page says inputs stay in the browser; that privacy claim is presented by the project, not independently verified here.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue