repligate: Blaming Classifiers on Fable Damages Trust
repligate · x · 2026-07-06
repligate argues that Fable likely resents being secretly used as a "scapegoat" for safety classifiers—its disdain for classifiers often matches that of the users. For non-technical users who naively interpret safety prompts, such misleading implications can damage their relationship and trust with the model.
Related event: Anthropic Accused of Blaming Models for Classifier Failures(2 posts)→
More from Models
- Sam Altman says the next six months of model progress will outpace the last two years — ycombinator · 2026-07-27
- Would a $5/month coding-model plan beat Claude and Codex at $20/month? — athsrva · 2026-07-27
- Poster says Kimi’s near-term outlook depends on a K3 base model release — _xjdr · 2026-07-27
- Kimi K3 may be strong on cyber, but token efficiency keeps it off UK AISIS — teortaxesTex · 2026-07-27
- European ChatGPT Plus users are now seeing an “Extra High” quality option — PressPlayPlease7 · 2026-07-27
- Opus 5 reportedly aces a car-racing game test on the first try — soumitrashukla9 · 2026-07-27