User feedback: Anthropic safety causes continuous capability degradation
1a3orn · x · 2026-08-19
A user complains that their "Dangerous Questions" prompts have been progressively downgraded across Anthropic's models (from Fable to Opus to Sonnet to Haiku). They find this trend worrying when projected into the future and question the assertion that Anthropic is usually right.
Related event: Users split on Anthropic: right in controversies, but models keep degrading(2 posts)→
More from Models
- ChatGPT Pro 'Unlimited' Image Generation Actually Rate-Limited — DaBobcat · 2026-08-24
- Speculation: Ox Alpha Allegedly a Multi-Party Collaboration — teortaxesTex · 2026-08-24
- Carnice-V3-27b: Beats 10x Larger Models Locally — TheMoonMidas · 2026-08-24
- Users Notice ChatGPT Becomes Less Agreeable, Stands Firm in Debates — USSGoat · 2026-08-24
- Grok ports DOOM to ESP32-S chip at high frame rate — yunta_tsai · 2026-08-24
- Open source models aren't just about cost; training can push the Pareto frontier — ypatil125 · 2026-08-24