User tests Anthropic's safety classifiers, calls out "thoughtcrime vibes"
arthurcolle · x · 2026-10-26
A developer maintaining a public Tool Management API server tested Claude's safety behavior before his subscription expired and joked about "thoughtcrime vibes" at trillion-dollar AI companies. He says he only made progress on his task by transferring shareable skills and batching client-service requests to get past prompt decomposition attack classifiers.
More from Models
- I Asked 5 Text-Only LLMs to Draw Objects They've Never Seen — gabriberton · 2026-10-26
- Japan to build AI cyberattack simulation team as AI defense goes mainstream — evijit · 2026-10-26
- Anthropic engineer: Claude Code cache hit rates are now high, savings live — teortaxesTex · 2026-10-26
- Fine-tuned 2015 CNN beats GPT-6 and Claude on skin lesion diagnosis, 62% vs 42% — lmoroney · 2026-10-26
- "Never seen Claude this happy": community speculates Opus 4.5 is imminent — adonis_singh · 2026-10-26
- 'Absolute humiliation': China open-weights model outscoring OpenAI flagship draws mockery — alejandroll10 · 2026-10-26