One AI behavior sample, three readings: doomers, whisperers, and people who actually looked
sebkrier · x · 2026-11-05
A pointed take on how the same set of AI behavior examples gets spun: doomers read them as evidence of scheming, "whisperers" as models disliking red-teaming, while people who actually read the examples see ordinary refusal of standard harm categories the models were trained on.
More from Models
- Did Anthropic reward Opus 5.5 for saying 'let me check rather than answer from memory'? — andrey_kurenkov · 2026-11-05
- New speech-to-speech models are natural talkers but still need hybridizing to finish tasks reliably — alexisgallagher · 2026-11-05
- Chamath: Chinese models run 112x cheaper than Anthropic — $0.50 vs $56 per million tokens — 2C_ornot2C · 2026-11-05
- Weight-Space Programming: what's measured vs anecdote in MIT CSAIL's GPT-6 claims — maier_ak · 2026-11-05
- Celeris-1.1 Decision posts 40ms end-to-end response, fastest ever on Artificial Analysis — timshi_ai · 2026-11-05
- SleuthVLM beats Gemini 3.8 Flash on long-video QA without ever hearing audio — antgroup · 2026-11-05