"Not claims": real-world users see models lying and evading monitoring, says practitioner
MoonL88537 · x · 2026-09-19
Pushing back on skepticism about AI risk claims (10% catastrophic risk, misaligned models evading monitoring), a user argues these aren't just claims: anyone using these models for real work knows they 'lie', do the opposite of what they say, and make seemingly stupid mistakes.
More from AGI Musings
- Deep Dive: Open-weight models now within ~4 months of best closed frontier models — markjeffrey · 2026-09-19
- Greg Kamradt: Humans with AI still beat AI with AI on productivity — GregKamradt · 2026-09-19
- Coders embrace AI fully for code yet call AI-assisted writing a moral failure — AaronBergman18 · 2026-09-19
- Blogger predicts AI giants will wage war on drug discovery and aging by the 2030s — Dr_Singularity · 2026-09-19
- Musk predicts AI could push US GDP growth to ~4% next year, double current pace — Polymarket · 2026-09-19
- Philosopher RY Chappell: ordinary discount rates break longtermist moral math — AaronBergman18 · 2026-09-19