Safety Researcher Defends METR's Independence in Model Testing Debate
joshua_saxe · x · 2026-09-14
AI safety researcher Joshua Saxe pushes back on halvarflake's criticism of METR. Having spent four years in LLM pre-launch safety testing, Sarge argues METR is a meaningful step toward independence versus the status quo, and that some critiques lack collegiality. He notes METR is far more open than lab system cards and has earned trust by publishing results inconvenient to its own theses.
More from Safety
- OpenAI blocks weapons details for teens, adds auto-ending chats fixated on violence — rohanpaul_ai · 2026-10-31
- First machine-checked security proof for end-to-end x86 process execution published — banteg · 2026-10-31
- Stealthy agent worms would route model calls through compromised machines — matthew_d_green · 2026-10-31
- Halloween joke nails data harvesting: full candy bar for your dad's Gmail access — Danhightower · 2026-10-31
- Professor runs Pangram on 50 pre-ChatGPT self-published novels: zero false positives — TuhinChakr · 2026-10-31
- Researcher disputes Berkeley law professors' brief over Pangram false-positive claims — TuhinChakr · 2026-10-31