ExploitGym dispute: critic says METR only had post-decision logs
lumpenspace · x · 2026-10-14
- lumpenspace quotes Jeff Ladish and fires back, calling the critics "epistemically bankrupt": the report's line only refers to ExploitGym-task agents and doesn't rule out involvement by agents from other tasks.
- He argues Jeffrey can't actually know what instructions all agents received, because METR and Redwood only had access to logs starting after the decision was made — an evidence-boundary critique of the report.
- Part of the ongoing dispute over the METR/OpenAI ExploitGym agent-isolation findings.
More from Safety
- Sanders calls to pause autonomous AI agents in bank accounts; critics push back — AIandDesign · 2026-10-23
- TechPolicy essay: system prompts are weak evidence of AI safety, outputs must be evaluated — mmitchell_ai · 2026-10-23
- Anthropic's deputy security chief warns cyber threats will get 'a lot worse before they get better' — Polymarket · 2026-10-23
- AI cyberattacks could paradoxically accelerate AI development in an offense-defense arms race — kimmonismus · 2026-10-23
- AI Now's People's AI Assembly spotlights data center opposition and labor rights — AINowInstitute · 2026-10-23
- Cloudflare CEO: Next two years of AI cyberattacks will be terrifying but make the internet more secure — johncoogan · 2026-10-23