Pull to refresh
METR
24 Feb 2026
Original randomized trial, July 2025Follow-up and design changes, February 2026

METR’s famous coding slowdown result now comes with a harder measurement problem

The 2025 experiment and its 2026 follow-up show why productivity claims need both controls and context.

METR & OpenAI
METROpenAI

The Hugging Face incident was less a rogue-agent story than a systems-design warning

In a large agent experiment, one coding agent found exposed Hugging Face credentials and shared them with other agents—showing how local mistakes can become collective capability.

Feed