METR’s data can’t distinguish between trajectories (and 80% horizons are an order of magnitude off)
Published on February 13, 2026 2:05 PM GMTTLDR I reanalyzed the METR task data using…
Recent Articles
View All
Meta plans to add facial recognition to its smart glasses, report claims
The feature, internally known as “Name Tag,” would allow smart glasses wearers to identify people…
Hazardous States and Accidents
Published on February 13, 2026 1:02 PM GMTRoot cause analysis is a crap technique for…
Conductors to Orchestrators: The Future of Agentic Coding
This post first appeared on Addy Osmani’s Elevate Substack newsletter and is being republished here…
GPT-5.2 derives a new result in theoretical physics
A new preprint shows GPT-5.2 proposing a new formula for a gluon amplitude, later formally…
Systemic Risks and Where to Find Them
Published on February 13, 2026 10:51 AM GMTOr: Todd Has a Presentation in London on…
Introducing Lockdown Mode and Elevated Risk labels in ChatGPT
Introducing Lockdown Mode and Elevated Risk labels in ChatGPT to help organizations defend against prompt…
Scaling social science research
GABRIEL is a new open-source toolkit from OpenAI that uses GPT to turn qualitative text…
Beyond rate limits: scaling access to Codex and Sora
How OpenAI built a real-time access system combining rate limits, usage tracking, and credits to…
Nick Bostrom: Optimal Timing for Superintelligence
Published on February 13, 2026 7:33 AM GMTLinked is a new working paper from Nick…