ResearchNovember 09, 2026
-26
🧊
OpenAI's Time Horizon Doubles Rapidly
An analysis suggests that OpenAI's internal AI R&D time horizons appear to double roughly every two months.
#OpenAI#METR#time horizons#AI R&D#benchmarks

🔥 What happened
OpenAI researcher Basil Halperin analyzed internal data: the 50%-accuracy time horizon doubles every 2 months—much faster than METR's 4-7 months. At 80% accuracy, it's only 15 minutes but rising rapidly.
💡 Why it matters
The gap between 50% and 80% is huge: OpenAI shows a 15x difference, METR only 4x. This could stem from different domains or contexts—and makes predicting AI capabilities harder.
âš¡ Our take
OpenAI is accelerating its own automation—a warning to anyone who thinks AI progress is linear. If you're not paying attention, you'll be run over by the curve.
The title, summary and analysis of this item were produced automatically by an AI system and have not been editorially reviewed. They may contain errors, bias or omissions — when in doubt, read the linked original source.
Deep Dives & Similar Intelligence

Anthropic's AI is accelerating itself 10x
ResearchNovember 06, 2026

Anthropic's AI R&D Doubles, But Evidence Hidden
ResearchNovember 06, 2026

OpenAI's AI patches 948 Linux kernel bugs
DevOpsNovember 05, 2026
ki-daily.
OpenAI's AI finds 948 Linux root exploits
Ethics & SecurityNovember 05, 2026

OpenAI Unleashes GPT-6 Cyber on Bugs
AI modelsNovember 06, 2026