Papers & Research

METR releases time horizon study: task length AI can complete doubles roughly every 7 months

Original title:Measuring AI Ability to Complete Long Software Tasks

AI summary

METR proposed measuring model capability by the length of tasks AI agents can complete, measured by the time required by human professionals, and found that this metric has grown exponentially over the past 6 years, with a doubling time of about 7 months, and less than 3 months on SWE-Bench Verified.

Key Takeaways

  • METR proposed measuring model capability by the length of tasks AI agents can complete,measured by…
  • and found that this metric has grown exponentially over the past 6 years
  • with a doubling time of about 7 months,and less than 3 months on SWE-Bench Verified.

Quick Facts

Published
Updated
Source
METR:Research(网页)
Category
Papers & Research
AI score
75
Read the original Read the full article at the original source
Why it matters

METR 用任务时长衡量模型能力,给出六年间指数增长趋势和约 7 个月翻倍速度,为理解模型真实工作能力提供一个可比较的框架。

Source & copyright

This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.

Original article · AIHOT entry

How to Cite

When citing this article, keep the title, date, and permalink, and credit Orange News as the source.

Orange News: METR releases time horizon study: task length AI can complete doubles roughly every 7 months. March 19, 2025. https://new.nat6.net/en/news/cmtyo7nb703vgrog0j0ha07cl/

View all →