METR 发布 MirrorCode 初步结果链接，称有证据显示 AI 已能完成部分持续数周的编码任务

MirrorCode: Evidence that AI can already do some weeks-long coding tasks

AI 导读

METR 发布 MirrorCode 项目链接帖，该项目由 METR 出资并与 Epoch AI 共同开发，详细结果见 Epoch AI 博客 https://epoch.ai/publications/mirrorcode-preliminary-results/。页面另提及 METR 对 349 名技术工作者的调查，自述因 AI 工具工作价值变化中位数为 1.4–2x，且预期随时间增长。

正文 · 原文

CONTRIBUTORS

David Rein

DATE

April 10, 2026

This is a linkpost for MirrorCode, a project that METR funded and co-developed with Epoch AI. See Epoch AI’s blog post for more detail: https://epoch.ai/publications/mirrorcode-preliminary-results/

Cite



@misc{metr-2026-mirrorcode-preliminary-results,
    title = {MirrorCode: Evidence that AI can already do some weeks-long coding tasks},
    author = {David Rein},
    howpublished = {\url{https://metr.org/blog/2026-04-10-mirrorcode-preliminary-results/}},
    year = {2026},
    month = {04},
}

Featured research

METR researches, develops and runs cutting-edge tests of AI capabilities, including broad autonomous capabilities and the ability of AI systems to conduct AI R&D.

Measuring the Self-Reported Impact of Early-2026 AI on Technical Worker Productivity

A survey of 349 technical workers finds a median 1.4–2x self-reported change in value of work due to AI tools, expected to grow over time, though there are reasons to be skeptical of the magnitude.