AI coding agents still need human cleanup on nearly three out of four tasks in a real-world test

Alibaba's new coding-agent benchmark, Co-Design Bench, tests AI agents on real user requests that already tripped them up in practice. Even GPT-6 Astra, the top performer, fully solves only 26.4 perc…

Alibaba's new coding-agent benchmark, Co-Design Bench, tests AI agents on real user requests that already tripped them up in practice. Even GPT-6 Astra, the top performer, fully solves only 26.4 perc…

可信度：待确认 · 来源 The Decoder · 事件时间 11-08 17:50（北京时间）

打开

The Decoder

浏览最新 AI 动态