Ethan Mollick· emollick·· 10 天前AI 评分26
AI 能力超越基准测试饱和
The few attempts to measure similar things (METR Long Tasks, GDPval) all became saturated earlier this year as AIs began to do days of wo...
AI 导读
少数尝试衡量类似指标的基准（METR Long Tasks、GDPval）都在今年早些时候达到饱和，因为 AI 开始以高质量水平完成数天的工作。
来源：Ethan Mollick · x.com