Why AI May Make Us Busier: Three Productivity Paradoxes
In a commentary from the WeChat account 卫夕指北, author Wei Xi uses the Cowan, Solow and Jevons paradoxes to examine why AI may increase workloads without yet producing clearly visible productivity gains. Wei argues that faster drafting and lower model prices can lead to higher standards, more output and greater usage, potentially raising aggregate spending. The commentary also notes disagreement over when measurable productivity improvements will appear.
The commentary provides an attributed framework for assessing AI productivity claims, while its observations about busier workers and rising spending are the author’s synthesis rather than universal measurements.

先说个我自己的直观感受:AI写文章越来越牛逼,三分钟就能吐出一篇像模像样的稿子。但要写出一篇走心的、大家愿意读的文章,一点也没有变得更容易。我很多大厂的前同事也天天在用AI,但很明显,老板的要求越来越高,很多产品经理都要求拿着能运行的demo去过评审会。最终,AI并没有给我们省时间,反而让我们越来越忙。今天卫......
本文来自微信公众号: 卫夕指北 ,作者:卫夕
先说个我自己的直观感受:
AI写文章越来越牛逼,三分钟就能吐出一篇像模像样的稿子。
但要写出一篇走心的、大家愿意读的文章,一点也没有变得更容易。
#DeepSeek#Nvidia#Microsoft#Google#OpenRouter#Epoch AI#Upwork#Workday#MIT#OpenAI