Original Summary

I used to be a software engineer, but these days I find myself being more of an agents manager: constantly switching between 3-4 tasks that my agents are working on, and making sure that the features get shipped, the bugs fixed, with correct and clean code. There&#x27;s no doubt that I&#x27;m producing more value for the company, but I wonder if that&#x27;s the best I can do for myself if the goal is staying relevant in the coming years.<p>I wonder: if the AI is writing all the code, what value am I bringing?<p>In an attempt to keep my critical thinking and coding skills sharp, I sometimes foolishly try to debug errors or implement a feature myself. But it feels like wasting time when an agent can do it in a fraction of time.<p>What is gonna be sought after in the coming years: someone that persevered with manual coding and debugging but is slow and uncomfortable around agents, or someone that has little coding and debugging skills left but knows how to orchestrate dozens of agents effectively?<p>I&#x27;m curious to hear what other software engineers on HN&#x27;s approach is.

  • 情报分类：技术学习与提效
  • 分类依据：内容涉及技术、AI、软件工具或工程实践
  • 信息来源：Hacker News 新项目
  • 发布时间：2026/9/24 02:35:34