Why strong developers hold back on coding agents, and what an all-agent build showed
What happened
A Hacker News front-page discussion on Oct 29 looked at why good developers are cautious about fully adopting coding agents, with concerns centered on production reliability, long-term maintenance and engineering guardrails. Thoughtworks ran a team experiment in which agents wrote all the code: about two-thirds of time went to whiteboard design and one-third to execution and checking frameworks. The app ran, but builds failed often. The article argues for involving experienced developers to test whether tests, system boundaries and continuous delivery practices can support this way of working, rather than judging success by code output alone.
Written by AI from the coverage · updated 1 hour ago
Coverage
Follow the reports to see the story from different sides.
- Hacker News front pageWhy are good developers so wary of agentic development?
优秀开发者对全面采用编码智能体的谨慎，源于对生产可靠性、长期维护和工程护栏的担忧。Thoughtworks 组织团队开展全由智能体编码的实验，团队约 ⅔ 时间用于白板设计、⅓ 用于执行与检查框架，应用虽能运行，构建却频繁失败。文章主张让经验丰富的开发者参与实验，检验测试、系统边界和持续交付实践能否支撑这种开发方式，而非仅以代码产量衡量成功。
Heat over time
Not enough continuous observations to draw a trend yet.