Mastodon上的几位用户正在讨论在编码中使用AI，普遍的共识是AI应该被视为一种工具，而不是人类程序员的替代品。一些用户表示个人喜欢编写代码，不会将这项任务委托给AI。其他人则强调对AI生成的代码进行彻底审查和测试的重要性，并将其与人类编写的代码的编译过程进行比较。提出的一个关键点是，随着AI编写代码的速度越来越快，需求的清晰度和特异性必须提高，以避免AI做出无益的假设。 AI
影响 AI越来越被视为一种编码工具，需要更清晰的需求和人工监督，以确保负责任和有效的集成。
排序理由 该集群包含用户在社交媒体上关于在编码中使用AI的意见和讨论，而不是主要的发布或重要的行业事件。
在 Mastodon — fosstodon.org 阅读 →
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
I like this comparison between # AI and # Perl in # research . I've always claimed that, depending on the programmer, Perl code can range from truly beautiful to ugly and unintelligible. I still stand by this statement. https:// davetang.org/muse/2026/07/18/b ioinformatics-and-ai…
Here's a big part of the reason* I don't use so-called # AI to write code for me: I actually enjoy writing code. I wouldn't offload this task to AI for the same reason I wouldn't use it to go for a walk in the woods, or eat a delicious meal for me. Why would any sane** person do …
Using # AI to write something you don’t fully understand can happen. Compilers do the very same thing to your code every day. But shipping AI-authored code without thorough review and testing is just as irresponsible as shipping your human-authored code just because it compiles w…
Hard truth: the faster # AI writes code, the less vague our requirements can be. 💎 Define the functionality. 💎 Define what "done" means. 💎 Define what's NOT included. AI won't notice the gaps; it will fill them with assumptions! 👉️ For more visit: www.softwareantifragility.com