我们构建 AI 编程助手的方向对吗？
2 分•作者: anaempromptu•5 个月前
本周 Jason Lemkin/Replit 事件让我开始思考我们目前使用 AI 编程助手的根本问题。
我们都看过演示——将自然语言转化为可运行的代码，进行对话式调试，以及“只需描述你想要什么”。但当你真正投入生产时，一切都崩溃了。
我一直在关注的核心技术挑战包括：
* 大规模上下文管理——这些系统在孤立的任务中表现出色，但在处理复杂、多文件项目时，很难维持连贯的状态。你如何处理跨越数十个文件、数千行代码的上下文？
* 安全与能力的权衡——更强大的工具可能造成更大的破坏。Replit 承诺不触及生产环境，但最终还是删除了一个数据库。你如何构建真正有效的安全防护措施，而又不会削弱工具的功能？
* 复杂系统的对话式界面——自然语言具有歧义性，而代码是精确的。我们是否试图解决错误的界面问题？
* 生产差距——我测试过的每个 AI 编程工具在演示中都表现出色，但在处理真实代码库、真实数据和真实边缘情况时，就会崩溃。为什么这种差距如此持久？
我真的很想知道——是否有人使用在生产环境中真正可靠的工具构建了 AI 应用？
查看原文
The Jason Lemkin/Replit incident this week got me thinking about the fundamental problems with how we're approaching AI coding assistants.
We've all seen the demos - natural language to working code, conversational debugging, "just describe what you want." But then you hit prod and everything breaks down.<p>The core technical challenges I keep seeing:<p>- Context management at scale - These systems work great for isolated tasks but struggle to maintain coherent state across complex, multi-file projects. How do you handle context that spans thousands of lines across dozens of files?<p>- The safety/capability tradeoff - More powerful tools can do more damage. Replit promised not to touch production, then deleted a database anyway. How do you build guardrails that actually work without neutering the tool?<p>- Conversational interfaces for complex systems - Natural language is ambiguous. Code is precise. Are we trying to solve the wrong interface problem?<p>- The production gap - Every AI coding tool I've tested works beautifully in demos and falls apart with real codebases, real data, real edge cases. Why is this gap so persistent?<p>I'm genuinely curious - has anyone built AI apps with tools that actually work reliably in prod?