个人软件作品集的时代结束了吗？
自从人工智能代理的推出以来，通过代理命令行接口（Agent CLIs）开发原型和小型软件应用（如常见的网页和Python应用）变得显著容易。我个人对软件的看法与三年前相比有些不同。技术或架构方面对我来说不再那么有趣，因为这些内容可能是借助人工智能生成的。这并不能说明这个人真正了解业务层下发生的事情，因为他可能并不是实际编写代码的人。是的，以前也可以通过强大的CTRL-C CTRL-V来实现，但现在我们有了自动化的CTRL-C CTRL-V，几乎不会将代码粘贴到错误的位置。
尽管如此，我觉得当一个应用达到特定规模时，人工智能仍然无法满足需求。这意味着开发者实际上需要对解决这些问题所需的概念有深入的架构理解，这在人工智能时代之前就是一种要求。但这是否意味着我现在对更雄心勃勃的软件的价值看得更低了呢？
在我看来，当前的软件开发时代相当不稳定且奇怪，没有人真正知道，所有的一切只是意见之争。
所以这是我的看法。这些工具的崛起是否降低了个人软件作品集的价值？你怎么看？
查看原文
Since the launch of AI agents it has become significant easier to develop prototypes and smaller software applications (of the most common like web and python apps) through Agent CLIs.<p>I personally view software a bit different now than I did over 3 years ago. The technical or architectural aspects is not as interesting to me anymore, since they could have been generated with the help of AI. This does not tell me that the person actually knows what is happening beneath the business layer, since the person might not have been the one who actually wrote it. Yeah that was doable before too with the help of almighty CTRL-C CTRL-V, but now we have an automated CTRL-C CTRL-V which rarely pastes the code in the wrong location.<p>Although I feel that when a specific scale is reached for an application, AI does not suffice anyway. That means the person actually requires to have deep architectural understandings of the concepts required to solve these problems, which was kind of the requirement before AI era. But does that just means I now value more ambitious software lower?<p>The current era of software development is quite shaky and weird IMO, and no one actually knows, everything is just an opinion-war.<p>So here goes my opinion. Has the rise of these tools lowered the value of personal software portfolios? What do you think?