And if your program doesn't work, that should be your first guess.

如果你的程序无法运行的时候，它们可以成为你的第一猜测。

麻省理工公开课 - 计算机科学及编程导论课程节选

The point I'm trying to make is, fixed-program computers is where we started, but it doesn't really get us to where we'd like to be.

我要指出的一点是，我们从固定程序计算机开始说起，但它并不能让我们到达我们想去的地方。

麻省理工公开课 - 计算机科学及编程导论课程节选

You can very quickly drive yourself nuts by sitting down, thinking, oh, I can write this program, no sweat, and then you write it and then there's a tiny little mistake here, but then there's a tiny little mistake here and here, and all of these stupid little things cascade until you try running your program, and the thing just doesn't work.

如果你只是仅仅干坐着，不费吹灰之力，凭空幻想着我可以把这个项目搞定，你会想地发疯的，也许你会写出一点东西出来，可能会有一点点错误,紧接着这或那又出现一点小错误，这些小错误一点一点地堆积了起来,直到你开始运行你整个项目，你会发现这堆东西根本就不工作。

哈佛公开课 - 计算机科学课程节选

And so similarly as Gmail's priority inbox, it's gonna be pretty flawed for a lot of us at least initially because it just doesn't know the answers a priori correctly unless we teach it or program it to be better. There is also another flaw in here or multiple flaws arguably.

类似于Gmail的优先收件箱，在起初使用时对大多数人来说它是有缺陷的,因为它不知道如何正确判断优先级,除非我们事先告诉它或编写程序改良它,另外这里存在一个或多个漏洞，可以说是缺陷。

哈佛公开课 - 计算机科学课程节选