From the a question I've rephrased here to apply to engineering schools:

...why on earth do we assess the way we do? If I gave an engineer a problem and said, "whatever you do, don't do any research on this, and, regardless of the amount of time it may need for competent work, you've got three hours," the engineer should say, "I can't do that." Yet, that's the dominant testing model...

One of the hardest things I've had to learn since graduating from college is how and when to give up, and how to say things like "that's impossible," or "that's not worth it, and we aren't going to try." Of course, that was only a good and valuable lesson to learn because I had already learned how to keep pushing through hard problems, how to not give up, how to keep on going even if somebody else said that it was impossible. The drive to push my limits had to be ingrained in me before I could learn for what reasons I should and should not push them.

Your mileage may vary.