You have seen them many times. Small snippets of code and the question “What would the the following piece of code print” beneath. You might have tried taking a guess, and perhaps, even failed.
I find those kinds of questions utterly pointless. They not only teach you nothing about real-world programming, but might also lead many newcomers to long-term frustration with programming.
All programming languages have baggage - obscure features that made it in the spec but were later deemed as hacks that should be avoided. There is a certain sense of pride that junior programmers feel when they find such hacks and use them to demonstrate problem-solving skills. I am all in favor of encouraging developers to be proactive and think out of the box. Yet, I often try to point out that using questions like the above as a way of judging one's skills is the wrong way to do it. The mere fact that an opportunity for such questions exists in the first place, should make one take a skeptical look at the language itself.
Contrary to what your teacher taught you, real-world programming is all but proving yourself at solving complex riddles. Much of the well-written production code I have seen is pretty trivial and boring-looking like. This makes it easy to follow and maintain years down the road, once its original creator no longer works on it.