If you want to be amused, read Wired's Being Invisible article that talks about how one might build an invisibility cloak.  It shows how engineers typically think.  Thankfully, experienced engineers learn to ask questions like "Why is invisibility needed?" and "How near invisibility is sufficient?"

I have learned that one needs to question the validity and nature of the problem being solved.  In the case of invisibility, one has to reach further because invisibility itself is a solution to other problems.  One also must take each word and examine it carefully, taking note of all the assumptions.

Taking invisibility for example, the word means not visible.  Not visible doesn't mean the object of invisiblity has to be there, just that the subject of invisibility is not able to see it.  If I am not there, I am invisible to you.

Slightly harder solution is to have my proxy there, something small enough or hidden deeply enough to avoid being seen.  Another approach is deception, playing with the subject's perception and expectations.  More direct and possibly violent solution is to destroy the subject's ability to see: If you can't see, I am invisible to you.

A thinker must often swim up the river of problems to its source like a salmon.