When I saw the title, I thought it would be about the information theoretic application of entropy, and talk about the maximum densities of useful work that can be expressed in a given number of characters. I further imagined the article talking about how abstractions are often good, and the move toward more expressive languages is in the large positive, but then exploring how it can be overdone like in the "Perl golf" exercise; and perhaps attempting to find a function or algorithm defining the optimal density for a balance between compactness and readability.
That'd be an interesting article.