While languages don't get better over time they do change which means that there are more of them, which means it is more likely that you will find some that suit you.
Someone who absolutely loves parentheses and capital letters would probably favor LISP 1.5 over Clojure.
On one hand, I can definitely look at C# right now and see how it is a vast improvement upon C# 1.0. I'll even go as far as to say I would probably refuse to write code in C# if it had not become what it is today.
On the other, I think you can only really change a language up to a certain degree before creating an inconsistent mess. If trends in the future dictate the need for a shift in paradigms, the extent to which existing languages can accommodate this shift is probably limited. I mean, you can often incorporate features that are based on the new approach to thinking about a problem, but once you start changing the core semantics of the language you're probably going to be better off with something completely new.