Mmmmm.
I enjoy theoretical aspects of science,even if i have an engineering background,i think that abstract and theoretical inclination is more beneficial in long term scale....
Even if the applied stuff is better for your pocket :),the theoretical breakthroughs are more seminal and their implications would shape almost every aspect of our lifes.
For instance,say that you prove and give an answer to an open problem-like Grisha Perelman did-and his breakthrough was of global respect and would shape the science for the upcoming years,although a small fraction of people knows his work and how seminal is...Albeit everybody knows Google,which is not that clever but is was breakthrough in ~1995-98 and uses obsolete mathematics:)-I mean it is advanced but not that advanced:)-.
For those who have an interest in theory you can find the blog of Luca Trevisan who is professor in UC Berkeley in CS Division somewhat amusing .Go check it....
