I have been amazed at how many relatively minor things related learn to my current work that were only developed from the 1990s or later. Having received my doctorate in 1993, "the 1990s" are kind of a bridge for me; "after I got my Ph.D." if truly disruptive discoveries are made will inevitably be a kind of comparison one makes with others, and makes in terms of one's skill set.
I cannot go into any of the details, of course, but I was surprised to find elements of optimization theory, the theory of matrices, and other areas that simply didn't exist 16 years ago, and how areas of topology are increasingly creeping into electrical engineering. Were I to get my Ph.D. today, I would certainly include courses in topology and functional analysis, rather than rely on the stuff I picked up in systems optimization and real analysis (though that did quite nicely). Ah yes - now I remember- I never took functional because I did get most of what I needed in real analysis. But nowadays it's gotten much more infused with topological concepts.
I suspect that the only real job security in engineering is kick-ass research: otherwise you get obsolete. And right now, that's just what I'm doing.
What to expect to see from Google's October 4th event
16 minutes ago