Whenever clever people work on new problem-spaces (ie. not yet mined-out), they have good chances of being the ones to have the fundamentally new insights; so your odds are good.
But the two kinds of technology-based businesses make me think of Google's technology: server implementation (fast results) and pagerank (relevant results).
Of course, there have also been insights that I think may be fundamental, but they aren't patentable as far as we know.