I don't care for Kurzweil but I don't see anything wrong with hypothesizing about something and then designing experiments to test that hypothesis, rather than

I don't care for Kurzweil but I don't see anything wrong with hypothesizing about something and then designing experiments to test that hypothesis, rather than just being a mindless "data scientist." :) Sounds pretty normal to me.