>>13173140t. someone who graduated from my department
I go to a relatively small state school, and my department tends to focus on combinatorics, network science/graph theory, and mathematical biology. Probably >50% of grad students coming out of my program are working as "data scientists".
It's not exactly what I was hoping to do, but I imagine that's the route I'll be going. Data science actually sounds kind of cool, mainly because you can work on everything from biology, to economics, to sociology, to physics, to political science, to computer science, to transportation, logistics, anything. Pure or applied. Industry or academia. Almost anywhere that is working on the forefront of science or engineering, you are likely to run into data science, and it's really cool that it's allowing us to approach a lot of topic in the humanities and social science in a rigorous manner for literally the first time in human history.
Admittedly, I haven't done any data science myself, but I've read a ton of papers that incorporate data science, and they're very repetitive a similar. It's seems like there is really no creativity or deep thinking involved. It seems like its just about applying mathematical tools to big data sets, and then using computers to crunch the numbers. Every single paper for the past 20 years is just:
(1) Collect 20,000 data points concerning some topic that you're interested in
(2) Pick some random feature of the data set that might be interesting, e.g. whether person A has retweeted person B on Twitter.
(3) Define a metric on your data set
(4) Use a computer to find the correlation between the metric you introduced and whatever feature of the data set that you're interested in.
It's all very meticulous, but it kind of reminds me of a more STEMish version of social psychology, where you have people using really meticulous experimental methods, but ultimately the field is over-saturated, not evolving, and many results tend to be trivial or obvious.
