I see a lot of graduate students focusing on practical uses of ML algorithms as a result of this. A lot of people don't realize that a good portion of the math is already figured out, and that it's in the implementation of these algorithms that they can find more interesting results.