I have just came across a short but nice paper by Larry Wasserman called Raise of the Machines. It is attached with a funny image of Terminator, and in some ways hilarious, but he does have some interesting points to talk about the “evolution” of Machine Learning during the years, from the perspective of a statistician:
1. ML and Statistic are eventually the same thing. Some people, including Geoffrey Hinton (at 1:14), still have some ways to discriminate the two, but the distinction is probably getting faded away as both fields getting close to each other.
2. The “conference culture” has pushed ML move forward in a fast pace.
3. “ML has much to offer Statistics. And Statisticians have a lot to offer ML”.
And some other points that can be found here in detail.