The author describes the old, pre-revolutionary way of doing mathematics as "relying on intuition and physical experience." In certain fields, such as algebra and number theory, the old way of doing it had another important trait: to prove the existence of something, such as the greatest common divisor or the unique factorization of a polynomial, one had to give an algorithm to construct it. The new, axiomatic way of thinking, by contrast, is content with showing that the assumption of non-existence leads to a contradiction. The rise of symbolic computation over the last few decades (Mathematica, Maple, etc.) can therefore be seen as a comeback of the old style of mathematics. I call it Kronecker's revenge. With respect to math education, well, considering the importance of computing, I am inclined to think that we should pay more attention to the 19th century way of thinking rather than less.