There is a well-known misconception that financial machine learning applications are mostly about predicting the price of an asset. As a result, the majority of investment professionals were disappointed with the applications of ML techniques in finance. In the meantime, those who apply AI for investment management in the right way achieve way better performance compared to their peers. In this lecture, I would like to show several examples of successful applications of ML in finance as well as several cases where AI applications lead to overfitting and poor out-of-sample performance.