Machine learning refers to methods which explore the study and construction of algorithms that can learn from and make predictions on data. Deep learning is a particular technique in machine learning, but becomes a very powerful hammer for artificial intelligence. In this talk, I begin with a story on success of deep learning in the task of speech recognition, computer vision, natural language processing. In addition, I will introduce a few exemplary products developed by global IT companies such as Google, Facebook, Amazon, and so on. In the second part of my talk, I will introduce recent advances in deep generative models, in the perspective of prescribed or implicit models. Highlights are given to variational autoencoders and generative adversarial networks, as well as their modifications.