Contrastive learning, exemplified by SimCLR, CLIP and many others, has achieved impressive results for learning visual representations from images and image-text pairs. In this talk, I will cover topics on contrastive self-supervised learning, including an overview of a few basic contrastive methods, important factors in contrastive learning, simple approaches for semi-supervised learning (with lots of unlabeled images and a few labeled images), some intriguing properties and open challenges of existing contrastive learning methods.