A scientist’s work is never done.
无法避免的苦难,越早承受越好。
A good learning algorithm should be able to generalize from a limited set of examples to unseen data.
The architect must be a guardian of the future.