Application of artificial intelligence has made revolutionary advances in many medical scenarios such as diagnosis, drug screening, medical imaging and nursing. Images of pathological sections (2D) are a preliminary breakthrough of artificial intelligence. Abundant medical and pathological resources, and standardization, digitization of pathological section images provide big data for the in-depth study of artificial intelligence. Through a series of research on artificial intelligence in pathologies of breast cancer, gastric cancer, and cholangiocarcinoma, we established a labeling standard of tumor cells and an in-depth learning process, and we also developed an artificial intelligence algorithm for hilar cholangiocarcinoma. However, there are still many problems and we tried to search for the solutions. With the improvement in precision of artificial intelligence diagnosis, pathology of artificial intelligence may be soon applied in future clinical practice.