The 2019-nCoV coronavirus epidemic from Wuhan WHO reported on January 9. It turns out that artificial intelligence knew about it almost 2 weeks earlier. Canada's BlueDot system, which uses data from over 100 different sources, builds complex models of disease incidence and spread, able to predict risk well in advance. It uses machine learning and natural language processing methods, for example by scanning local news in search of signals indicating an epidemiological threat, data from hospitals, climate and demographic data, local and air transport statistics, and internet forums.