I explain the (conjectural) Fano/LG correspondence and how it leads to a program to classify Fano varieties by first classifying their mirrors.