The artist uses an adversarial generative neural network (GAN) algorithm to paint the damaged Dunhuang portraits with thousands of years of history, exploring whether it is possible for information technology to make up for the irreversible time, the love and loss of civilization. The GAN-generated portraits with curly hair and deep eyes resemble tiny Johann Strauss and Beethoven, looking at the shore in the background of the algorithm. The artist also wants to express in this series: the history that is happening now is in history per se. Time is not purely linear. All that we have in the present is the result of the future.