We modelled the material lead titanate based on the underlying quantum mechanics of our world with the help of machine learning. Our approach depicted the lead titanate with accurate details like a realism painting, in contrast to the traditional methods that only provided qualitative description of a material in the spirit of impressionism. I will compare our simulation results to experimental data to show the accuracy of our modelling approach and its predictive power for ideal materials that were difficult to be synthesized perfectly in the lab.