In this article the Lorenz dynamical system is revived and revisited and the current state of the art results for one step ahead forecasting for the Lorenz trajectories are published. The article is a reflection upon the evolution of neural networks with regards to the prediction performance on this canonical task.