This booklet introduces the most important basics of deep learning. It describes the very frequently used method of how a computer can learn using neural networks and training data and apply what it has learned to other questions similar to the training data. Two simple about 'one-pager' examples in Python show how training a neural network with forward and back propagation works and how the trained system can process simple forms of artificial thinking. The two short Python programs "Learning truth tables" and "Recognizing a questionnaire" are printed in full and are easy to follow.