Living systems hinge on self-organization and learning. We will show basic principles of self-stabilization and learning, using two complementary systems — Living Neural Networks that have been optimized for emergent information processing, and Societal Networks, which might not always be optimal. We will examine their emergent dynamics, how they (fail to) self-stabilize with homeostatic mechanisms, and how they can learn in a self-organized (unsupervised) manner. Thereby, we shed light on how self-organization shapes these living systems so that they can navigate their complex environment across scales.