The Primordial Soup of Vanilla RNNs and Reservoir Computing
Using past experience for improved future performance is a cornerstone of deep learning and of machine learning in general. One definition of machine learning lays out the importance of improving with experience explicitly:
A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E.