RNN (Recurrent Neural Network)

Friendly Description: A Recurrent Neural Network is a kind of AI built for things that come in order, like sentences, music, or stock prices. Unlike most networks, an RNN keeps a small "memory" as it goes, so each new piece of information is understood in the context of what came before. RNNs were one of the big building blocks before transformers took over.

Example: Early voice transcription systems used RNNs to listen to audio one tiny chunk at a time, remembering the sounds they had just heard so they could correctly stitch them into words. The memory is what helped them understand that "I scream" and "ice cream" sound similar but mean different things in context.