Deep learning

Deep learning is machine learning built from many layers of simple units, each one passing its result to the next. “Deep” = lots of layers, not lots of complexity.

Each layer learns slightly more abstract features. In a vision model: edges → shapes → parts → objects. In a language model: characters → words → phrases → meaning.

That stacking is what unlocked the modern era — it’s how models got good enough at language, images, and audio to feel almost human.