A Model is a Function
At its core, a machine learning model is just a function: it takes input and produces output.
INPUT → [ ? MODEL ? ] → OUTPUT
Temperature, Humidity → [Model] → Will it rain?
Image pixels → [Model] → Cat or Dog?
Text → [Model] → Positive or Negative sentiment?The "magic" isn't IN the model - it's in HOW the model's parameters got set during training.