So, I am a total ML noob. The thing I haven't found a straight answer to is, what is a model. I mean when it is in production? Is it just some random blob that you pipe data into and get data out?
Your input value is X, you multiply it by your slope and add your intercept to get the output (the Y value on the line).
The 'training' of an ML algo is really just finding the line-of-best-fit so that you can make predictions. So your line-of-best-fit is encoded in these two parameters, allowing you to make predictions about what the output would be for arbitrary input.
The problems people are throwing at ML have many more parameters and dimensions, but the training is a matter of finding those parameters that come closest to predicting the outcome. The 'model' is this set of parameters that allows the function to make predictions.
(disclaimer: also an ML noob, correct me if I'm wrong)