Machine learning
Also known as: ML
Machine learning is the practice of building software whose behaviour is derived from examples rather than written as rules. Instead of a developer specifying "reject the invoice if these six conditions hold", a model is shown many past invoices and their outcomes, and it infers the pattern.
This is the right approach when the rules are real but nobody can state them exhaustively — spotting fraud, ranking search results, predicting which customer is about to leave. It is the wrong approach when the rules are known and stable, because a rule you can read and edit will beat a model you have to retrain.
The practical cost of machine learning is rarely the model. It is the training data: collecting it, labelling it, keeping it current, and proving it is not quietly biased in a way that will embarrass you later.