What is a model? What is an algorithm?
P10.orientation.01 · Audience: guest, it-ml, language-pro
Two words appear everywhere in AI — model and algorithm — and almost every later idea rests on them. Take them one at a time, starting from things you already know well: a map, a grammar, a recipe. No maths, no code. Your glossary is one click away whenever a term is new.
🗣️ From a linguist's perspective: a model is like a grammar
A grammar is not the language itself — it is a compact description that captures how the language tends to behave, useful enough to teach, translate and correct with, while quietly leaving out countless exceptions. That is exactly what a model is: a deliberate simplification of something real, kept because it is useful, not because it is complete. When you reach for a grammar rule instead of every sentence you have ever heard, you are already using a model.
A modelModelA set of patterns learned from examples that turns an input into an output. Unlike a fixed recipe (an algorithm), a model's behaviour comes from the data it was trained on, not from rules someone wrote by hand. is a simplified stand-in for something in the real world — small enough to reason with, close enough to be useful. The classic example is a map. A city map leaves out almost everything: the weather, the traffic, the colour of the doors. It keeps just enough — streets and their connections — to answer one kind of question well: how do I get from here to there?
The famous saying is “the map is not the territory”. A model is never the real thing; it is a chosen simplification. A good model keeps what matters for its job and drops the rest on purpose. So the useful question is never “is this model true?” but “is it useful for this task, and where does it break?”
Worked example — three models of the same city
A street map, a metro map and a tourist map all model one city. The metro map even distorts distances so the lines stay readable — and that distortion makes it better at its one job. Same reality, different simplifications, each right for a different question. A language model is a model in exactly this sense: a simplified stand-in for how language behaves, built to answer “what text is likely to come next?”
📚 Go Further
Plain-language explainers on models and algorithms — no maths required.
Ask the mentor about this module
Ask a question about this content. The mentor explains and grounds its answer in what you are studying; asking is recorded as a learning signal, not a grade.
Where next?
Later in Orientation — concepts before the maths