The trained core of an AI: a big bundle of numbers that turns an input into an answer.
In AI, a model is what you get after training. Think of it as a large file: a fixed recipe for how to calculate, filled with numbers (parameters) that training has tuned. Give it a question or a photo, and it runs those numbers to produce an answer.
The chat app you use is the window; the model is what actually writes the reply inside it. An app can swap one model for another, and one model can power many apps. Models come in all sizes, from huge ones that only run on servers to small ones that fit on a phone.
Think of it like this
Think of a music app and a song file. The app gives you buttons and a screen; the actual sound lives in the file. With AI, the app is the window and the model is the part that makes the answer.
For example
When an app says it has "switched to a new model", the window stayed the same but the core inside changed.
Common mix-ups
Mix-upAn AI app and a model are the same thing.
ActuallyBesides the model, an app adds its own features, such as web search, saving your chat history or filtering out harmful answers. So the same model can do different things depending on which app you use it through.
Mix-upA model keeps learning and changing every time you ask it something.
ActuallyA model has a cut-off date for the material it learned from, so it doesn't know later news unless a tool like search helps. If an AI seems to remember you, it's usually the app saving notes separately and attaching them to your question.
Try it in class
For teachers at the start of a lesson — or for students working on their own.
- Question
- If the same app answers in a different style today than yesterday, what might have changed?
- Activity
- Use the music-app-and-song-file comparison to make a two-column table. In the "window (app)" column, list what you can see — the text box, buttons, chat list. In the "core (model)" column, write what computes the answer. Then pick one use that suits a big model running only on servers and one that suits a small model inside a phone, and explain why. Lesson DAY 09, "Is a bigger model always better?", picks this up, and DAY 10, "Could this AI fit on my laptop?", goes on to open and closed models.