Basics
What is a large language model (LLM)?
Also called
- LLM
- language model
Definition
An AI model trained on huge amounts of text so it can read, write, summarize and answer questions in everyday language.
Key takeaways
- An LLM is the engine inside most chat assistants.
- It predicts which words are likely to come next, so it writes fluently but can state wrong facts with confidence.
- The app you use and the model underneath are not always made by the same company.
In plain English
An LLM is the engine inside most chat assistants. It works by predicting which words are likely to come next, which is why it can write fluently but can also state wrong facts with confidence. The app you use and the model underneath are not always made by the same company. Example: the assistant that summarizes a long supplier contract for you is powered by an LLM.
A small business example
Picture a cafe owner pasting a long supplier contract into a chat assistant and asking for the key terms. The LLM underneath produces a clear summary quickly. Because it predicts likely words rather than looking facts up, the owner checks the payment terms and notice periods against the contract itself before relying on the summary.
Questions
Why does an LLM sometimes sound sure about something wrong?
It works by predicting which words are likely to come next, not by checking facts. That is why it can write fluently and still state wrong facts with confidence.