How does AI work with language?
Natural language processing (NLP) covers translation, classification, working out meaning and generating text.
What is it?
Natural language processing (NLP) is the area of AI that deals with human language: translation, summarising, classifying, finding meaning and writing text. Today it is driven mainly by large language models, but language stays hard because of ambiguity and culture.
An everyday example
An online translator turns a menu abroad into your language, and your email program suggests a short reply.
Why does it matter?
Language technology opens up knowledge in other languages and is crucial for smaller languages such as Slovenian.
A common misconception
That a fluent translation is also an accurate one. A smooth sentence can change the meaning, tone or legal content of the original.
In more detail
Is NLP older than LLMs?
Yes. Working with language is a broad field with a long history.
What does it mean for a model to recognise intent?
From your words, it estimates what you want — for example to make a booking or get an answer to a question.
Do all languages work equally well?
No. Quality depends on how much varied training data there is, on the language and on the task.
How do I check a translation?
For an important text, use a native speaker or an expert, and check names, numbers and legal terms.
Remember: A fluent translation can still change the meaning.
Reliable sources
Last reviewed 1 November 2026. An explainer, not professional, legal or medical advice. For rules and figures that change, check the official version.