Natural Language Processing (NLP)
The field of getting computers to understand and generate human language — the decades-old discipline whose task-specific methods collapsed, in a few years, into a single general model.
Reviewed July 15, 2026Stable
When not to use it
- As a synonym for large language models. NLP is the field; an LLM is the current technology within it, not the whole discipline.
- For tasks better framed as pure information retrieval or structured-data queries, where language modeling adds cost without benefit.
- When a simple deterministic method (a regex, a lookup) solves the problem, reaching for a language model is overkill.
Reach for something else instead
- Computational linguistics overlaps heavily but leans toward the scientific study of language rather than building applications.
- Information retrieval handles finding relevant documents, a related but distinct problem from understanding their content.
- Speech processing covers the audio side, often treated alongside NLP once speech is transcribed to text.
Read more on the blog
- What linguistics predicts about where AI failsLinguistic features stopped improving models around 2019, because models learn the structure themselves. The categories survived anyway, as the best available map of where these systems break.
- What is natural language processing (NLP)?Every search, translation, voice assistant, and chatbot runs on natural language processing, the field of getting computers to work with human language. Its defining modern story is a quiet revolution: the dozens of separate, task-specific methods that made up NLP for decades collapsed into a single general approach, the large language model.
- Why AI answers the question you asked, not the one you meantAsk a model whether it can do something and it may tell you rather than doing it. The gap between literal meaning and intended meaning has a name, a fifty-year literature, and a pattern in where models fail.
Further reading
- Jurafsky & Martin, Speech and Language Processing — the standard textbook covering the field's tasks and history.
- Vaswani et al. (2017), Attention Is All You Need — the transformer that reshaped NLP into the language-model era.
- Devlin et al. (2019), BERT — pretraining that made one model transferable across many NLP tasks.
Primary sources, listed so you can check the claims on this page rather than take them on trust.
Where people go wrong
- Treating high benchmark scores as proof that language is understood or that NLP is solved.
- Assuming progress in English transfers to other languages; low-resource languages lag far behind.
- Forgetting the task vocabulary once one model does everything, which leaves you unable to specify or evaluate what you actually want.
At a glance
FieldLanguage & LLMs
Scopeunderstanding and generating human language
Erasrules → statistical → deep learning → transformers/LLMs
Defining shiftmany methods into one model
DifficultyBeginner → Intermediate
Where this sits
A starting point. Nothing needs to come before it.
0Levelstarting point
0Needs firstnothing
76Opens up25% of 310
7Areascrosses fields
Computed from the prerequisite graph, not assigned. How this works