Natural Language Processing (NLP)
The branch of AI focused on enabling computers to understand, interpret, and generate human language. Powers chatbots, translation, sentiment analysis, and text summarization.
Why it matters
NLP is the reason machines can finally handle everyday human language instead of rigid commands. It powers the tools you use daily: autocomplete, spam filters, voice assistants, and chatbots. It matters because it bridges the gap between how people naturally speak and how computers process information. Without NLP, you'd still be clicking through menus instead of simply asking for what you want in plain words.
A concrete example
You type 'find cheap flights to Rome next month' into a travel app. NLP breaks that sentence apart, recognizing 'Rome' as a destination, 'next month' as a date range, and 'cheap' as a price preference. The app returns relevant results without you filling in a single dropdown. That everyday convenience, understanding a messy human sentence and acting on it, is NLP quietly doing its job in the background.
How to use it
Most of what used to require a dedicated NLP pipeline — classification, entity extraction, sentiment, summarisation — can now be done by prompting a general model, which is why so much bespoke NLP tooling disappeared. The trade-off is worth knowing: a purpose-built classifier is cheaper and more predictable at very high volume, while a prompted model is far faster to build and adapts without retraining.
The common mistake
Reaching for a general model for a simple, enormous-volume task. Classifying ten million short strings is a job for a small dedicated model; using a large language model for it is slower and costs orders of magnitude more for no accuracy gain.
Related terms
Large Language Model (LLM)
An AI model trained on vast amounts of text data that can generate, summarize, translate, and analyze human language. Examples include the GPT, Claude, Gemini and Llama families.
Inference Cost
The computational expense of running a trained AI model to generate outputs. Measured in cost per token or per request. Varies dramatically between models — GPT-4 class models cost 10-50x more per token than smaller models.
Structured Output
AI model responses formatted in a specific schema like JSON, XML, or tables rather than free-form text. Essential for integrating AI into software pipelines where downstream systems need predictable, parseable data formats.
Artificial Intelligence (AI)
The simulation of human intelligence by computer systems, including learning, reasoning, and self-correction. Modern AI is primarily powered by machine learning and neural networks.
Context Window
The maximum amount of text (measured in tokens) an AI model can consider at once. Larger context windows allow the model to reference more information in a single conversation.
Prompt
The text instruction you give to an AI model to generate a response. Prompt quality directly impacts output quality — better prompts yield dramatically better results.