If you've spent any time reading about AI over the past couple of years, you've encountered the term "large language model" or its abbreviation, LLM. It's used as though everyone already knows what it means, which leaves a lot of business owners nodding along without a clear mental model of what's actually being described. That mental model matters; because understanding what an LLM is tells you a great deal about what AI can do and what its limitations are.
What "language model" means
A language model is a type of AI that has been trained to understand and generate text. It learns by processing enormous amounts of written material; books, articles, websites, code, conversations; and identifying the statistical patterns that govern how language works. After processing enough text, it can generate new text that follows those patterns: answering questions, continuing a thought, writing in a particular style, summarizing a document, or translating between formats.
What this means practically
Because LLMs are fundamentally language systems, they're remarkably good at anything that involves reading, writing, understanding, or generating text. Drafting communications. Summarizing documents. Answering questions. Engaging in natural conversation. These are the things LLMs do well, because these are exactly what they were built to do.
They're less reliable for precise factual recall (they can "hallucinate"; generate plausible-sounding but incorrect information) and real-time information (they have training cutoffs and don't know about recent events without additional tools).
Why this matters for how you use AI
Understanding that AI is fundamentally a language system helps you use it more effectively. Tasks that involve language; writing, summarizing, organizing, communicating; are where AI will consistently add value. Tasks that require precise factual accuracy; financial calculations, legal citations, medical information; require human review of AI output before acting on it. This is not a limitation that makes AI less useful. It's a characteristic that helps you deploy it intelligently.