Large Language Model (LLM)
An AI system trained on massive amounts of text that can understand and generate human language.
What It Is
A large language model is the engine behind tools like ChatGPT, Claude, and Gemini. It’s a piece of software that has been trained on enormous amounts of text, from books and websites to code and conversations, until it learned patterns in how humans use language. When you type a prompt, the model predicts the most likely helpful response based on everything it absorbed during training. “Large” refers to both the size of the training data and the billions of internal parameters the model uses to make predictions. These models don’t think the way humans do. They are very sophisticated pattern-matching systems that produce remarkably useful output.
Why It Matters
LLMs are the foundation of nearly every AI tool you will use as an operator. Understanding that they are pattern-based (not reasoning-based) helps you set realistic expectations. They excel at drafting, summarizing, translating, coding, and brainstorming. They struggle with math, real-time facts, and tasks that require true understanding. Knowing what an LLM actually is helps you use it as a tool rather than treating it as an oracle.
In Practice
When you open Claude or ChatGPT and type a question, you are sending a prompt to an LLM. Choosing which model to use (Claude Opus vs. Haiku, GPT-4o vs. GPT-4o mini) means choosing different LLMs with different tradeoffs in speed, cost, and quality. Operators pick the right model for the job, not always the biggest one.