AI glossary

LLM

An LLM, or large language model, is the type of AI model that understands and generates text. It's the engine behind chatbots and assistants like ChatGPT and Claude, trained to predict language and use that to answer questions.

An LLM doesn't have a database of facts. It has learned patterns in language. That's why it can answer and reason fluently, and also why it can confidently say something that isn't true. That's not a defect but a characteristic of how it works. It predicts which word is most likely to come next.

So for business use, you combine an LLM with your own data. That way it answers based on facts that are correct, not just on what sounds likely.

In practice.

The AI applications I build run on current models from Anthropic, grounded in the client's own knowledge base. That way you get the power of the model without letting it guess.

Frequently asked questions.

What is an LLM?

A large language model is an AI model that understands and generates text. It's the engine behind assistants like ChatGPT and Claude, trained to predict language and use that to answer questions.

Can an LLM remember facts?

Not reliably. An LLM has learned patterns in language, not a database of facts. That's why, for business use, you connect it to your own data, so it answers based on facts that are correct.

Related terms.

Want to see how this works in practice? Have a look at AI assistant and AI receptionist.

Curious what this could do for you?

Walk me through one process. In half an hour I will show you where this kind of work gets stuck in your business, and what it would save to hand it off.

Walk me through one process