LLM (Large Language Model)
An LLM (large language model) is an AI that has learned from a vast amount of text. It can understand, summarise and rewrite text.
Explained simply
An LLM is the core of tools like ChatGPT, Microsoft Copilot or Claude. It was trained on huge amounts of text. From that it learned which words typically follow each other.
When you ask a question, the model works out a fitting answer step by step. It does not "know" anything in the human sense. It writes what fits best according to its training data.
That is why answers often sound convincing, even when they are wrong. Such invented answers are called hallucinations.
An example: a hotel in Tyrol has replies to guest enquiries drafted by an LLM. That saves time. But a person still checks room rates and availability.
What does this mean for your business?
LLMs are well suited to text: drafts, summaries, translations. A person should always check figures, facts and legal statements. It also matters which data you enter. Settle that before your teams start working with LLMs.
Frequently asked questions
Is ChatGPT an LLM?
Which LLM is best for which task?
How JJW helps
AI in businessIn our AI training, your teams learn how LLMs work and where their limits are.
See AI training →