How it works
An LLM (large language model) reads tons of text and learns one skill: given the words before, predict the next word.
Do that billions of times and the model learns grammar, facts, style, even simple reasoning patterns.
Why "large"
More text + more parameters = better predictions. The word LLM simply refers to models at a scale where this works well.
What this means for you
- Ask in full sentences, not keywords.
- Give context: it has no memory of your life.
- Treat its answers as drafts, check the facts.