Home / Guides / How AI works, in plain words

How AI Chatbots Work, in Plain Words

A chatbot has read an enormous amount of text and learned patterns of language. It predicts a helpful reply one piece at a time.

Step by step

  1. It learned from text. Books, websites and more.
  2. It predicts the next word. Over and over.
  3. It does not look things up by default. Some tools can search.
  4. It can sound sure and be wrong.
  5. You steer it with your question.

What to expect

Expect an image: a very well-read assistant that predicts helpful words. It does not look up facts unless it has a search feature, and it does not truly understand your life.

Mistakes to avoid

Do not assume confidence means accuracy. Avoid sharing private information. Do not treat it as an expert in law, medicine or finance.

A good next step

Try asking the same question two different ways and notice how the answer changes.

A prompt you can copy

Type this into ChatGPT, Claude or Gemini

Explain in three short paragraphs how a chatbot like you works, using an everyday analogy, for someone who has never used one.

Good to know. Think of it as a very well-read assistant who sometimes guesses.

Want to try this with a friendly teacher?

Join a free live class on Zoom. Small group, no experience needed, no sign-up fee.

Reserve a free seat

Common questions

Does it think?

It produces text; whether that is thinking is debated.

Does it learn from me?

Not in the moment, though some services use chats to improve.

More guides