Course › Lesson 2
🔒 Full course lesson ⏱ 9 min
Lesson 2: How chatbots like ChatGPT work
A chatbot is the world's most well-read autocomplete.
🧸 The simple picture
You know how your phone suggests the next word when you text? Type "Happy" and it offers "birthday."
Now imagine that autocomplete read a gigantic library: millions of books, websites, and articles. It got so good at guessing the next word that it can write whole paragraphs, answer questions, and keep up a conversation. That's basically what ChatGPT, Gemini, Claude, Grok, and other chatbots are.
🔑 Key ideas
- Chatbots run on a large language model (LLM). "Large" because it learned from a huge amount of text. "Language model" because it models how words usually go together.
- When you ask a question, it builds an answer one small piece at a time, each time picking a likely next word.
- It doesn't look up a single "right answer" in a database. It writes a fresh answer every time, which is why asking twice can give two different replies.
- Some chatbots can also search the web, read files you upload, or look at pictures. Without those tools, it only knows what it learned during training, up to a certain date.
- It sounds human because it learned from human writing, but it is not thinking or feeling the way a person does.
🛠️ Try it (2 minutes)
Open any free chatbot (ChatGPT, Gemini, Claude, Copilot, or Grok). Ask it the same question twice in two new chats:
Give me a fun name for a pet goldfish.Notice the answers are different? That's the "fresh guess every time" idea in action. Then ask:
Explain how you came up with that answer, in simple words.
📝 Quick quiz
Tap an answer to check it.
1. A chatbot is most like…
✅ It predicts likely next words, over and over, very well.
2. Why can you get a different answer if you ask the same thing twice?
✅ Each answer is generated new, with a little built-in variety.
3. What does "LLM" stand for?
✅ Large Language Model: a big program that learned how language works.