What Is Generative AI?
First Steps with AI
From zero to your first AI conversation — what AI is, which tools exist, and where the limits are.
AI is not Google
Section titled “AI is not Google”When you use a search engine, it scans the web and shows you links to existing pages. Generative AI works fundamentally differently: it creates new text, word by word, based on patterns learned from billions of texts.
That’s the first paradigm shift: AI is not Google. It doesn’t find answers — it constructs them.
How does it work?
Section titled “How does it work?”Think of the autocomplete on your phone. It suggests the next word based on what you’ve typed so far. Generative AI works on the same principle — just at a completely different scale.
Instead of predicting from your messages, an LLM Large Language Model — a neural network trained on massive amounts of text that can generate new text. uses billions of books, articles, websites, and conversations as its foundation. The result: a system that can write entire essays, emails, or analyses — choosing the statistically most likely next word, one at a time.
Another analogy
Section titled “Another analogy”Imagine a student who has read thousands of textbooks but never attended a single class or lab. When you ask them a question, they write a confident, well-structured answer that sounds like a textbook. Most of the time it’s right — but sometimes they invent facts, because they’re reconstructing from patterns, not from understanding.
The building blocks: Tokens
Section titled “The building blocks: Tokens”AI doesn’t read and write in whole words but in Token The smallest text units an AI model processes — usually a word or part of a word. . A token is usually a word or part of a word:
- “apple” = 1 token
- “friendship” = 2 tokens (“friend” + “ship”)
- A typical page of text ≈ 300 tokens
Why this matters: every AI model has a limited “token budget” — the so-called Context Window The maximum amount of text an AI model can process at once — its short-term memory. . Think of it as the desk the AI works on. What doesn’t fit on the desk, it can’t see.
Why does AI give a different answer each time?
Section titled “Why does AI give a different answer each time?”If you ask the same question twice, you’ll often get slightly different answers. That’s not a bug — AI calculates thousands of possibilities for each next word and then picks one.
How adventurous it is depends on a parameter called temperature:
- Low (conservative): AI picks the most likely words → consistent, predictable answers
- High (creative): AI also picks less likely words → more surprising, varied answers
Think of it like a dial between “follow the recipe exactly” and “improvise in the kitchen.”
Try it yourself
Section titled “Try it yourself”- Open an AI tool of your choice (ChatGPT, Claude, Gemini, or Copilot — all have free versions).
- Ask the same question twice: “Explain in two sentences what Generative AI is.”
- Compare the answers. What’s the same? What’s different?
Think further
Section titled “Think further”Generative AI AI systems that can create new content — text, images, code, music — rather than just searching or classifying existing information. is a tool, not an oracle. It reflects patterns from human knowledge back at you — it has no thoughts, opinions, or beliefs of its own. This understanding is the foundation for everything that follows.
In the next lesson, we’ll look at the major tools: ChatGPT, Claude, Gemini, and Microsoft Copilot.