Module 1 · 20 min

What is AI?

Definitions without jargon: AI, ML, LLM. What a model does and doesn't do.

You'll learn

Understand the AI / ML / LLM distinction
Know what a "token" is and why it matters
Recognize when a model is "hallucinating"

Modern AI, in short: programs very good at continuing text. That's it. The rest (answering, coding, summarizing) are applications of this fundamental ability.

AI vs ML vs LLM

The terms nest inside each other like Russian dolls:

┌─────────────────────────────────────────────┐
│  AI (any system that "seems" intelligent)   │
│  ┌───────────────────────────────────────┐  │
│  │  ML (learns from data)                │  │
│  │  ┌─────────────────────────────────┐  │  │
│  │  │  Deep Learning                  │  │  │
│  │  │  ┌───────────────────────────┐  │  │  │
│  │  │  │  LLM ← we are here        │  │  │  │
│  │  │  └───────────────────────────┘  │  │  │
│  │  └─────────────────────────────────┘  │  │
│  └───────────────────────────────────────┘  │
└─────────────────────────────────────────────┘

AI is the broad category; LLMs are one specific technique inside it.

AI = any system that mimics a cognitive capability (a "smart" thermostat technically counts).
ML = a system that learns patterns from data (instead of being manually programmed).
LLM (Large Language Model) = an ML model trained to predict the next word in text. ChatGPT, Claude, Gemini are LLMs.

What is a "token"?

Models don't see words — they see "tokens": chunks of words. "Business" = 1 token, "Analyst" = 1 token. Rule of thumb: 1 English word ≈ 1.3 tokens.

What it can and CAN'T do

✓ Can: summarize, translate, write code, structure information, brainstorm.
✗ Can't: access the live internet (unless given tools), remember past conversations (without explicit memory), do serious math without a calculator.
✗ Can't: honestly tell you "I don't know". It will confidently make things up — this is called hallucination.

Mini-check

Which of the following is NOT an LLM?

Select an answer.