Lesson 03 — Ways to use AI

Chat · API · CLI · Code & Agents — four doors into the same building.

AI for Dummies Module 3

By the end of this lesson you’ll know which door to open for any task — and why beginners should start at the front door and work outward.

Four doorways — one building

All four paths reach the same AI model. The door you pick determines who drives.

The four doorways — click to explore

☝ Click a card above to see details

Doorway 1 — Chat (the front door)

Chat in action — a real 3-turn conversation

Watch how a follow-up message refines the answer without starting over.

Doorway 2 — the API (kitchen window)

When you actually need the API

API in code — the canonical curl call

This is the actual Anthropic API call, straight from the documentation. Every library wraps exactly this.

Wrap this in a for loop → process your entire spreadsheet automatically.

curl https://api.anthropic.com/v1/messages \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "claude-sonnet-4-6",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Explain what an API is in one sentence."}]
  }'

The Python equivalent — 7 lines

Install once, then call Claude from any Python script. This is a complete, working program.

# Install the library (one-time)
pip install anthropic

import anthropic

client = anthropic.Anthropic()
# reads ANTHROPIC_API_KEY from env

message = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=1024,
    messages=[{
        "role": "user",
        "content": "Write a 2-line welcome email."
    }]
)

print(message.content[0].text)

The API key — your password and credit card

An API key is a long secret string. Every call made with it is billed to your account — even if someone else makes the call.

Doorway 3 — CLI (the terminal)

The terminal is a text window where you type commands. It looks scary. It’s just… typing.

ollama list          # installed models
ollama pull mistral  # download model
ollama serve         # run as a server

Gentle CLI intro comes in Lesson 04 when we install Ollama. You don’t need to touch it today.

Doorway 4 — Agents (AI that acts)

Chat answers. An agent executes. Give it a goal and step through how it works.

Agent vs Chat — the structural difference

Agent example — Claude Code session

Claude Code is Anthropic’s coding agent. It runs in your terminal, reads your project, and edits files. Here’s what a session looks like:

All four doorways compared

Cloud vs local — cost at scale

Cloud API costs money per token. Local models (Ollama) are free per call after setup. Here’s what 10,000 short calls looks like:

~50 input + 50 output tokens per call. Prices illustrative — check provider docs for current rates.

The beginner path — start at Chat, branch when needed

Lesson 03 — key takeaways

• Chat — browser window, zero setup, start here. Handles ~80% of everyday AI needs.
• API — your code sends the prompt automatically; the doorway for automation, apps, and batches.
• CLI — the terminal; type ollama run llama3.2 for a local, free, private AI session.
• Agents — AI that plans multi-step actions on real files; powerful, needs your oversight.