The interface you know
When you use Claude.ai or ChatGPT, you're using a web application. There's a text box. You type something in. The AI responds. You copy the output and paste it wherever you need it.
That interface is convenient, but it has a hard limit: everything passes through your clipboard. The AI can't reach into a folder on your computer, process 50 files in sequence, or hand its output directly to another tool.
What a CLI actually is
A CLI — command-line interface — is a text-based window for talking to your computer. Instead of clicking icons and menus, you type short commands and press Enter. The results appear as text.
On a Mac, this window is called Terminal. On Windows, it's Command Prompt or PowerShell. You've probably seen it in movies — the black screen with green text. That's the same thing, dressed up for dramatic effect.
The terminal isn't the AI. It's where you run the AI — and everything else on your computer.
A journalism analogy
Imagine you're researching a story and you have 200 interview transcripts saved as text files. Using a web browser interface, you'd open each one, copy the text, paste it into Claude.ai, ask your question, copy the response, and paste it somewhere. That's 200 copy-paste cycles.
With a CLI tool, you write one command that says: "Read every file in this folder, find mentions of [source name], and save a summary to a new file." The AI processes all 200 transcripts while you do something else.
What changes when you use a CLI tool
The underlying AI model is the same whether you access it through a browser or a terminal. Claude Sonnet is Claude Sonnet either way. What changes is what you can do with the output.
CLI tools can read files directly from your computer and write results back to disk. They chain together — one command's output becomes the next command's input. And they can be scripted: save a workflow once and run it again next week without retyping anything.
For journalists, this is the difference between using AI as a chat partner and using it as a tool that fits into how you already work.
What you don't need to know
You don't need to understand how computers work at a deep level. You don't need to write software. The commands in this course are short — usually one line — and you'll have a reference to look them up.
The main adjustment is mental: instead of clicking buttons, you'll be typing short instructions. The computer follows them exactly. If something doesn't work, the error message usually tells you what went wrong.
READY? Head back to Module 1 to install the tools and run your first CLI command.