The AI Landscape

By the end of this lesson, you’ll understand the basics of how AI CLI tools work, why they matter for your career, and which learning path is right for you. No coding background required, and no math beyond everyday arithmetic — we start from zero.

Why this matters

The workplace is changing. Industry reports suggest that teams using AI tools see meaningful productivity gains, and people who can direct AI agents are increasingly sought after by employers. This isn’t about becoming a programmer — it’s about expanding what you’re capable of. By the end of this course, you’ll have a practical skill that very few people have developed yet.

The elephant in the room

Let’s be direct: AI is changing job markets. Some tasks are being automated. Some roles are shrinking. If you’re feeling anxious about that, you’re paying attention — not overreacting.

But here’s what the headlines miss: your domain expertise is your moat. AI can generate code, summarize reports, and analyze data. It cannot understand your patients, your research questions, your business context, or your users the way you do. The people being displaced aren’t the ones who understand their domain deeply — they’re the ones who can’t adapt their tools.

That’s what you’re doing here. You’re not learning to code. You’re learning to operate new machinery — machinery that amplifies what you already know. The question isn’t whether to adapt, it’s how fast.

💬If you're here because you're worried about your job

That’s a valid reason to be here. Plenty of people in this course are in the same position — scientists watching grants shrink, analysts seeing dashboards auto-generated, developers watching junior roles disappear. Acknowledging the fear is step one. Building the skills to stay relevant is step two. You’re already on step two.

Which kind of AI work this is

“Working in AI” covers several kinds of work, and a lot of the anxiety comes from confusing them. Three matter for placing this course:

Who	What they do	What the work usually asks for
Researchers	Invent new models and new ways to train them	Deep math and a research setting
ML engineers	Build, evaluate, deploy, and maintain machine-learning systems — sometimes training or tuning the models themselves	Statistics, data pipelines, and often sizable data and computing budgets
People who direct finished models	Use existing models to build and fix real things	Clear thinking and their own domain knowledge

This course is the third row, and it never trains a model. The way of working it teaches for that row has a name — orchestrator — and Module 2 defines it.

What are AI CLI tools?

In a standard chat window, AI tools have limited direct access to your local project files — you typically need to copy-paste or upload content manually. AI CLI tools are different — they run inside your Terminal and work directly with your project files:

Read your existing code and documents
Create new files and folders
Edit multiple files at once
Run commands, tests, and builds
Fix errors and try again automatically

You describe what you want in plain English, and the AI builds it right inside your project — no copy-pasting between windows. Because these tools can modify files and run commands, always review proposed changes and use version control (you’ll learn Git in Module 7).

The mental model

Think of your project folder as a workshop. A chatbot is like calling a friend on the phone — they give advice but can’t touch your tools. An AI CLI tool is that friend standing in your workshop, picking up tools and building alongside you.

The core loop: Describe → Build → Review

Every interaction follows the same three-step pattern:

Describe what you want — you can start in plain English; advanced controls are introduced later.
The AI builds it — reading your project, writing code, running tests.
Review and refine — check the result, ask for changes, iterate.

Your primary job is to think clearly, give good instructions, and verify results. Coding skill becomes a force multiplier, but it’s not required to get started.

💡This is a skill, not magic

Anyone can type a question into ChatGPT. Directing an AI agent to build real software takes practice — knowing what to ask for, how to break down problems, and when to course-correct. That’s the skill this course teaches.

See it in action

Here’s what an actual AI CLI session looks like. Scroll through this transcript — a user launches Claude Code and builds a complete contact form through conversation:

Claude Code — AI CLI Demo

/home/user/my-project $ claude

Claude Code v1.0 /help for commands, Ctrl+C to exit >

/home/user/my-project $ I need a contact form for a small business site. Name, email, and message fields. Dark theme, clean and modern.

I'll create a contact form for you. Created: contact-form.html (47 lines) Created: styles.css (38 lines) The form includes: - Name, email, and message fields - Client-side validation - Responsive dark theme - Submit button with hover effect Open contact-form.html in your browser to see it.

/home/user/my-project $ Now add email validation and a success message that appears after submission

Updated: contact-form.html Added: - Email format validation - Green success banner after form submit - Form fields clear after submission - Smooth fade-in animation Refresh your browser to test the changes.

/home/user/my-project $

Notice what didn’t happen: the user didn’t learn HTML, didn’t install frameworks, didn’t read documentation. They described what they wanted in plain English and refined it with follow-up requests. That’s orchestration. (Use the tool switcher above to see how different tools handle the same task — all four are full agentic coding tools that read, write, and edit files directly.)

⚠What if the AI gives a wrong answer?

AI tools sometimes produce incorrect code, made-up facts, or results that look right but have subtle bugs. This is normal. Always test what the AI builds — run it, click through it, check the output. You don’t need to read every line of code, but you do need to verify the result works. We’ll cover verification strategies in detail starting in Module 8.

💡How to evaluate AI claims (including ours)

You’ll hear a lot of bold claims about AI — including in this course. A healthy habit: when someone cites a statistic or makes a prediction about AI, ask “where did this number come from?” and “is this still current?” AI capabilities and adoption numbers change fast. We’ll flag when information may go stale, and we encourage you to verify claims independently.

The four tools you’ll learn

As of September 2026, these are the leading AI CLI tools. This space moves fast — features and pricing change regularly, so always check the latest documentation.

Tool	Maker	What makes it special
Claude Code	Anthropic	Deep codebase understanding, long context, terminal-first
Antigravity CLI	Google	Succeeds the consumer Gemini CLI; huge context window (model-dependent), async multi-agent runs
Codex CLI	OpenAI	Autonomous task execution, agent-first architecture
Copilot CLI	GitHub / Microsoft	GitHub ecosystem integration, free tier with no credit card

You don’t need to pick just one. Effective orchestrators know which tool to reach for and when.

🧬In Your Field: Biotechclick to expand

Lab scenario: You have RNA-seq results in a CSV and need a heatmap. Instead of wrestling with R/ggplot2 syntax, describe the visualization to an AI CLI tool and get a working first draft — expect some iteration to refine labels, colors, and formatting. The Biotech track walks you through bioinformatics tools exactly this way.

📊In Your Field: MIS / Businessclick to expand

Business scenario: Your manager wants a dashboard showing quarterly sales by region. Instead of starting from scratch in Excel, describe the dashboard to an AI CLI tool and get an interactive prototype with charts and filters. For straightforward cases with clean data, this can save hours of work. The Business track shows you how.

🏛️In Your Field: Government / State Devclick to expand

Government scenario: Your team needs an internal tool to manage consular case intake. Instead of waiting months for IT, describe the tool and get a working prototype for internal review. Production deployment will still require security review and ATO, but you’ll have something to demo quickly. The Government Dev track covers building internal tools for government environments.

Where you stand right now

Using AI is no longer rare. Three professional developers in four already reach for it (SlashData, Q1 2026), so the tools themselves are not the advantage any more.

What is less well measured is systematic use — handing an agent a defined task, a place to run, and a way to check its own work. The figure below starts with everyone alive, then changes scale to show two dated proxies for it among developers: daily agent use in Stack Overflow’s 2025 survey, and use of one named terminal agent in JetBrains’ 2025 survey. Neither measures this course’s exact multi-session terminal workflow, and both were fielded over a year ago. Watch for the moment the scale changes: every professional developer on earth is about eleven dots on the first grid.

ℹYou're closer than you think

By the end of this course, you’ll be among the early adopters who don’t just use AI, but direct it to build things. That’s a meaningful advantage in any field.

2,500 DOTS = 8.2B HUMANS · EACH DOT ≈ 3.3M PEOPLE

1 / 7

8.2B

All humans alive

8.2 billion people (UN, 2024)

SPACE / ARROWS

🔧

Common first-timer concerns

Use the Symptom → Evidence → Request pattern: describe what you see, paste the error, then ask for a fix.

Symptom

I don't know what a 'terminal' or 'CLI' is

Evidence

You see references to 'command line' and 'terminal' and feel lost

What to ask the AI

"That's perfectly normal. Module 3 teaches you the terminal from absolute zero. For now, just know it's a text-based way to talk to your computer — and we'll walk you through every step."

Symptom

I'm worried I'll break something on my computer

Evidence

You've heard that terminal commands can be dangerous

What to ask the AI

"AI CLI tools work inside project folders, not your whole computer. We'll teach you safe habits in Module 3, and everything you build in this course is in isolated folders you can delete anytime."

Choose your path

Not everyone starts from the same place. Three questions to figure out where to jump in:

1. Have you ever opened a terminal?

No → you’re in the right spot. Continue to the next lesson and work through Modules 1–11 in order. Everything is explained from zero.

2. Comfortable in a shell but new to AI CLI tools?

Yes → consider the developer track. It moves faster through the basics and spends more time on agentic patterns.

3. Already running Claude Code or Codex daily?

Yes → Modules 1–11 will read as review. Go straight to Module 20: The Prompting Playbook and Module 19: Agentic Engineering, which cover advanced prompting technique, then MCP servers, skills and hooks, multi-agent orchestration, headless/CI pipelines, and spec-driven development.

💡You can always come back

Even experienced practitioners find value in the shared modules — especially Module 7 (Git), Module 8 (Understanding What AI Builds), and Module 11 (Ship It). Skim the objectives and skip what you already know.

Key takeaways

AI CLI tools run in your terminal and can read, create, edit, and run your project files directly
The core loop is Describe → Build → Review — plain English in, working software out
Four tools (Claude Code, Antigravity, Codex CLI, Copilot CLI) — you’ll learn when to use each
No coding or math prerequisite — the skill is clear thinking, not syntax
You’re joining a small group of early adopters who direct AI agents to build real things

KNOWLEDGE CHECK

What makes AI CLI tools different from regular AI chatbots like ChatGPT?

Next step: Head to the next lesson to have your first real conversation with an AI tool.

What you'll learn