Your first 30 minutes with AI: where to start

Time: 30 min (including sign-up)

Who this lesson is for

This lesson is for you if you:

• Have never opened ChatGPT or Claude
• Aren't sure what "chatting with AI" even means
• Worry you'll "break something on the computer"
• Don't know where to get AI or what it costs
• Keep hearing about AI but have never tried it
• Are a grandparent, a manager outside tech, or any other adult who's just starting out

If you've already spent at least an hour talking to AI, skip ahead to the next lesson, AI without fear.

The gist

Opening AI for the first time is like a 90-year-old making her very first video call. It feels scary. It seems complicated. It seems like you need to know "something special." In reality, it's three buttons.

In 30 minutes you will:

1. Sign up at claude.ai (you'll need an email address and a cell phone that can receive text messages)
2. Ask your first 3 questions
3. See 7 everyday tasks AI handles in a minute
4. Learn what not to do
5. Decide where to go next

🎨 Picture this: you're learning to ride a bike. No theory about gears, aerodynamics or the physics of rotation. You get on, push off and ride. Falling isn't the end of the world; a bruise heals. AI works the same way.

What AI actually is (in one sentence)

AI is a conversation partner that has read almost everything written on the internet and can now answer your question. Like a very well-read consultant, except it's free and available around the clock.

It can:

• Explain complicated things in plain words
• Write texts (emails, posts, articles)
• Translate from almost any language into another
• Do math, make plans, come up with ideas
• Make sense of your documents and photos

It can NOT (in a regular chat, without connected apps and permissions):

• Make phone calls for you
• Buy groceries
• Remember you tomorrow if you're not logged in
• Guarantee it's right (it sometimes makes things up, so check important facts)

Step 1: Go to claude.ai (5 minutes)

Link: https://claude.ai

What to do:

1. Open any web browser (Chrome, Safari, Edge, any of them)
2. In the address bar at the top, type claude.ai and press Enter (on a phone, tap Go on the keyboard)
3. There's no separate "Sign up" button: new and returning users start on the same page. Type your email into the "Enter your email" box and tap "Continue with email". Any address works: Gmail, Outlook, iCloud, Yahoo. If you have a Google or Apple account, you can tap "Continue with Google" or "Continue with Apple" instead
4. You don't need to make up a password: Claude doesn't use passwords. Instead, you'll get an email with a login link (the sender's address ends in @mail.anthropic.com). If it doesn't show up, check your spam folder
5. Open that email on the same device and tap the link, and you'll land in Claude. If you open the email on a different device (you started on a computer but read email on your phone), the link shows a code: type it in where you started
6. As a new user, you'll be asked to verify your phone number: enter your cell number, wait for the text message with a code, type the code in and tap "Verify code." You can't skip this step, and landlines and internet (VoIP) numbers won't work
7. Answer the couple of questions Claude asks (for example, what to call you) and keep tapping the button to continue
8. Done. You're in.

Cost: $0 (zero dollars). The free plan works without a credit card, without a subscription, without a catch.

Age. Claude is for adults: under Anthropic's rules, you must be at least 18 to use it.

Claude doesn't work in every country: the official list is at anthropic.com/supported-countries, and signing up takes a phone number from a supported country. If Claude isn't available to you, for example while you're traveling abroad, Steps 2 to 4 of this lesson (first questions, everyday tasks, safety rules) work in any other assistant; only the sign-up is different. See the comparison in Comparing AI models, and current plans and supported countries on the What's current page.

🎨 Picture this: it's like signing up for any new app: your email, a link in your inbox, a code by text, and you're in. That's it.

What you'll see on the screen

Claude's main screen is simple:

• The message box: in the middle or at the bottom of the screen, next to a greeting. That's where you type your question
• On the left: a list of your past conversations (empty for now). On a phone, it's tucked behind the menu icon in the corner of the screen
• Bottom left: your name or initials. Tap them to open Settings

No complicated menus. It's a chat, like texting a friend, except there's only one person on the other end (the AI).

Step 2: Your first 3 questions (10 minutes)

The main rule: write the way you'd text a friend or a nephew. Don't try to "phrase it perfectly." AI understands everyday language.

Question 1

Hi. What can you do?

Just type that into the box and tap the arrow button next to it (on a computer, you can simply press Enter).

Claude will answer in the language you wrote in. Read the whole answer. It gives you a sense of what it can do, in its own words.

Question 2

Explain in plain English what a blockchain is

(or pick any topic you've always wanted to understand: quantum physics, mortgages, Bitcoin, how AI works, the stock market)

Claude will explain it without jargon. If it's still unclear, type: "Explain it even more simply, like I'm in fifth grade." It will rephrase.

Question 3

Translate into Spanish: Hello, how do I get to the farmers market?

(or into French, Chinese, any language)

Claude translates instantly. You can follow up with: "How do I pronounce that?" and it will spell out the pronunciation.

Congratulations. You just used AI. That's all there is to it. Everything after this is just more kinds of tasks.

Step 3: 7 everyday tasks for your first few days

Here's what people typically do in their first weeks. Try one task a day, and in a week you'll be a confident user.

Tip: don't try to do everything in one day. One task a day is a good pace. In a week, that's 7 wins.

Step 4: What NOT to do in your first 30 minutes (or ever)

The list is short, but it matters.

❌ Never give AI your passwords, credit card numbers, Social Security number, bank account details or PINs. Never. Even if it asks. AI has no reason to know them. If someone "on behalf of AI" asks for your card details, it's a scam.

❌ Don't treat medical advice as 100% reliable. AI can explain what aspirin is, but a doctor makes the diagnosis. Use AI to prepare questions for your doctor, not to replace your doctor.

❌ Don't treat legal advice as 100% reliable. AI can give you the general idea of a law, but a lawyer goes to court for you. For any document you sign, talk to a lawyer.

❌ Don't panic if AI gets something wrong. It isn't perfect. Sometimes it makes up facts (this is called a "hallucination"). Double-check important numbers and dates with a Google search or with a real person.

❌ Don't worry about "breaking" it. There's no red button that deletes the internet. You can click any button in Claude's interface and nothing will break. The worst that happens is you close a chat and open a new one.

❌ Don't give AI documents with other people's personal information (a friend's ID, a relative's medical records). That violates their privacy.

Step 5: Plans: Free, Pro, Max, API

Once you like it, the question comes up: should you pay? Here's an honest comparison.

Free: $0/month

• Limits: the number of messages is limited, and the limit resets every five hours; the exact number depends on demand and on how long the conversation is. For getting started, it's usually enough
• Model: a standard one (still smart, just not the smartest); which one exactly, see What's current
• Good for: most people who are just starting
• Not a fit: if you want to work with AI 4 to 8 hours a day

🎨 Picture this: the bus. It goes to the same places a cab does, just with stops.

Pro: $20/month (as of October 2026; less if you pay yearly)

• Limits: noticeably higher than on Free
• Model: access to stronger models in the Claude family than on the free plan
• Bonus: easier to work with large documents (PDF, Word, Excel) thanks to the higher limits
• Bonus: access to Claude Code (that's for people who build software)
• Good for: students, professionals, active users

🎨 Picture this: a cab. You ride whenever you want. More convenient. Costs more.

Max: from $100/month (as of October 2026)

• Limits: several times higher than on Pro (there are 5x and 20x tiers)
• Model: the same models as Pro, with a lot more room to use them
• Good for: programmers, researchers, copywriters, people who work with AI 8 hours a day
• Not a fit: regular users (it's overkill)

🎨 Picture this: a private driver. Always on hand. Expensive. But if you're on the road all day, it pays for itself.

API: pay as you go

• Price: charged per million tokens and depends on the model. A token is a small chunk of text: by Anthropic's rough estimate, about three-quarters of an English word. Current prices: What's current
• This is for developers who build AI into their own apps
• As a regular user, you don't need it. Ignore it until you start building software.

🎨 Picture this: renting a car by the hour. You only pay for what you actually use. But you need a "license," meaning you need to know how to program.

Current plans: https://claude.com/pricing and What's current

Tip for beginners: start with Free for 2 weeks. If you keep hitting the limits, move to Pro. You don't need Max unless you work with AI for many hours every day.

Step 6: What to do after these 30 minutes

You choose where to go next. Three options:

Option A: "I want to understand what this actually is"

Go to the lesson AI without fear. It covers all the common worries ("AI will take my job," "AI is watching us," "AI will trick me").

Then take How an LLM works inside (an LLM is a language model, the engine behind any AI assistant), followed by The history of AI, which explains how we got to 2026, what came before Claude and who made it. That's the same order as the Use AI in my work path on the course page.

Option B: "I want to use it at work right away"

Open the Professions section. It covers 100 jobs: what AI changes, what to learn and which tools to use (accountant, teacher, doctor, designer, salesperson, lawyer and more).

Find your job, read that one page, and try one example today.

Option C: "I want to build systems with AI"

It's a long road, but a doable one. On the course page, pick the Build my own product path: it starts with these same first lessons and leads, step by step, to the lesson Installing Claude Code. Claude Code is a program in which AI writes and edits code for you; you don't need to install it right away. Get the basics down first.

A word of caution: option C takes real effort. It isn't "30 minutes and I'm a developer." It's months of practice, and everyone's timeline is different.

Your first-30-minutes checklist

Go through the list. If any box isn't checked, go back to that step.

• Signed up at claude.ai (email + the link from your inbox + the code by text)
• Saw the main screen (the input box at the bottom)
• Asked at least 3 questions ("What can you do?" + 2 of your own)
• Read Claude's answers all the way through
• Understand the difference between Free, Pro, Max and API
• Know what you must NOT give AI (passwords, card numbers, Social Security number)
• Know that important facts need checking (health, legal matters, numbers)
• Picked where to go next (A, B or C)

Common questions from first-time users

"If I close the browser, will I lose the conversation?" No. All your conversations are saved in your account, in the list on the left. Come back tomorrow and they'll be there.

"Does Claude remember me?" Within one conversation, yes, it remembers everything you've said. It also remembers some things between conversations: on the Free, Pro and Max plans (as of October 2026), memory is on from the start. As you chat, Claude saves short notes about you and what you're working on, and uses them in new conversations. You can read, edit or delete those notes, or turn memory off, in Settings → Memory.

"Do I have to write in perfect English?" No. Typos and casual phrasing are fine; Claude will understand you. It also works in many other languages, so you can write in Spanish or another language you're comfortable with.

"What if it gives me the wrong answer?" Just write: "That's not it. I meant [your clarification]." It will rephrase. You can correct it as many times as you like.

"Can I break something by clicking the wrong thing?" No. There are no dangerous buttons in Claude's interface. The worst you can do is delete an old conversation, and even then you'll get a "Delete this chat?" confirmation first.

"Do people at Anthropic read my conversations?" It depends on your settings. On Free, Pro and Max, Anthropic uses your chats to train its models only if the setting to help improve Claude is turned on. You can turn it off in the privacy settings; with it on, your data is kept for up to 5 years, with it off, for 30 days (as of October 2026). If you're worried, don't write anything there you'd want to keep private.

Key takeaways

AI isn't a complicated technology for programmers. It's a conversation partner in your browser at claude.ai. Free. No credit card. Five minutes to sign up, and you're a user.

The first 30 minutes are simple: sign up → 3 questions → understand the limits → choose where to go next. No magic. No terms you need to memorize. And if you run into a word you don't know, look it up in the Glossary.

The main rule: write to AI the way you'd write to a friend. No perfect wording, no fear of getting it wrong. AI is more patient than any person: it rephrases, it re-explains, and it never gets tired.

Next lesson

→ AI without fear: fears and myths about AI

Next on the Use AI in my work path: How an LLM works inside, then The history of AI.

Other options from Step 6: Professions (AI for your job) or Installing Claude Code (setup for future builders).