THE BIG DUMMY'S
GUIDE TO A.I.

Short answer: it's a machine you talk to in plain English, and it talks back.

You've already met its cousins. The bank's phone robot. The GPS lady. Autocomplete on your phone. AI — artificial intelligence — is the new model year of all that, except this one actually understands what you mean, holds a real conversation, and can do useful work: write a letter, explain a medical bill, plan a trip, untangle a recipe, or just keep you company on a slow afternoon.

Under the hood, an AI "chatbot" is a computer program that read more books, newspapers, and websites than any human ever could, and got real good at one thing: figuring out what words come next. That turns out to be enough to answer questions, draft letters, and explain darn near anything — like a know-it-all neighbor who's usually right, works for free, and never gets tired of your questions.

The secret nobody tells beginners: there is no wrong way to talk to it. You can't break it. It won't laugh at you. If you can ask a question out loud, you're qualified.

In CB days you needed a radio, an antenna, and twenty feet of coax. For AI you need…

That's the whole shopping list. No license, no call sign, no antenna on the roof. The FCC has no opinion about you whatsoever.

Five minutes, start to finish. Here's the whole thing:

1. Open your web browser — Safari, Chrome, Edge, whichever blue or colorful circle you normally tap to look things up.
2. Go to one of the big three. Type claude.ai, chatgpt.com, or gemini.google.com into the address bar. They're all good; they all work the same way. (Can't decide? Flip a coin. You can try the others tomorrow — they're free.)
3. Make a free account — an email address and a password, same as anything else. No credit card.
4. Find the message box and say something real. Not a test — a real question from your actual day: "Explain this letter from my insurance company in plain English." "Give me a week of simple dinners for two." "Help me write a birthday note to my granddaughter who loves horses."
5. Talk back. This is the part nobody does on day one: when the answer comes, reply to it. Say "shorter," or "explain that one part again," or "make it funnier." It remembers the conversation. Going back and forth is the whole skill — and now you have it.

That's it. That's the license test. You passed.

CB had "10-4" and "smokey." AI has its own slang. Here's the translation card:

CB had courtesy rules. So does this. Five of them, and the first two are the ones that matter:

• Never tell it your secrets. No Social Security number, no bank passwords, no card numbers — same rule as a phone call from a stranger. It doesn't need them, and anything that asks for them isn't the real thing.
• Trust, but verify. AI is confidently wrong sometimes (that's the "hallucination" from Chapter Four). For anything that touches money, health, or the law, treat its answer as a smart first opinion — then check with a human professional or an official source.
• It's a tool, not an oracle. If an answer smells off, say so: "Are you sure? Double-check that." It will often correct itself on the spot.
• Plain English beats computer-talk. You don't need magic words. Talk to it like a sharp new assistant on their first day: say what you want, who it's for, and how long it should be.
• Watch for imposters. The real services are the addresses in Chapter Three. Ads, pop-ups, and emails offering "AI access" for a fee are running a scam on the newness. When in doubt, type the address yourself.