Free to read
Inside AI Without the Jargon
All 12 chapters: what each one opens on, the questions it answers, and the one small thing to do when you have finished it. About 21,500 words in the book itself.
Chapter 1 · about 1,500 words
AI in Plain English
It opens on Frank, who is seventy-four and spent his working life as a plumber, told us he'd avoided the whole subject for two years because of one word: algorithm.
What it answers
- If it just predicts words, how does it get anything right at all?
- Because the writing it learned from was mostly accurate, so the most likely next words are usually true ones. Most of the time, prediction and truth line up. The trouble is only in the gaps, where the likely-sounding words happen to be false. That's why it's right far more often than not, and also why you never fully drop your guard on specifics.
- Is one of these tools smarter than the others, should I hunt for the best one?
- Not in any way that matters for what you'll do. They trade the lead back and forth, and the differences are the kind only enthusiasts notice. Pick whichever is easiest for you to open, learn its feel, and don't spend a minute of worry about missing out. The skill you're building travels with you to any of them.
When you have finished it
Don't open anything yet. Just think of one thing you've explained to another person more than once in your life. How compound interest works. Why the good knife shouldn't go in the dishwasher. The route to your house. Anything. Hold that in mind. In the next chapter you're going to ask an AI to explain that same thing, and you're going to judge its answer as the expert you already are. That's the posture we want you in: not a student hoping for a passing grade, but a seasoned person checking an eager assistant's work.
Chapter 2 · about 1,300 words
Your First Useful Conversation
It opens on The blank box is the hardest part, and it's hardest for a reason that has nothing to do with age.
What it answers
- Do I have to pay for this?
- No. Every mainstream tool has a genuinely free version that does everything in this book. Paid tiers exist and buy small extras, but you never need one to start, and you should never hand over a card just to have a basic conversation.
- What if I type the wrong thing and break something?
- You can't. There's nothing to break. The worst outcome of a bad message is a bad answer, which you fix by asking again or closing the tab. Type freely, experiment, make a mess, it costs nothing and leaves no mark.
- Does it remember me between visits?
- If you make a free account and stay signed in, it can keep your past chats so you can pick threads back up. If you don't, each visit starts fresh. Both are fine; it's your preference.
When you have finished it
Have that one conversation. Ask your familiar thing, read the answer as the expert you are, then send exactly one follow-up to improve it. Shorter, plainer, warmer, whatever it needs. The goal isn't a perfect answer. The goal is to feel, once, the small click of talking back and being understood. Everything else in this book is built on that click.
Chapter 3 · about 1,300 words
The Five Parts of a Better Request
It opens on If the last chapter got you talking, this one makes you good.
What it answers
- I gave it all five parts and the answer's still off. Now what?
- Tell it exactly what's wrong and ask again, that's the sixth part nobody mentions. "Too formal, warm it up." "You missed that it's for my landlord, not a friend." The conversation is the tool; a disappointing answer is the middle of the process, not the end of it.
- Won't a long, detailed request annoy it or slow it down?
- The opposite. More detail gives it more to work with, and it never tires or judges. The people who get thin answers are almost always the ones who asked in five words. Err toward telling it too much.
When you have finished it
Take one request you'd genuinely like help with this week. A letter, a plan, an explanation, a list. Before you type it, add just two things you'd normally leave out: one piece of context about your situation, and one limit (a length, a tone, a thing to avoid). Send it. You'll feel the answer arrive with more edges than usual. That's the five parts working, and it only gets more natural from here.
Chapter 4 · about 1,300 words
Everyday Tasks AI Can Make Easier
It opens on Let's come down out of the clouds and into the kitchen, because this is where these tools earn their keep.
What it answers
- Can it actually do things for me. Book the appointment, send the email?
- Mostly no, and that's a good boundary to keep. A plain chat tool drafts and plans; you're the one who presses send, makes the call, or places the order. Treat it as the assistant who writes the letter and hands it to you, not the one who mails it.
- How do I get the answer out of the chat and into a real letter or document?
- You copy it. There's usually a small copy button, or you can select the text the old-fashioned way, then paste it into your email or a document. A grandchild can show you the copy-and-paste motion once, and after that the tool's work flows anywhere you need it.
When you have finished it
Pick the most boring, most postpone-able writing task on your plate, the duller the better. Describe it to the tool in a sentence or two, including who it's for, and ask for a first draft. Don't send whatever it gives you. Just notice how much easier it is to fix a draft than to face a blank page. That relief, repeated across a hundred small chores, is most of what these tools are actually good for.
Chapter 5 · about 1,200 words
Learning Almost Anything with AI
It opens on There's a specific kind of pleasure that these tools can give back to you, and it's one many people quietly lose track of somewhere in a busy life: the pleasure of understanding something you didn't understand before, at
What it answers
- Isn't this just cheating myself out of really learning?
- No more than a patient tutor or a good encyclopedia ever did. You're still doing the understanding; the tool just answers your follow-ups without sighing. The danger isn't leaning on it to learn, it's trusting a specific fact without checking. Understand freely; verify the details that matter.
- What if I ask something that sounds stupid?
- There is no such thing here, and that's the quiet gift of it. The tool has no memory of your last question, no raised eyebrow, no impatience. Ask the most basic thing you've been embarrassed to ask a person for years. That privacy is exactly what makes it such a good place to finally learn.
When you have finished it
Pick one thing you've genuinely always wanted to understand and never quite got. Ask the tool to explain it "simply, like I'm curious and capable but new to it." Then. And this is the whole exercise, ask one real follow-up question about the part that stayed murky. Feel the difference between being told an answer and having a conversation toward understanding. That difference is the door to learning anything you like, for the rest of your life, on your own terms.
Chapter 6 · about 1,400 words
Health, Money, and Legal Questions — Handled Wisely
It opens on Now we come to the questions that matter most, and the ones where the stakes are real enough that we're going to slow right down and be careful together.
What it answers
- Can I at least ask it what my symptoms might mean, just to prepare?
- You can ask it to explain what a condition generally involves, or what questions to bring. But never treat what it says about your body as a diagnosis. It hasn't examined you and can be confidently wrong. Use it to walk in informed; let the doctor tell you what's actually going on.
- What about my money, can it look at my situation and advise me?
- It can explain concepts and help you organize your questions, and that's genuinely useful. It cannot know your full picture or the current rules, and a wrong specific here costs real money. So: understand more, decide nothing, and take your sharpened questions to a person who's actually responsible for the advice.
When you have finished it
Think of one appointment or decision on your horizon. A check-up, a policy renewal, a form that's been sitting there. Ask the tool to help you prepare good questions for it, and notice that it's helping you get ready rather than deciding anything. That posture, informed, prepared, and still firmly in charge. Is exactly how a wise person uses these tools on the things that count.
Chapter 7 · about 1,100 words
Photos, Voice, and Other AI Tools
It opens on Everything so far has happened by typing.
What it answers
- Will it hear my accent, or if my voice shakes?
- Usually yes. These tools are trained on an enormous range of voices and handle accents and unsteady speech far better than the old voice systems did. If it mishears a word, you just fix that word by typing or saying it again. No penalty, no fuss.
- Is it safe to photograph a document and show it?
- For an ordinary confusing letter, yes. For anything carrying account numbers, your Social Security number, or full identity details, no, the same postcard rule applies to pictures as to typing. Show it the puzzling notice; don't upload your bank statement or your ID.
When you have finished it
Do one thing without the keyboard. Tap the microphone and ask a question out loud, or point the camera at something puzzling in your home and ask what it is. Just once. Notice whether that way in feels easier or more natural than typing. If it does, you've just found your door, and you're allowed to use it for everything from here on.
Chapter 8 · about 1,300 words
Mistakes, Misinformation, and the Confident Wrong Answer
It opens on We've circled this from the start, and now we land on it directly, because it's the single most important skill in the book.
What it answers
- How am I supposed to know when it's wrong, if it always sounds sure?
- You don't judge by the tone. The tone is useless, always confident. You judge by the category. Numbers, dates, names, quotes, sources, laws, medical and money facts: these get checked, every time, no matter how certain they sound. General explanations of stable ideas: relax. The sound of the answer tells you nothing; the kind of answer tells you everything.
- Does it get worse over time, or 'lie' on purpose?
- Neither. It isn't deceiving you, there's no intent behind it at all. It simply has no way to feel the difference between remembering and inventing, so both come out in the same steady voice. Knowing that, you supply the one thing it lacks: a sense of which claims to verify.
When you have finished it
Deliberately ask the tool something it's likely to get wrong. A very specific fact, an obscure quote, the details of a small local thing. Then look at how confidently it answers, and go confirm it against a real source. The goal is to feel, once, the gap between how sure it sounds and whether it's right. Once you've felt that gap with your own eyes, you'll never again mistake fluency for truth, and that single lesson makes you safe everywhere else.
Chapter 9 · about 1,500 words
Scams, Privacy, and Staying Safe
It opens on This chapter is protective, not frightening, and the difference matters.
What it answers
- If the fake voices are that good, how can I ever trust a call again?
- You trust the channel, not the voice. A real emergency survives you hanging up and calling back on a number you already have. So you stop trusting caller ID and urgent voices, and you start trusting your own call-back and your family safe-word. That shift, from believing what reaches you to verifying through a path you chose. Keeps you safe no matter how good the fakes get.
- Is it dangerous to use these tools at all, privacy-wise?
- Used with the postcard rule, no. Ordinary conversations, recipes, letters, learning. Carry little risk. The risk lives in what you volunteer, so you simply don't type passwords, account numbers, or full identity details. Keep the keys in your pocket and you can use everything else with an easy mind.
When you have finished it
Two small things, and they're among the most valuable in the book. First, set up a family safe-word, send one message or make one call to your closest people and agree on it today. Second, pick the account that matters most to you, find its real customer-service number, and save it in your phone under a name you'll recognize, so that if a scary message ever comes, you already have the trusted path in hand and never have to use theirs.
Chapter 10 · about 1,200 words
Building Your Personal AI Routine
It opens on By now you have the pieces, what these tools are, how to talk to them, where they help, where they lie, how to stay safe.
What it answers
- How often 'should' I use it?
- There's no should. The readers who are happiest with these tools often use them only a few times a week, for two or three specific things, and forget about them otherwise. Frequency is not the goal; fit is. If it earns its place twice a month, it's earning its place.
- I keep forgetting it exists. Am I doing it wrong?
- Not at all. Unlike a language or a fitness routine, this skill doesn't fade when you ignore it. It waits in the drawer and works just as well the day you return after a month away. Forgetting for stretches is normal and fine. Hitch it to one thing you already do each week, and let the rest come or not.
When you have finished it
Start the one-week scrap-of-paper log described above, just begin collecting your "sighed at a chore" and "wish I understood that" moments, without acting on them yet. At the end of the week you'll have something more valuable than any list we could give you: a map, in your own handwriting, of exactly where this tool belongs in your particular life.
Chapter 11 · about 1,300 words
How to Check an Answer Before You Trust It
It opens on We gave you the principle in chapter eight: fluent doesn't mean correct, so verify anything that matters.
What it answers
- Isn't all this verifying exhausting?
- It would be if you checked everything, so you don't. Most answers. The recipes, the reworded notes, the explanations, need no checking at all. Verification is reserved for the handful of specific, high-stakes claims where being wrong costs something, and those you were going to want a real source for anyway. In practice it's a quick instinct, not a chore.
- Can I trust the sources it gives me?
- Treat them as leads, never as proof. Inventing real-looking sources is exactly what these tools do most confidently. If it names a study, an author, or a website, go find that thing yourself before you rely on it. A citation from an AI is a starting point for checking, not the end of it.
When you have finished it
The next time you get a specific, checkable fact from one of these tools. A number, a date, a source, run the four questions, out loud if it helps, and actually go confirm it at a real source. Do it once deliberately and slowly. After that, it starts happening on its own, quietly, in the background, exactly when it's needed and never when it isn't.
Chapter 12 · about 1,200 words
AI for Family History, Creativity, and the Projects You Keep Meaning to Start
It opens on We'll end where the tools are most purely joyful, because after all the sensible cautions you've earned a chapter that's mostly about delight.
What it answers
- Will using it to write my stories make them less mine?
- Only if you let it write instead of you. Used well, it does the opposite. It asks the questions that pull your memories out and tidies your own words without replacing them. The stories stay yours; the tool just helps them onto the page. Read the result and change anything that doesn't sound like you, and it will be entirely yours.
- I've learned all this, how do I keep from forgetting it?
- Use one thing this week, and teach one thing to someone else. Nothing sets a skill like handing it on. Beyond that, keep the Quick-Reference Card near your chair, and trust that the big ideas. Language machine, check what matters, don't be rushed, are few enough to carry in your head for good.
When you have finished it
Point the tool at something you want rather than something you need. Ask it to help you start the one project you keep meaning to begin. The family stories, the poem, the garden, the thing you've carried for years. You have every skill this takes. Let the last exercise in the book be the one that's purely for your own joy.
AI Without the Jargon
Learn what AI actually is, how to ask it for useful things, and when to check its work before you trust it.