From The No-Hype AI Playbook

Chapter 1 — Your skepticism is a feature, not a bug

The eye-roll test

You don't need an account or a computer for this one. Over the next few days, every time you see the letters "AI" attached to a product — a toothbrush, a mattress, a bank commercial, a pair of running shoes — ask one question: "What does it actually do?" Not what it promises. What it does.

You'll find that most of the time, nobody says. That's how you spot hype: it describes the future and forgets the verb. Meanwhile, everything in this book passes the test with plain verbs — it plans, it explains, it translates, it drafts, it shows you pictures. Keep asking that question and you'll never be sold snake oil again, by the AI industry or anyone else.

Congratulations: you're already doing the "critical evaluation of AI capabilities" thing the experts charge consulting fees for. You knew it as common sense.

Three openers for your first conversation

You'll pick your tool in chapter 2 and learn what makes a prompt work in chapter 3 — so consider these a preview, not homework. When you have an account and a free ten minutes, type in any of these, exactly as written or adjusted to taste:

1. The honest introduction

I'm retired, I'm skeptical of AI, and I'm testing whether you're actually useful. I like plain answers without cheerleading. In three short sentences, what kinds of everyday tasks are you genuinely good at for someone like me — and what's one thing people wrongly assume you can do?

2. The bluff check

I've heard AI sometimes makes things up confidently. Explain, in plain English and no more than five sentences, when I should double-check your answers and when I can relax.

3. The dealer's choice

Ask me five questions about my house, hobbies, and weekly routine. Then, based on my answers, suggest three specific ways you could save me time or aggravation this month. One question at a time, please.

That third one is a small preview of the most useful trick in this book — making the machine ask you questions. Notice that all three prompts tell it who you are, what you want, and how you want it delivered. There's a reason for that shape. It's called chapter 3.

Works in ChatGPT, Claude, or any of their cousins.