Lesson 2 — Speak Plainly: Your First Conversations with AI

“A word fitly spoken is like apples of gold in a setting of silver.” — Proverbs 25:11 (ESV)

This lesson takes about 25 minutes.

Here is the open secret about getting real help from an AI assistant: there is no magic incantation, no insider vocabulary, no trick. The people who get good results simply say what they mean — who the thing is for, what they actually need, and how it should sound. The people who get generic mush type “write me an email” and take what they get.

Scripture has always treated speech as a craft worth practicing. This lesson makes you a craftsman of one very modern kind of speech: the request.

What you’ll learn today

You’ll learn the three-part shape of a request that works — and the follow-up habit that turns a mediocre first answer into exactly what you needed. This is the single most practical skill in the book, and you will use it every remaining lesson.

The three-part request

Think of how you’d ask a capable, brand-new church secretary for help on their first day. You wouldn’t say “write a letter.” You’d say who it’s for, what it needs to accomplish, and the tone the moment calls for. The machine needs exactly the same three things — because unlike that secretary, it knows nothing about your life except what you put in the message box.

1. Who and what for. “I’m writing to the parents of our Sunday school kids about the schedule change to 9:15.”
2. What you need. “Draft a short email — under 150 words — that gives the new time, the start date, and who to contact with questions: Ruth, at the church office.”
3. How it should sound. “Warm and unfussy. No exclamation points.”

Put those together and send it as one message. The difference in what comes back — compared to “write an email about the schedule change” — will make you smile. Not because the machine got smarter, but because you finally told it what you know.

Coach’s Note — The single most common mistake is leaving the important facts in your head. When an answer comes back generic, don’t hunt for a magic phrase — ask yourself, “What did I know that I didn’t say?” The date, the audience, the reason, the tone. Nine times in ten, the fix is one more sentence of plain context. You are not programming; you are briefing.

The follow-up: where the real power is

The first answer is a draft, not a verdict. The machine remembers the conversation, so you can shape the result the way you’d shape it with a helper across the desk:

• “Shorter — half that length.”
• “Warmer. It reads a little corporate.”
• “Add a line thanking the volunteers who taught last quarter.”
• “Give me three different opening sentences to choose from.”

There is no limit and no cost. Ask for five revisions; ask for ten. The tool never tires, never sighs, never bills you. This is where AI genuinely serves the person who knows what they want: you keep the judgment, it does the typing.

Two habits to keep from day one:

• You send it, not the machine. Copy the final words into your own email or document, read them one last time with your own eyes, and put your name on them only if they’re true and they’re you. What goes out under your name is yours — a thread that runs through this whole book.
• Nothing private goes in. No passwords, no account numbers, no other people’s medical or financial details, nothing told to you in confidence. You can describe a situation generally (“a friend recovering from surgery”) without pasting anyone’s private life into a company’s computer. Lesson 9 and Appendix B draw the full map; the fence starts now.

Words as stewardship

Why does a book with Scripture on every page spend a lesson on phrasing requests? Because the Bible cares intensely about aimed words. “Let your speech always be gracious, seasoned with salt, so that you may know how you ought to answer each person” (Colossians 4:6, ESV) — each person: the fit of words to hearer is a spiritual skill, not a secular one. A well-briefed request that produces a clear, kind letter to a grieving neighbor is better stewardship than an hour lost to a blank page — and the hour saved can go to the visit the letter can’t replace.

Practice right here

Below this lesson is a practice panel called “Brief It Well.” It gives you a vague request — the kind everyone types on day one — and five switches you can flip to add the missing pieces: who it’s for, the goal, the key facts, the tone, the shape.

Flip them one at a time and watch the request sharpen — and watch the likely result transform from generic filler into something you could actually send. Then hit reset and, before you touch anything, predict which single switch matters most for that scenario. That prediction reflex is the skill.

Now do it for real

Take a real writing task from your own week — a thank-you note, a reply you’ve been putting off, a notice for your small group. Then:

1. Open your assistant (chatgpt.com, gemini.google.com, or claude.ai — the same free account from Appendix A).
2. Send a three-part brief: who and what for, what you need, how it should sound. One message, plain words. (Leave out any private details — initials or “my neighbor” work fine.)
3. Read the draft, then push back twice. At minimum: one change to length or tone, one added fact it couldn’t have known.
4. Copy the final version out, read it once with your own eyes, and send it from your own account — a real letter, really sent, in a fraction of the usual time.
5. Optional extra rep: ask for a Scripture-memory plan — Give me a two-week plan to memorize Philippians 4:4–9, ESV, about 15 minutes a day. Notice how the who/what/how shape works for plans as well as letters.

Faith foundations

The Old Testament’s wisdom literature reads like a manual for this lesson. “Death and life are in the power of the tongue” (Proverbs 18:21, ESV); a fit word is gold-work (Proverbs 25:11); even the timing of speech matters — “a word in season, how good it is!” (Proverbs 15:23, ESV). Words are never neutral tools in Scripture; they build and they wound, they bless and they deceive.

The New Testament raises the stakes further: “on the day of judgment people will give account for every careless word they speak” (Matthew 12:36, ESV). That verse now has a modern corollary worth sitting with: words you generate and send are words you speak. The machine drafted it; you uttered it. Care in what we ask for, honesty in what we sign, grace in what we send — the old disciplines fit the new tool exactly.

And there is a quiet apologetic here too. Language — real language, meaning aimed at persons — is God’s invention and gift; the eternal Son is himself called the Word (John 1:1). A machine can shuffle our words because our words were already saturated with meaning it never made. Every useful answer it gives you is, at bottom, borrowed light.

For your group

1. What’s the difference between briefing a helper and commanding a machine? Which posture produces better requests — and why might it also be better for your own heart?
2. Share your before-and-after: what did the vague request produce, and what did the three-part brief produce?
3. Read Colossians 4:6. What does “seasoned with salt… each person” imply about form letters — human or AI-drafted?
4. Where’s the line between AI helping you say something and AI saying it for you? When would using a drafted letter feel dishonest?
5. What writing task have you been avoiding that this skill could unstick this week — and what would you do with the saved hour?

You did it

You now know the three-part brief, the push-back habit, and the two fences (your name, your privacy). You’ve sent a real letter with half the friction. That’s fluency beginning.

Up next: In Lesson 3, we open the hood — because to trust a tool rightly, you need to know why it sounds so sure when it’s wrong.