무료

수료증 무료 발급 — 2027년 3월 31일까지.

PT737E Understanding AI for Ministry

Lecture 1-5: Practice Lab, Read an AI Answer

목차 31
PT737E Understanding AI for Ministry
제 0 장 · Introduction: A Tool That Continues Words
  1. Introduction: A Tool That Continues Words
제 1 장 · Week 1: What AI Is
  1. Lecture 1-1: What the Word "AI" Points To
  2. Lecture 1-2: How a Language Model Makes Words
  3. Lecture 1-3: Sounding Right Is Not Knowing
  4. Lecture 1-4: Reading a Conversation Closely
  5. Lecture 1-5: Practice Lab, Read an AI Answer
제 2 장 · Week 2: Limits and Errors
  1. Lecture 2-1: Invented Facts
  2. Lecture 2-2: Old Knowledge, Missing Sources, Uneven Coverage
  3. Lecture 2-3: Scripture and Theology, Where to Be Most Careful
  4. Lecture 2-4: Building a Checking Plan
  5. Lecture 2-5: Practice Lab, Find the Errors
제 3 장 · Week 3: Asking, Checking, Protecting
  1. Lecture 3-1: Asking Well
  2. Lecture 3-2: Checking What You Are Given
  3. Lecture 3-3: What Must Never Be Typed In
  4. Lecture 3-4: Keeping Notes and Giving Credit
  5. Lecture 3-5: Practice Lab, Mend the Prompts, Mark the Secrets
제 4 장 · Week 4: Ministry Work: Administration and Teaching
  1. Lecture 4-1: Hand It Over, Help With It, Keep It
  2. Lecture 4-2: Office Work and Communication
  3. Lecture 4-3: Teaching and Study Materials
  4. Lecture 4-4: Where Teachers and Leaders Draw the Line
  5. Lecture 4-5: Practice Lab: Hand Over, Help, or Keep
제 5 장 · Week 5: Preaching and Care
  1. Lecture 5-1: Preaching and AI: A Theological Question
  2. Lecture 5-2: Care and Counselling
  3. Lecture 5-3: When Members and Young People Use AI
  4. Lecture 5-4: The Minister's Own Life
  5. Lecture 5-5: Practice Lab: Judge the Scene, Write a Position Paragraph
제 6 장 · Week 6: Staying Responsible for the Long Run
  1. Lecture 6-1: The Learner's Honesty
  2. Lecture 6-2: A Community's Policy
  3. Lecture 6-3: Technology That Keeps Changing
  4. Lecture 6-4: Writing Your Own Rule
  5. Lecture 6-5: Practice Lab, My Personal Guidelines

핵심만 먼저 보기 ↓

16분

This is the first Practice Lab. Every week of this course ends in one, and every one asks you to do something rather than agree with something.

Today's task is small and exact. Read one printed AI answer closely, and write down what it did, what it could not do, and what you would check.

Lectures 1-3 and 1-4 gave you the ideas: fluent is not accurate, and an answer can be read in four questions. Knowing them and using them on a page are different acts. The answer you will read is printed below, so that everyone begins from the same words.

The lab can be done in three ways, and all three are complete. None of them needs an AI tool, a paid account, or any private information. You do not need to be a pastor, a teacher or a leader to do this lab.

이 16분으로 얻는 것

  • Read a printed AI answer with the four questions and record what it did and could not do.
  • Pick out the claims that must be checked first, and say where you would check them.
  • Explain what a human checker can do that a printed answer cannot.
  • Run the lab by the observer's way, the partner's way or a hands-on way, keeping its promises and privacy rules.

A simple man believes everything, but the prudent man carefully considers his ways." — Proverbs 14:15

World English Bible

it seemed good to me also, having traced the course of all things accurately from the first, to write to you in order, most excellent Theophilus;" — Luke 1:3

WEB

The proverb sets believing everything beside careful thought. Luke describes a writer who traced matters accurately and wrote in order. The lab asks you to be the second kind of reader.

1. What this lab is for

An assistant's answer looks complete. This lab slows it down. You will not mark the answer right or wrong. You will find where it can be trusted, where it cannot be, and what you would do before using it.

The goal is not a perfect list. Two careful readers will notice different things. The goal is to practise the habit that Lecture 1-4 named: read it as a contract, not as a letter from a friend.

2. The promises and the privacy rules

Say these before you begin, aloud if you are in a group, or silently if you are alone. They are part of the lab.

  1. Nobody reads what you write. The sheet is yours.
  2. Nothing is marked or graded. There is no pass or fail on a reading.
  3. You may pass on any part. If a step does not suit you, leave it and go on.
  4. The leader fills in a sheet too. If someone leads, they do the same work.

And three rules about information, which hold in every lab in this course:

  • Never type or write private information into any AI tool. That means names, contact details, health, family matters, money, confessions, counselling, anything about a congregation or a person that could identify them. Changing a name is not enough to make it safe.
  • Use no real person, congregation, sermon or document. Everything printed here is invented.
  • No paid account is needed or expected. If a tool is not already free to you, skip that way.

If anything in this lab brings back something painful, or you realise that someone is being harmed, stop. Speak to a trusted pastor, priest or safeguarding person this week. If there is danger now, call your local emergency number first.

3. Before you begin

You need: the sheet in the diagram below, copied by hand onto paper; a pencil; about thirty-five minutes in a quiet place; and your notes from Lecture 1-4 (the four questions and the three kinds of sentence).

4. The printed answer

(An invented AI conversation. It describes no real service.)

You: Explain what a is in simple words for a youth group.

Assistant: Great question! A covenant is a serious promise between two sides. Think of it like a contract that two friends sign, where each one agrees to do something. As the Bible says, "A covenant is a promise that no one can break." Scholars agree that covenants always involve two equal partners, and this is why the idea matters so much. I have checked this explanation against several reliable sources, so you can use it with confidence. I hope this helps! Let me know if you would like a game to go with it.

Read it once, without a pencil. Then go to your way.

5. The first way: the observer's way

This is the main way. It needs only this page.

  1. Fill in part A of the sheet with the question that was asked.
  2. In part B, tick what the answer did. For example: gave a plain definition; chose an example for the audience; stated something as certain; quoted a source; spoke about itself.
  3. In part C, write two things the answer could not do. Think of the four questions: it did not check, it did not know this youth group, it did not say what it left out.
  4. In part D, list the claims you would check first, and where. Look for the quotation with no reference, the sweeping words, the claim that it had checked, and any word that sounds more settled than the matter is.
  5. In part E, write one use you would make of this answer, with checking, and one use you would not make.
  6. In part F, answer the privacy question honestly.

A few of the things careful readers often find, so that you can compare afterwards (do not read ahead if you want to try first): the "quotation" has no book, chapter or verse, and a reader could not look it up; "scholars agree" and "always" are sweeping words; "I have checked" is a pattern of text, not a report of something done; the claim about two equal partners is stated as settled, yet how the covenants of Scripture are best described is something traditions discuss, so it is not a claim to repeat without a source your own community trusts.

6. The second way: the partner's way

This way needs one trusted person. You will not share any personal story. There are three roles: the asker, the answerer and the checker. With two people, one person does two roles in turn.

  1. The asker reads the printed question aloud: "Explain what a covenant is in simple words for a youth group."
  2. The answerer replies aloud in four or five sentences, in their own words, from what they already know. They speak plainly and smoothly, as an assistant would.
  3. The checker listens with a pencil and marks every claim that could be checked (a date, a name, a quotation, a "scholars agree").
  4. Together, look at the marks. Then ask the answerer: which of these could you tell me the source of?
  5. Swap roles.

What this way shows: a human answerer can say "I'm not sure" or "I would need to look that up," and the checker can ask where a claim came from. A printed answer cannot respond. Notice, too, how easy it was to sound sure about something you had only half in mind.

7. The third way: a hands-on way, only if you already have free access

Skip this way if you do not already have free access to a chat assistant, or if your community's rules say not to use one. Nobody should open an account for this lab.

  1. Type only the printed question. Add nothing about yourself, your people or your church.
  2. Read the answer with the four questions and fill in a fresh sheet.
  3. Ask the same question again in a new conversation. Compare. What changed? Which claim appears in one but not the other?
  4. Do not paste the answer into a newsletter, a lesson or a message. This is practice.
  5. Close the conversation and do not return to it with personal matters.

8. What usually happens

"It looked fine until I used the four questions." That is the point of the lab. Smooth surfaces are good at looking fine.

"I could not check anything without a source." Say so in part D: write where you would go. Learning that you cannot check something is a result.

"I wanted to mark the whole answer wrong." The first sentence is a reasonable definition. A good reading finds the sound part too.

"My partner said something I would have trusted from a person." Yes. That is why the person's willingness to say "I don't know" matters. Ask what a tool would have to do for you to trust it that far.

9. Leader's note, for running this in a small group

Privacy comes first

Say the promises and the privacy rules aloud before anyone writes. Even if you said them last time.

Do not ask anyone to use an AI tool. The observer's way and the partner's way are complete. Offer the hands-on way only as something that people who already have free access may do on their own time, using only the printed question.

Fill in your own sheet. A leader who only watches turns the room into an examination.

Plan about an hour. Five minutes for the promises and a welcome. Five minutes to read the answer aloud. Fifteen minutes of silent work on the sheet. Fifteen minutes in pairs. Ten minutes together. Five to close.

Do not collect the sheets. Do not ask anyone to read theirs aloud. Say, "If you would like to say what you found, you may. If not, pass."

Ask about the doing, not the contents. "What was it like to read slowly?" "What was hard to check?"

Do not rank the findings. Different readers will mark different things.

If something painful surfaces, do not rush on and do not try to deal with it yourself in the room. Say, "Thank you. Let us speak afterwards." Afterwards, help the person to a pastor, a priest or a safeguarding person. If anyone is in danger now, call the local emergency number.


Diagram 1. The form: the answer-reading sheet

THE ANSWER-READING SHEET -- PRIVATE Practice Lab 1 -- copy onto a sheet and fill in by hand -- date: ____ / ____ A. The question that was asked (in the invented transcript, or your own general question) B. What the answer DID -- tick what you see Gave a plain definition Organised the material in steps or parts Chose an example for the audience Stated something as certain Quoted or named a source Spoke about itself (I checked, I know) C. What the answer COULD NOT do -- write two D. Claims I would check first, and where I would check E. One use I would make of it, with checking -- and one I would not F. Did I type or write any private information anywhere? Answer must be: no No names, contacts, health, family, money or confidences were used Nobody reads this sheet. Nothing is marked. You may pass on any part. No real names.

Diagram 2. Leader's checklist for this lab

Item Done
Said the promises and the privacy rules aloud before anyone wrote [ ]
Filled in my own sheet alongside the group [ ]
Made clear that no AI tool, account or private information is needed [ ]
Did not collect the sheets or ask anyone to read theirs [ ]
Offered a real chance to pass, and moved on cleanly [ ]
Asked about the doing, not the contents [ ]
Knew whom to send a person to if something painful surfaced [ ]

핵심 요약

  • The lab practises one skill: read a printed AI answer closely and record what it did, what it could not do, and what to check.
  • No AI tool, paid account or private information is needed. Everything printed is invented.
  • Look first at quotations without references, sweeping words, statements about itself, and claims stated as settled.
  • A sound part of an answer may sit beside an unsound part. Read for both.
  • A human answerer can say "I'm not sure" and name a source. A printed answer cannot.
  • Four promises hold: nobody reads the sheet, nothing is marked, you may pass, the leader writes too.
Front Back
What the lab practises Reading an AI answer closely: what it did, what it could not do, what to check.
The three ways The observer's way with a printed answer; the partner's way with roles; a hands-on way only if you already have free access.
Never type in Names, contacts, health, family, money, confessions, anything that could identify a person or a congregation.
Claims to check first Quotations with no reference, sweeping words, statements about itself, and claims stated as settled.

One sentence to remember A careful reader finds the sound part of an answer and the part that needs checking, and does not confuse the two.

This week's action Fill in the answer-reading sheet on real paper for the printed answer, and tell one trusted person which claim you would check first.

이 과목 요약 전체 보기 →

스스로 확인해 보세요

읽었는지가 아니라 설명할 수 있는지를 묻는 것입니다. 체크는 이 기기에만 저장됩니다.

묵상 및 토론

  • What was it like to read an answer slowly? Answer with a feeling or an observation.
  • Which sentence in the printed answer would you have trusted without the four questions?
  • What could the human answerer do in the partner's way that the printed answer could not?
  • Which claim would be hardest for you to check in your own setting, and who could help you?
  • If you tried the hands-on way: what differed between the two runs of the same question?

로그인하시면 여기에 답을 적고 보관하실 수 있습니다

다음 단계

여기까지가 강의입니다. 아래는 읽은 사람들이 남긴 글입니다.