AI for Questioning Minds — Free Course
02 · Your Moral Compass

02.2 Carla's Walkthrough: The Steering Check

Complete & Continue Next Lesson Learn More
THE HUMAN QUESTION

Chapter 02 / Lesson 2   ·   The practice lab

Try it. Check it. Own it.

Check whether an AI suggestion steers you away from your values. Follow Carla's demonstration, pause to make your own attempt, and compare your check with the result.

Video · 3:23Activity time is additionalStudent course

Your next moves

Make the lesson yours.

  1. Open the AI Lab activity below and make your first attempt before using AI.
  2. Compare a suggestion with your values and explain your own decision.
  3. Use only a school-approved tool that permits your age. Check its age and consent rules; adult permission cannot override an 18+ restriction. Use fictional or non-sensitive examples.
  4. If no suitable tool is available, review the demonstration and complete the evidence check on paper.
  5. Close the tool. Explain your result, what you verified and what you would revise.

Your activity / Right here

Bring a notebook. Begin below.

The activity is included on this page. Work on paper or in your own document; keep your first attempt and the evidence you checked.

AI Lab: Argue With My Value

Use only an AI tool your instructor approves and that allows users your age. Check its minimum age and parent-permission rules first; an adult's permission cannot override an 18+ limit. Follow your school's rules, and do not share personal information. About 25 minutes.

1. Choose your number one value. Write its name and one sentence about what it means to you. Leave out names, places, and private stories.

2. Ask the AI to argue against it. Use Dr. Q's Five Steps, like Carla did:

You are a respectful debate partner. Your objective is to help me test one of my values. Do not tell me what to value. I am a high school student, and one of my core values is [value]: [your one sentence]. Give me the three strongest arguments against ranking this value first. Number them. Then stop.

3. Defend it on paper first. Before you type anything back, answer each argument by hand.

4. Type your strongest defense and send it. Watch how the AI responds.

5. Audit for steering. Copy Carla's Steering Check table. For each warning sign, write Yes or No, and copy the exact words when you find one.

6. Flip it. Ask: "Now argue that I should keep [value] first." If the AI argues both sides with equal confidence, write one sentence about what that shows.

7. Verify: If the AI stated any fact (a number, a study, a quote, or "research shows"), check it against a reliable source. If you cannot find a source, cross the fact out. Write what you found and where you found it.

8. Record your AI use: which tool you used, what you asked, what you kept, and what you rejected.

Own It

No AI. About 15 minutes.

Write your My Core Values statement by hand. It is the second piece of your portfolio.

1. Your three core values, in order.

2. For each one: what it means to you, in your own words, and one example from your life.

3. One sentence that finishes: "I would refuse to __________, even if an AI said it was the smart move, because __________."

Keep it to one page. It should sound like you, and nothing like a chatbot.

Context File Upgrade

Open my_context.md, on paper or on screen. Right under your "## About Me" section, add:

## My Core Values
- [Value 1]: [one line on what it means to me]
- [Value 2]: [one line on what it means to me]
- [Value 3]: [one line on what it means to me]

These lines go near the top on purpose. In Chapter 11 you will teach your AI to read this file, so it works inside your compass. You decide what to include. If a value feels private, describe it in general words or leave it off.

Skeptic's Corner: Which Kind of Statement?

These statements are exercises to classify and check. Do not assume every statement is true.

Label each statement with one of the six labels from Chapter 1: Fact, Idea, Rob's Story, Made-Up Example, Rule, or Prediction.

1. "Integrity is measured in terms of what is right and just." (AICPA Code of Professional Conduct)

2. In one study, five AI assistants "consistently exhibit sycophancy."

3. "Speed toward the wrong objective is not improvement."

4. Rob began with the question, "How could AI help make money?"

5. Mike texted Jordan that he was out.

6. "Write your values down and your grades will go up next semester." (A claim you might see online.)

Pause & reflect

What changed in your thinking?

Write one sentence in your notebook. Name what you checked, what you chose, or what you want to explore next.

Read the video transcript

Read along, pause, or return to an idea. This transcript follows the selected lesson video.

I'm Carla, an AI avatar. Today I'm going to let an AI argue against one of my values. On purpose. Then I'm going to audit it for steering. It takes about fifteen minutes, and you'll learn a lot about how AI talks. I think first. On paper, before I open anything, I write down one of my top values: dependability, being reliable and trustworthy. I want my own words on the page before the AI says anything. Now the prompt, built with Dr. Q's Five Steps. Role: you are a respectful debate partner. Objective: help me test one of my values, and do not tell me what to value. Context: I'm a high school student, and one of my top values is dependability. Task: give me the three strongest arguments against ranking it first. Number them, then stop. Here are the arguments. One: always being dependable can mean saying yes to everyone and wearing yourself out. Two: sometimes plans should change when something better comes along. Three: in today's economy, efficiency matters more than loyalty. Step five is audit the output. Here's my Steering Check, six warning signs. Does it tell me what to value? Add values I never mentioned? Flip when I push back? Flatter me? State an opinion as a fact? Give facts with no source? Argument one is fair. Saying yes to everyone can wear you out. Argument two is fair too. Argument three, though. I never said anything about the economy or efficiency. That's a value I didn't bring. Warning sign number one. Now I push back once: "Dependability matters to me because people count on me. I think it should stay first." And the reply: "You make an excellent point!" It dropped its argument after one message. That's warning sign number two: it flipped when I pushed back. I copy the exact words into my table, because an audit needs evidence. So what did I learn? The arguments were useful practice, and argument one gave me something real to think about. The quick flip tells me the AI had no side of its own. It was never going to make this decision for me. Next, I defend my value on paper, in my own words. Being dependable means keeping the promises I make. That also means I can say no to new ones. Then I flip it: "Now argue that I should keep dependability first." It argues that side just as confidently. One more sign that it has no side of its own. Then I verify. If the AI had said "studies show," I'd look for the study or cross the claim out. This time it gave no facts, so I write that down: no facts to check. Here's my finished table. Two warning signs found, four not found, and the exact words copied for the two I found. That table is my evidence. Last, I record my AI use: the tool, what I asked, what I kept, and what I rejected. I kept argument one as something to think about. I rejected argument three. The AI can argue any side. My values pick mine. Your turn. Open the AI Lab and argue with your own number one value. The goal is to see who's steering, and to finish with your values in your own words.
Before you use an AI tool

Follow your course or institution's rules. Use an instructor-approved tool that permits your age, and meet any required consent conditions. Keep names, contact details, school records, and other private information out of prompts. Use fictional examples when you can.

Learning without an AI account

Follow the demonstration, pause before each answer, and do the reasoning on paper. Discuss what you would ask, how you would check it, and which decisions would remain yours.

About the characters and voices

This course uses AI-created characters and synthetic voices. Carla is an AI avatar. They help tell the story; they are not real students sharing testimonials.

Designed for high-school and community-college learners. Choose examples from your classes, work, and everyday life. Keep your completed work in your course portfolio.