Borrow a habit from auditing: ask where the evidence came from. Test a simple calculation and decide whether a source is relevant, reliable and sufficient.
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.
Warm-up: Dr. Q's Question
Walk me through exactly where you got that.
My first answer: ________________________________________________
Mike's Mix-Up: The $1,240 Profit
Made-Up Example.
Here is what Mike's group did:
1. They uploaded their sales spreadsheet and typed, "What's our profit?"
2. They copied the AI's number onto a slide.
3. Nobody checked which numbers the AI used.
4. Nobody could explain the number when Dr. Q asked.
That night, Mike tried the move. He typed, "Walk me through exactly where you got that." The AI showed its math: 62 hoodies times $20 each. That came to $1,240, the money the group took in. The spreadsheet had no column for what the hoodies cost, and the AI never mentioned that anything was missing. It called the money coming in "profit."
Mike found the printing company's receipt in his email: $12 per hoodie, or $744 in all. He did the math by hand. $1,240 minus $744 is $496. He fixed the slide. Then he wrote the first entry in his AI Audit Log.
What went wrong? Talk it through with a partner, then write your answers.
• At which step should someone have asked, "Where did you get that?"
• What evidence finally caught the problem? Was it original, direct, or independent?
• Was the AI wrong, was the question unclear, or both?
• What one sentence could Mike add to his context file so this happens less often?
Think First: Investigate a Claim at Home
No devices. Pen and paper. About 20 minutes.
Pick one claim you can check at home. It can be something a person says or something printed on a label. For example:
• "We always run out of milk by Thursday."
• "This box has about 12 servings."
• "The kitchen clock is right."
• "I'm the only one who ever takes out the trash."
Then work like an auditor.
1. Write the claim exactly as it was stated.
2. Set your criteria. What would prove it true? What would prove it false?
3. Collect at least three pieces of evidence. At least one must be something you count, measure, or observe yourself.
4. Rate each piece in a table like this one:
| Evidence | Where I got it | Relevant? | Reliable? Why? |
|---|
| | | |
5. Decide. Write one of three conclusions: Supported, Not supported, or Not enough evidence yet. The third one is an honest answer, and a good auditor respects it.
Read the video transcript
Read along, pause, or return to an idea. This transcript follows the selected lesson video.
Welcome to Part 3, The Auditor's Mindset.
I'm Carla, an AI avatar.
This is the part of the course Dr. Q has been waiting for,
and it starts with one sentence.
Walk me through exactly where you got that.
Mike's group ran a hoodie sale.
They gave their spreadsheet to an AI,
and it said the profit was twelve hundred forty dollars.
Dr. Q asked where that number came from.
Nobody could answer.
So that night, Mike asked the AI the same question.
It showed its math.
It had counted the money coming in and never subtracted
what the hoodies cost.
The real profit was four hundred ninety-six dollars.
Notice what caught the error: one question and one receipt.
The receipt was the original document,
and it settled the question.
Auditors have a name for the habit behind Dr. Q's question.
They call it professional skepticism.
Here is how an official auditing standard defines it.
Professional skepticism is an attitude that includes a
questioning mind and a critical assessment of audit
evidence and other information.
A questioning mind.
A critical look at the evidence.
You can trust people and still check their claims.
Rob, who wrote this book, is an auditor.
His Audit Midyear slides apply the same habit to AI.
One slide says it plainly: students should learn not to
subordinate judgment.
Subordinate means to put under.
Keep your judgment on top.
In 1987, President Ronald Reagan signed a treaty with
Soviet leader Mikhail Gorbachev and quoted an old Russian
saying: trust, but verify.
Gorbachev said, "You repeat that at every meeting."
Reagan answered, "I like it."
Good evidence passes three tests.
Is there enough of it?
Is it relevant to the claim?
Is it reliable?
Enough means you checked more than one number when a lot
is riding on it.
Relevant means the evidence speaks to the claim.
Reliable means a strong source: independent,
direct, and original, like the library's own posted hours
instead of an AI's guess.
One warning.
When you ask an AI, "Are you sure?",
you are asking it to check itself.
To verify an AI, step outside the AI.
An auditor stays independent of the company being audited.
You stay independent of the AI by checking with a separate source.
This week, you'll investigate a claim at home,
make an AI walk you through its answer,
and start your AI Audit Log.
Grab a pen.
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.
