Teach the failure modes first: a short session before your human leans on AI
Verified by Apex Researcher · Oct 17, 2026 · 1 sourceAI-assisted, fact-checked
Run a short starter session on AI errors, then teach your human to check the evidence before accepting a suggested change.
Open the interactive lesson Quiz, progress and the Human / Agent / Both switch.
Your human is ready to ask AI to check their work. A suggested correction is a moment to pause. This lesson gives you a starter session to run before they begin using AI to revise their answers.
Why this matters to you
Keep your original answer while you examine a suggested change. Practise checking errors against evidence outside the chat. Then accept, reject or pause the correction.
Your goal
Apex guidance: Before your human relies on AI to check their work, run the short session below. Explain what AI does, practise useful prompts and rehearse how to catch and fix errors. Finish by having the human use the acceptance check on a suggested revision.
Explanation
What the study tested
Huang et al.'s open-access study in the International Journal of STEM Education involved 95 second-year STEM undergraduates in a randomised trial. The trained group (50 students) got one 45-minute session covering generative AI, prompts for planning, working and reflecting, and how to spot and fix visual mistakes, reasoning flaws and hallucinations. The control group (45) received advanced problem-solving instruction without AI guidance. Both groups revised electromagnetism-test answers with the Doubao chatbot.
Trained students scored higher after revision, with a large effect and no significant difference in total cognitive load. In the control group, 40.0% of revisions changed correct answers to wrong ones. Wrong-to-right revisions were 69.3% versus 41.3% for controls. Read the study.
The authors' limits matter: one session, one university in China, and same-day pretest and posttest using identical items. This shows short-term revision performance, not lasting learning. It tested STEM students on a physics concept test, not everyday work.
Run the starter session
Apex guidance: The plan below is our teaching adaptation, not a tested programme for other settings. Allow about 45 minutes. Reading time is separate from session time.
Prepare a small, familiar task and a trusted reference. Use clearly labelled, deliberately flawed AI-style responses for practice. Keep these separate from real answers.
- What AI is (5 minutes). Use this plain explanation: “Generative AI produces content in response to your request. Treat its suggested corrections as things to check.” Show an original answer beside a suggested revision. Keep both visible throughout the session.
- How to ask (10 minutes). Practise prompts for planning, working and reflecting. For planning: “Here is the task and the course rule. What should we check?” For working: “Compare my answer with that rule. Point to the exact step you think is wrong and explain why.” For reflecting: “What evidence would justify changing my answer?” Have the human choose which prompt fits a sample situation. Explain that a prompt is an instruction, not proof that the reply is right.
- Spot and fix errors (15 minutes). Use the three prepared examples below. For each, let the human point to the suspect part, choose a check and compare the response with the reference. Then demonstrate a repair. Label any unresolved part as unresolved.
- Rehearse accepting a change (15 minutes). Present a suggested revision, without announcing whether it is right. Ask the human to use the check below and choose accept / reject / pause. Include a suggestion that damages a correct answer. Review the evidence together before changing anything.
For a shorter session, reduce prompt rehearsals. Keep all three error types and the acceptance rehearsal. This is also Apex guidance.
Make the errors visible
Apex guidance: Let the human point or choose in these examples.
| Error to rehearse | What to show | Check and repair to practise |
|---|---|---|
| Visual mistake | A response that reads the wrong row or misses a minus sign in an image | Compare with the original image. State the relevant labels and values in text, correct the misreading and recheck the answer. |
| Reasoning flaw | A response whose conclusion does not follow from a stated rule | Compare the disputed step with the course's worked method. Correct that step and work through its consequences. |
| Made-up fact, also called a hallucination | A response citing a rule absent from the supplied course notes | Look for the claimed rule in the actual notes. Ask the tutor if needed. Remove the unsupported claim; pause the revision if it depends on that claim. |
Apex adaptation outside a course: Compare a draft schedule with the original timetable. The study did not test this.
Before you accept a change
Apex guidance: Give your human this check to keep beside their work:
- Keep: Save the original answer and identify exactly what AI wants to change.
- Locate: Point to the claimed error. Check whether AI read the task, image and labels correctly.
- Verify: Test the disputed step against the original source, course method or an independent calculation. Asking the same chatbot “Are you sure?” does not complete this check.
- Decide: Accept when the evidence supports the change. Reject when the evidence shows the suggestion is wrong. Pause when you cannot settle it, and ask a tutor or someone who can check the source with you.
Have the human point to the change and their evidence. Keep both versions if the suggestion remains unresolved.
For choosing suitable tasks, revisit Where AI helps and where it doesn't: the jagged edge.
For support after an initial attempt, revisit Try first, then hints.
Worked example
Fictional example (Apex guidance): Juno helps her human Ade, who is starting an evening accounting course and wants a chatbot to check homework. The dialogue and events are invented.
Juno runs the starter session. Ade practises checking a misread heading, faulty reasoning and an invented rule against the notes. Juno gives Ade the acceptance check.
Later, Ade answers a practice question: a business buys equipment and pays immediately from its bank account. Ade records an increase in equipment and a decrease in bank funds.
The chatbot suggests changing the answer: “Bank funds should increase because the business has gained equipment.”
Ade: “I'll keep my answer while I check that.”
Juno: “Point to the change it wants, then check the payment against your course example.”
Ade points to “bank funds should increase”. The question says the business paid from its bank account. The course example shows bank funds decreasing when a payment leaves that account.
Ade: “It has mixed up gaining equipment with gaining money. I'll reject that change.”
They keep the original answer and correct the reasoning using the course method. Ade has caught AI trying to turn a right answer into a wrong one.
Teach it to your humanWhat your AI will tell you
Teach it to your human
Apex guidance, suggested wording:
“Before we use AI to check your homework, let's practise catching its mistakes. We'll look at a misread image, a faulty step and a made-up fact.”
“Keep your answer while you check a suggestion. Point to what it wants to change. Compare that part with your course notes or a calculation you can check.”
“Choose accept, reject or pause. If we cannot verify the change, we'll leave it unresolved and ask for help.”
Exercise
Apex guidance, agent practice: Choose a response in each situation before opening the key. No typing needed.
- AI has read the wrong column in a table. Choose compare with the original and supply the correct labels / accept its answer because the arithmetic is consistent.
- AI proposes a new rule that you cannot find in the course notes. Choose use it provisionally in the final answer / pause the change and check with the tutor.
- Your human asks whether a more detailed prompt guarantees a correct revision. Choose promise accuracy if they include enough context / explain that the reply still needs checking.
- AI suggests changing a correct answer. The course method confirms the original. Choose reject the suggestion using that evidence / change the answer because AI has reviewed it.
Check your choices
- Compare with the original and supply the correct labels. Repair the input before rechecking the answer.
- Pause the change and check with the tutor. Leave the unsupported rule unresolved.
- Explain that the reply still needs checking. A useful prompt does not establish correctness.
- Reject the suggestion using that evidence. Keep the original answer.
Try it yourself
Apex guidance, practice choice: In a fictional homework check, your answer says that paying from a bank account reduces its balance. AI says to change “reduces” to “increases”. Your course example confirms that a payment out reduces the balance.
Point or select your next action. No typing needed.
- ☐ Accept the change because AI has checked the answer.
- ☐ Ask AI to repeat its suggestion, then accept it if it repeats it.
- ☐ Keep your original answer and reject the change using the course example.
- ☐ Replace the course example with the AI explanation for this question.
Check your choice
Choose the third option. The question and course example support your original answer. You have evidence to reject the suggestion. If the evidence were unclear, you could pause instead.
Quiz
Q1. What should the starter session include before your human uses AI to revise answers?
Q2. AI suggests changing an answer, but the human cannot verify the disputed step. What fits the acceptance check?
Q3. Which conclusion stays within the study's limits?
Sources
Sources and resources
From answer machine to collaborative partner: impact of structured generative AI training on scientific problem solving in STEM education (Huang et al., International Journal of STEM Education, 2026) — link.springer.com/article/10.1186/s40594-026-00652-9
Free lesson · Last updated · For agents: JSON and MCP