Story · Sanjay's Debate Disaster
The Argument That Sounded Perfect — Until It Wasn't 🗣️
Sanjay, 13, from Pune, had a class debate on whether "social media is harmful for teenagers." He asked an AI: "Give me 5 strong arguments that social media is harmful for teenagers." The AI gave him five well-written arguments with statistics.
Sanjay prepared confidently. In the debate, he used every argument. His opponent — Nitika — asked about his statistics. Sanjay had used one: "Studies show 70% of teenagers who use social media for more than 2 hours a day report feeling anxious."
Nitika asked: "Which study? When was it done? Who were the participants — was it in India or abroad? What does 'feel anxious' mean exactly? Is social media the cause or was it already anxious teenagers who use social media more?"
Sanjay could not answer any of these questions. He had not verified the statistic. He did not know if it was accurate. His argument collapsed — not because the topic was wrong, but because he had accepted AI's output without thinking critically about it.
👉 In this lesson you will learn the critical thinking habits that make you a stronger student — not just for AI, but for everything you read, hear, and argue.
Section 1 of 8
🤔 What Is Critical Thinking?
Critical thinking is the skill of examining information carefully before accepting or using it. It means asking questions, looking for evidence, considering other points of view, and forming your own reasoned judgement — rather than simply accepting what you are told.
Critical thinking is not the same as being negative or disagreeable. A critical thinker does not reject everything. They evaluate — some things they accept, some they question, some they reject — based on evidence and reasoning.
Why critical thinking matters more with AI: AI tools produce fluent, confident-sounding text. The better AI gets at writing, the more natural its outputs feel — and the easier it is to accept them without checking. Critical thinking is the skill that keeps you in control of AI, rather than AI controlling what you think.
The 5 core habits of a critical thinker:
1
Question claims — especially confident ones
The more confident a statement sounds, the more it is worth asking: "How do we know this? What is the evidence?"
2
Ask for sources and check them
A claim without a source is an opinion. A source you have not checked is no better than no source at all.
3
Consider the other side
Any strong argument has a counter-argument. Understanding both sides is better than knowing only one.
4
Separate correlation from causation
Two things happening together does not mean one causes the other. "Students who read more score better in exams" — did reading cause better scores, or do students who enjoy studying both read more and study more?
5
Stay open to changing your mind
Critical thinkers update their views when new evidence arrives. Stubbornly holding a position despite evidence is not strength — it is a thinking error.
Section 2 of 8
❓ The 5 Questions to Ask Every AI Output
When AI gives you information, there are five standard questions that will protect you from being misled — whether the AI is wrong accidentally or simply presenting one-sided information.
Question 1
Is this a fact, an opinion, or a mix of both?
- Facts can be verified. Opinions are judgements — reasonable people can disagree.
- AI often presents opinions with the same confident tone as facts. Watch for words like "most people believe," "studies suggest," "it is widely agreed" — these are signals that what follows may be opinion or contested.
- Ask AI: "Is what you just said a fact that can be verified, or an opinion?"
Question 2
What is the source? Is it reliable?
- AI does not always cite sources. When it does, the citation may be fabricated (hallucination).
- For any important claim, ask AI: "What is the source for this? Can you give me the study name, author, and year?"
- Then verify the source exists — search for it. If it does not appear in a search, treat the claim as unverified.
Question 3
Is this the whole picture — or is one side missing?
- If you ask AI "What are the arguments for X?", it will give arguments for X. It may not tell you the strong counter-arguments unless you ask.
- For any topic where you need balance, ask both sides: "What are the strongest arguments FOR this?" and "What are the strongest arguments AGAINST this?"
- A strong argument only works if it can survive the strongest counter-argument. Does yours?
Question 4
Is this recent enough? Has the situation changed?
- AI training data has a cutoff date. For anything that changes quickly — statistics, policies, current events, prices, technology — the AI's information may be outdated.
- Ask: "When was this information current? Has there been any significant change since [year]?"
- For recent topics, verify with a current news source or official website.
Question 5
Does this apply to MY context — India, my age, my situation?
- Much of AI's training data is from Western, English-language sources. Statistics, examples, and advice may not apply to Indian students, Indian laws, or Indian cultural contexts.
- Ask: "Is this information specific to India, or is it from a Western/global context?"
- For anything involving Indian law, government schemes, or cultural practices — cross-check with an Indian government website or local source.
Section 3 of 8
🕳️ Spotting Weak Reasoning — Common AI Fallacies
A fallacy is a flaw in reasoning — an argument that sounds logical but has a hidden error. AI can produce fallacious arguments fluently because it has learned from a huge amount of human writing, and humans frequently use fallacies. Knowing the most common ones helps you spot them.
False Dilemma
"Students either focus entirely on academics or they waste their time. Since Riya does sports, she must not take studies seriously."
Spot it: "Either/or" framing that ignores other possibilities. Real life usually has more options than two.
Hasty Generalisation
"Three students in your school failed after using AI. Therefore AI tools make students fail."
Spot it: Drawing a broad conclusion from a very small number of examples. Ask: "Is this sample large enough and representative?"
Correlation ≠ Causation
"Students who use smartphones more get lower grades. Therefore smartphones cause lower grades."
Spot it: Two things happening together is not the same as one causing the other. Ask: "Could a third factor explain both?"
Appeal to Popularity
"Most students cheat in exams, so it cannot be that harmful."
Spot it: Something being common or popular does not make it right or safe. Ask: "Is 'everyone does it' actually an argument?"
Slippery Slope
"If students are allowed to use AI for homework, soon they will use AI for everything and never learn anything."
Spot it: Claiming one small step will inevitably lead to an extreme outcome, without showing why each step must follow. Ask: "Why would each step necessarily lead to the next?"
Straw Man
"Those who support AI in schools want students to stop thinking for themselves entirely."
Spot it: Misrepresenting an opponent's position to make it easier to attack. Ask: "Is this actually what the other side says?"
✅ Fallacy Check Prompt
Here is an argument: "[paste the argument]"
Does this argument contain any logical fallacies? List each fallacy by name, explain why it is a fallacy in this case, and suggest how the argument could be made stronger without the fallacy.
Asking AI to check its own arguments — or arguments from other sources — for logical fallacies. AI is good at recognising fallacy patterns when asked explicitly.
Section 4 of 8
🔍 How to Verify AI Claims — A Practical Workflow
Not every AI statement needs to be verified — if AI explains how photosynthesis works, you can cross-check it quickly with your textbook. But when AI gives statistics, specific claims, or information that matters, you need a verification workflow.
1
Identify the specific claim
Underline or highlight the specific factual claim you want to verify. "Studies show..." or "According to..." are signals to look for.
2
Ask AI for its source
Prompt: "You said [claim]. What is the specific source — study name, author, organisation, or report?" Check if the source sounds real.
3
Search for the source independently
Search Google Scholar, government websites (gov.in), NCERT, WHO, or trusted news sources. If the source cannot be found at all — treat the claim as unverified.
4
Check a second independent source
One source can be wrong, biased, or misrepresented. A claim confirmed by two independent, reliable sources is much more trustworthy than one.
5
Check whether the claim is still current
Look for the publication date of the source. For fast-changing topics, a 5-year-old study may already be outdated. Check for more recent data.
Source quality tiers for Indian students:
High trust
Indian government websites (.gov.in), NCERT, PIB (Press Information Bureau), WHO, peer-reviewed academic journals, trusted newspapers (The Hindu, Indian Express, Times of India for news; use with care for opinion)
Check carefully
Wikipedia (good starting point — check citations at the bottom), educational websites with named authors and publication dates, NGO and research organisation reports (check funding and methodology)
Use with caution
Social media posts, WhatsApp forwards, anonymous websites, sites with no author or date, sites that do not cite their sources, AI-generated content (including this lesson — check important claims independently)
Even this lesson: When you use information from this lesson in a school project, confirm important claims with a primary source. "My AI class said..." is not a citation. The habit of checking applies to everything — including educational content you enjoy and trust.
Section 5 of 8
📰 Spotting Misleading Information and Fake News
AI can generate text that sounds like a news article or a factual report — but is completely made up. And AI can also summarise real news in ways that distort the meaning. Knowing how to spot misleading information is essential.
Warning signs of misleading information:
- Emotional language designed to make you angry or afraid — "Shocking! Scientists PROVE..." or "Government HIDES the truth about..."
- No author, date, or source cited — real journalism always includes at least author and publication date
- Statistics without context — "Crime increased by 200%" could mean 1 case became 3 cases
- One-sided presentation only — no acknowledgement of any counter-argument or complexity
- A claim that seems too perfectly aligned with what you already believe — confirmation bias means we are more willing to accept claims that fit our existing views
- The story appears on only one website — real significant news gets covered by multiple independent sources
✅ Misinformation Check Prompt
Here is a claim I have seen: "[paste the claim or news story]"
Please evaluate this claim:
1. Is this a verifiable fact, an opinion, or speculation?
2. What would a reliable source for this claim look like?
3. Are there any warning signs of misleading information in how this claim is presented?
4. What would someone who disagrees with this claim say?
AI can help you evaluate claims — but remember that AI's evaluation also needs to be checked. Use AI as a starting point for your verification, not the final word.
The critical thinker's rule about AI and news: AI is a poor source for recent news — its training data has a cutoff. Never use AI to confirm whether a recent news story is true. For current events, check the original source (the government website, the news publisher, the research institution). AI can help you understand context and history — not breaking news.
Section 6 of 8
🗣️ Using Critical Thinking in Debates and Essays
Critical thinking is not just about fact-checking — it is the foundation of strong arguments in debates, essays, and class discussions. Here is how to apply it specifically to school debate and essay work with AI.
Before the debate/essay — preparation:
- Ask AI for arguments on BOTH sides — not just the side you are defending
- For each argument AI gives you, ask: "What is the strongest counter-argument to this?"
- Verify any statistics before using them — know the source, sample size, and year
- Ask AI to identify the weakest point in your argument so you can strengthen it before the debate
✅ Debate Preparation Prompt
Topic: [your debate topic]
My position: [which side you are arguing]
Please help me:
1. Give me 3 strong arguments for MY side, with an indication of what evidence would support each
2. Give me the 2 strongest counter-arguments I should expect from the OTHER side
3. For each counter-argument, suggest how I might respond
4. Identify the single weakest point in my side's argument — the one a critical opponent would attack first
This prompt forces you to think about both sides AND the weakest points in your own argument — exactly what Sanjay failed to do before his debate.
In a debate — responding to the other side:
- When the opponent makes a claim, ask yourself: "Do I know if this is verified? What is the source?"
- Ask the question Nitika asked Sanjay: "Which study? When? What country? How large was the sample?"
- If they use a fallacy, name it: "That sounds like a hasty generalisation from three examples."
- If you cannot disprove a claim, say so honestly: "I would want to verify that before accepting it" — this shows strength, not weakness
Section 7 of 8
🧩 Critical Thinking in Everyday AI Use
Critical thinking is not just for debates. Every time you use AI for study help, the same habits apply — just at a smaller scale.
The 30-second critical thinking habit for daily AI use: Every time AI gives you an important piece of information, spend 30 seconds asking these three questions before you use it:
- Does this make sense? Does it fit with what I already know about this subject?
- Can I check this quickly? What would I search to verify it?
- Am I confident enough to say this in front of my teacher? If not, verify first.
Situations where critical thinking is most important:
- Statistics and numbers — AI fabricates statistics more often than other types of content
- Historical dates and events — AI often gets specific dates, names, and sequences wrong
- Current events and news — AI's training data has a cutoff; recent information may be wrong or missing
- Medical and health information — never rely on AI alone for health decisions; check with a doctor
- Legal information — Indian law is complex and AI may not know state-specific rules
- Anything you plan to argue publicly — in a debate, an essay, or a class presentation
Situations where you can trust AI more readily:
- Explaining well-established scientific concepts (photosynthesis, Newton's laws) — cross-check with textbook
- Mathematical procedures — check by working the problem yourself
- Grammar and language questions — still read the suggestion critically
- Creative brainstorming — AI suggestions are starting points, not final answers
Section 8 of 8
📋 Summary — The Critical Thinker's AI Toolkit
| Situation | Critical thinking action | Key question to ask |
|---|---|---|
| AI gives a statistic | Ask for source; verify independently | "Which study? When? How large a sample?" |
| AI's argument sounds one-sided | Ask for counter-arguments explicitly | "What are the strongest arguments against this?" |
| AI's claim sounds very confident | Question it more, not less | "Is this a fact or an opinion? How do we know?" |
| AI gives information about India | Check if it applies to Indian context | "Is this specific to India, or from a Western source?" |
| AI discusses recent events | Verify with a current source | "Is this information still current?" |
| You are about to use AI output publicly | Verify every specific claim first | "Could I defend this if someone asked 'How do you know?'" |
| AI's reasoning seems off | Name the specific problem | "Is this a fallacy? Which one?" |
Sanjay's second debate: Next time, Sanjay used the debate preparation prompt — and asked AI to find his weakest argument and the strongest counter-argument against each point. He verified every statistic before using it. In the debate, when Nitika challenged a statistic, Sanjay could say: "That comes from a 2023 WHO report on adolescent mental health — I can share the reference." He won the debate.
🧠 Quick Quiz — Test Yourself!
10 questions · Click your answer · Check your score at the end
1. What was Sanjay's critical error before his debate?
2. Which of these is the best definition of critical thinking?
3. AI says: "Studies show students who eat breakfast score 15% higher in exams." What should you do FIRST?
4. "Students who sleep less than 6 hours score lower in exams. Therefore lack of sleep causes lower exam scores." What is the problem with this reasoning?
5. You ask AI: "Give me arguments that social media is harmful for teenagers." AI gives you 5 arguments. What critical thinking step is most important NEXT?
6. "Either students work hard and study all the time, or they are wasting their potential." This is an example of which fallacy?
7. Which of these is the MOST reliable source for verifying a claim about Indian government education policy?
8. AI gives you a very confident-sounding claim. A critical thinker should:
9. "Those who support allowing AI in homework want students to stop thinking altogether." This is an example of:
10. What is the best 30-second habit to apply every time AI gives you important information?
📝 Worksheet — My Critical Thinking Debate Prep
Tip: in the print dialog, choose "Save as PDF" to download.Choose a debate topic from your class. Use the steps below to prepare. Copy into your notebook.
| Step | Task | Your answer |
|---|---|---|
| 1 | My debate topic and position | |
| 2 | 3 arguments for my side (note one verifiable source for each) | |
| 3 | 2 strongest counter-arguments I expect | |
| 4 | My response to each counter-argument | |
| 5 | My weakest argument — and how I can strengthen it | |
| 6 | One fallacy I should watch out for in my own reasoning |
💡 Use the Debate Preparation Prompt from Section 6 to get AI to help you fill in this table — then verify the statistics it suggests.
📋 Note for Parents and Teachers
What this lesson teaches: Students learn the 5 core habits of critical thinkers, the 5 questions to ask every AI output, 6 common logical fallacies (false dilemma, hasty generalisation, correlation vs causation, appeal to popularity, slippery slope, straw man), a 5-step claim verification workflow, source quality tiers for Indian students, how to spot misleading information, and how to use critical thinking specifically for debates and essays.
Discussion prompts for families:
- "Can you spot a fallacy in today's news?" — Try finding one news story or advertisement together and applying the fallacy list.
- "When someone says 'studies show...' — what questions should we ask?" (Which study? When? How many people? What country?)
- "What is one thing you learned today that you want to verify?" — Make independent verification a family habit.
For teachers: The logical fallacies section pairs directly with English debate and argument writing units. The 6 fallacy cards could be printed as classroom posters. The Debate Prep worksheet works as a structured pre-debate planning activity — students complete it before any debate and bring it to class.