{"course":{"slug":"ai-for-students","title":"AI for Students","url":"https://addaly.com/learn/ai-for-students","level":"beginner","track":"work"},"lesson":{"slug":"two-kinds-of-help","title":"Two Kinds of Help","url":"https://addaly.com/learn/ai-for-students/two-kinds-of-help","position":1,"of":73,"minutes":7,"answer":"After a study session, ask what changed: the document, or you.","keyIdea":"After a study session, ask what changed: the document, or you."},"body":{"format":"markdown","text":"## The question that actually decides it\n\nMost advice about AI and study starts with rules. Allowed, not allowed, disclose it, don't. That is a real question, and we cover it later. But it is the wrong *first* question, because the answer tells you nothing about what happened to you.\n\nHere is a better one.\n\n**After the session, what changed: the document, or you?**\n\nBoth kinds of help produce a finished assignment. Only one produces a person who could do it again.\n\n## The same tool, two directions\n\nSame subject. Same chatbot. Same ten minutes.\n\n| Avoiding the work | Doing the work |\n|---|---|\n| \"Solve question 4.\" | \"I got x = 3, the answer key says x = −1. Here is my working. Which line is wrong? Don't tell me why yet.\" |\n| \"Write an introduction on the causes of the Nigerian Civil War.\" | \"Here is my introduction. Argue against my thesis as harshly as an examiner would.\" |\n| \"Explain recursion.\" | \"Explain recursion, then give me three problems and withhold the solutions.\" |\n| \"Summarise chapter 6.\" | \"I've read chapter 6. Ask me six questions on it, one at a time, and don't accept a vague answer.\" |\n\nThe right column is not slower because it is more virtuous. It is slower because you are the one doing the expensive part, and the expensive part is the point.\n\n## The load-bearing part\n\nEvery assignment has one thing it exists to build. Find it, and you know what not to hand over.\n\n- In a problem set, the load-bearing part is the solving. Not the arithmetic, not the neat presentation.\n- In an essay, it is the argument: which claims you make, in what order, and what you leave out.\n- In a lab report, it is the interpretation. Not the method section you have written eleven times.\n- In a translation exercise, it is choosing between two defensible renderings.\n\nEverything else is scaffolding, and scaffolding is fair game. Reformatting a bibliography into APA is not a skill anybody is examining. Turning your bullet points into a table is not either. If you spend the saved hour on the load-bearing part, the tool has done its job.\n\n## The transfer test\n\nHere is a check you can run today, and it costs ten minutes.\n\nFinish a piece of work with AI help. Close the laptop. Wait ten minutes: get water, walk somewhere. Then take blank paper and redo the central thing: solve the problem again, or write your thesis and its three supports from memory.\n\nWhatever comes back is yours. Whatever does not, you watched somebody else do.\n\nThis is uncomfortable the first time, because the gap is usually much bigger than it felt. Reading a clear explanation produces a strong, immediate feeling of understanding. That feeling is real, and it is not knowledge. Cognitive psychologists call it the fluency illusion, and AI is very good at generating it, because AI is very good at being clear.\n\n## Neither extreme survives contact with reality\n\nRefusing to touch these tools is not integrity, it is just a tax you pay in hours. A student in Lagos or Lima or Lucknow with no tutor, no office hours, and a class of 300 has something now that did not exist before: a patient thing that will explain the same idea six different ways at two in the morning. Telling that student to switch it off is not advice, it is a lecture.\n\nAnd going the other way (running every task through the model and shipping the output) reliably produces someone who cannot tell when the output is wrong. That is not a moral failure. It is a mechanical one, and we will look at the evidence in the next lesson.\n\nThe useful position is in the middle and it is specific: **delegate the scaffolding, keep the load-bearing part, and test yourself on paper afterwards.**\n\n## One habit to start with\n\nBefore you type anything into a chat box, finish this sentence out loud:\n\n> \"I am asking for this because I want to be able to _____ afterwards.\"\n\nIf the honest ending is \"submit it,\" you already know what kind of help this is. That is not always the wrong answer: some weeks are like that. But say it plainly to yourself, because the cost is real and it arrives later, in a room where the tool is not allowed."},"check":{"question":"Priya asks a chatbot to solve five calculus problems and reads each solution until every step makes sense. Ravi solves the same five himself, gets three wrong, then asks only what went wrong with his method. A week later there is an unassisted test. Why is Ravi likely to do better?","options":[{"text":"Priya's mistake was not putting the solutions into her own words, so none of it became hers.","correct":false,"tempting":"Treats the problem as one of paraphrasing and ownership, as if learning were a plagiarism question rather than a memory one."},{"text":"Ravi got three wrong, and mistakes are what teach; had he solved all five correctly he would have gained nothing.","correct":false,"tempting":"Over-reads the advice to learn from mistakes, making error the mechanism when the mechanism is effortful retrieval, right or wrong."},{"text":"Ravi produced the steps himself, and producing something under difficulty is what makes it available later; Priya recognised steps she never generated.","correct":true},{"text":"Priya did not read deeply enough. Had she studied each solution until it was completely clear, she would be equally prepared.","correct":false,"tempting":"Assumes clarity while reading equals knowing, which is exactly the fluency illusion a well-written solution creates."}]},"prev":null,"next":{"slug":"what-happens-to-your-memory","title":"What Happens To Your Memory","url":"https://addaly.com/learn/ai-for-students/what-happens-to-your-memory"},"authorship":"Founder and series editor: Dr Mustafa Mun. Written by AI models to a written standard. Exam keys checked blind by a second model: 3,665 of 3,666 agreed. Academic review invited.","review":{"status":"Draft, open for academic review"},"cite":"Addaly (2026) 'Two Kinds of Help', in AI for Students. Edited by M. Mun. Addaly. Available at: https://addaly.com/learn/ai-for-students/two-kinds-of-help (Accessed: 31 October 2026).","licence":{"name":"CC BY-NC-SA 4.0","url":"https://creativecommons.org/licenses/by-nc-sa/4.0/"},"reader":{"schema":1,"title":"Two Kinds of Help","courseTitle":"AI for Students","module":{"index":1,"slug":"what-the-machine-is-doing","title":"What the machine is doing when it helps you"},"answer":null,"keyIdea":"After a study session, ask what changed: the document, or you.","hero":{"kind":"plate","image":null,"alt":null,"caption":null},"blocks":[{"t":"h2","id":"the-question-that-actually-decides-it","runs":[{"text":"The question that actually decides it"}]},{"t":"p","runs":[{"text":"Most advice about AI and study starts with rules. Allowed, not allowed, disclose it, don’t. That is a real question, and we cover it later. But it is the wrong "},{"text":"first","i":true},{"text":" question, because the answer tells you nothing about what happened to you."}]},{"t":"p","runs":[{"text":"Here is a better one."}]},{"t":"p","runs":[{"text":"After the session, what changed: the document, or you?","b":true}]},{"t":"p","runs":[{"text":"Both kinds of help produce a finished assignment. Only one produces a person who could do it again."}]},{"t":"h2","id":"the-same-tool-two-directions","runs":[{"text":"The same tool, two directions"}]},{"t":"p","runs":[{"text":"Same subject. Same chatbot. Same ten minutes."}]},{"t":"table","head":[[{"text":"Avoiding the work"}],[{"text":"Doing the work"}]],"align":["left","left"],"rows":[[[{"text":"“Solve question 4.”"}],[{"text":"“I got x = 3, the answer key says x = −1. Here is my working. Which line is wrong? Don’t tell me why yet.”"}]],[[{"text":"“Write an introduction on the causes of the Nigerian Civil War.”"}],[{"text":"“Here is my introduction. Argue against my thesis as harshly as an examiner would.”"}]],[[{"text":"“Explain recursion.”"}],[{"text":"“Explain recursion, then give me three problems and withhold the solutions.”"}]],[[{"text":"“Summarise chapter 6.”"}],[{"text":"“I’ve read chapter 6. Ask me six questions on it, one at a time, and don’t accept a vague answer.”"}]]]},{"t":"p","runs":[{"text":"The right column is not slower because it is more virtuous. It is slower because you are the one doing the expensive part, and the expensive part is the point."}]},{"t":"h2","id":"the-load-bearing-part","runs":[{"text":"The load-bearing part"}]},{"t":"p","runs":[{"text":"Every assignment has one thing it exists to build. Find it, and you know what not to hand over."}]},{"t":"ul","items":[[{"text":"In a problem set, the load-bearing part is the solving. Not the arithmetic, not the neat presentation."}],[{"text":"In an essay, it is the argument: which claims you make, in what order, and what you leave out."}],[{"text":"In a lab report, it is the interpretation. Not the method section you have written eleven times."}],[{"text":"In a translation exercise, it is choosing between two defensible renderings."}]]},{"t":"p","runs":[{"text":"Everything else is scaffolding, and scaffolding is fair game. Reformatting a bibliography into APA is not a skill anybody is examining. Turning your bullet points into a table is not either. If you spend the saved hour on the load-bearing part, the tool has done its job."}]},{"t":"h2","id":"the-transfer-test","runs":[{"text":"The transfer test"}]},{"t":"p","runs":[{"text":"Here is a check you can run today, and it costs ten minutes."}]},{"t":"p","runs":[{"text":"Finish a piece of work with AI help. Close the laptop. Wait ten minutes: get water, walk somewhere. Then take blank paper and redo the central thing: solve the problem again, or write your thesis and its three supports from memory."}]},{"t":"p","runs":[{"text":"Whatever comes back is yours. Whatever does not, you watched somebody else do."}]},{"t":"p","runs":[{"text":"This is uncomfortable the first time, because the gap is usually much bigger than it felt. Reading a clear explanation produces a strong, immediate feeling of understanding. That feeling is real, and it is not knowledge. Cognitive psychologists call it the fluency illusion, and AI is very good at generating it, because AI is very good at being clear."}]},{"t":"h2","id":"neither-extreme-survives-contact-with-reality","runs":[{"text":"Neither extreme survives contact with reality"}]},{"t":"p","runs":[{"text":"Refusing to touch these tools is not integrity, it is just a tax you pay in hours. A student in Lagos or Lima or Lucknow with no tutor, no office hours, and a class of 300 has something now that did not exist before: a patient thing that will explain the same idea six different ways at two in the morning. Telling that student to switch it off is not advice, it is a lecture."}]},{"t":"p","runs":[{"text":"And going the other way (running every task through the model and shipping the output) reliably produces someone who cannot tell when the output is wrong. That is not a moral failure. It is a mechanical one, and we will look at the evidence in the next lesson."}]},{"t":"p","runs":[{"text":"The useful position is in the middle and it is specific: "},{"text":"delegate the scaffolding, keep the load-bearing part, and test yourself on paper afterwards.","b":true}]},{"t":"h2","id":"one-habit-to-start-with","runs":[{"text":"One habit to start with"}]},{"t":"p","runs":[{"text":"Before you type anything into a chat box, finish this sentence out loud:"}]},{"t":"quote","runs":[{"text":"“I am asking for this because I want to be able to _____ afterwards.”"}]},{"t":"p","runs":[{"text":"If the honest ending is “submit it,” you already know what kind of help this is. That is not always the wrong answer: some weeks are like that. But say it plainly to yourself, because the cost is real and it arrives later, in a room where the tool is not allowed."}]}],"check":{"question":[{"text":"Priya asks a chatbot to solve five calculus problems and reads each solution until every step makes sense. Ravi solves the same five himself, gets three wrong, then asks only what went wrong with his method. A week later there is an unassisted test. Why is Ravi likely to do better?"}],"options":[{"text":[{"text":"Priya’s mistake was not putting the solutions into her own words, so none of it became hers."}],"correct":false,"tempting":[{"text":"Treats the problem as one of paraphrasing and ownership, as if learning were a plagiarism question rather than a memory one."}]},{"text":[{"text":"Ravi got three wrong, and mistakes are what teach; had he solved all five correctly he would have gained nothing."}],"correct":false,"tempting":[{"text":"Over-reads the advice to learn from mistakes, making error the mechanism when the mechanism is effortful retrieval, right or wrong."}]},{"text":[{"text":"Ravi produced the steps himself, and producing something under difficulty is what makes it available later; Priya recognised steps she never generated."}],"correct":true,"tempting":null},{"text":[{"text":"Priya did not read deeply enough. Had she studied each solution until it was completely clear, she would be equally prepared."}],"correct":false,"tempting":[{"text":"Assumes clarity while reading equals knowing, which is exactly the fluency illusion a well-written solution creates."}]}],"order":[1,2,0,3],"why":null},"prev":null,"next":{"slug":"what-happens-to-your-memory","title":"What Happens To Your Memory","minutes":8}}}