AI Prompts for Students

Start with a clear task

There's an obvious way to use AI as a student and it's the one that doesn't work: having it produce the answer. You get the assignment done and learn nothing, which shows up later in exams where the tool isn't available.

The useful version inverts it. Learning research consistently finds that retrieval — being tested — builds durable memory far better than re-reading, and that explaining a concept exposes the gaps in your understanding faster than reviewing notes does.

Every prompt below puts AI in the role of tutor, examiner, or critic rather than author. They're harder to use than "write my essay," and that's exactly why they work.

The technique: make it quiz you, not tell you

Flip the direction of information. Instead of asking the model to explain a topic to you (passive), ask it to test you on it and correct your answers (active). The difficulty is the point — the effort of retrieval is what forms the memory.

The strongest version is the Feynman method: you explain the concept in your own words, and the model identifies exactly where your explanation breaks down. That's feedback you can't easily get from a textbook.

• Weak prompt

  Explain photosynthesis to me.

• Strong prompt

  I'm going to explain photosynthesis to you in my own words. Don't correct me until I'm finished. Then: point out exactly which parts were wrong or vague, which important steps I skipped entirely, and ask me two questions that target the gaps you found. Here's my explanation: [your explanation]

• Why it works

  Reading an explanation feels productive and produces almost no retention. Attempting an explanation and getting it critiqued surfaces precisely what you don't know — which is the only thing worth studying.

The prompts

• Feynman check

  I'll explain [concept] in my own words. Don't correct me until I finish. Then tell me exactly what was wrong, what was vague, and what I skipped — then ask me two questions targeting those gaps. My explanation: [paste]

• Quiz me

  Quiz me on [topic] at [level]. Ask one question at a time and wait for my answer before the next. Start easy and get harder based on how I do. After each answer, tell me if I'm right and why — briefly.

• Exam-style questions

  Write [N] exam-style questions on [topic] matching the format of [exam/course]. Include mark allocations. Don't give me the answers yet — I'll attempt them first.

• Mark my answer

  Here's an exam question: [paste]. Here's my answer: [paste]. Mark it as an examiner would: what marks would I get, what did I miss, and what specifically would move this to full marks?

• Find my gaps

  Here are my notes on [topic]: [paste]. What important concepts are missing or under-covered relative to a solid understanding of this topic? Don't rewrite my notes — just tell me what to go study.

• Spaced-repetition cards

  Turn this material into flashcards: [paste]. Each card: one specific question, one concise answer. Test understanding, not just recall of definitions. Format as Q: / A: pairs.

• Concept connections

  How do [concept A] and [concept B] relate in [subject]? Where do students commonly confuse them, and what's the clearest way to tell them apart?

• Worked example walkthrough

  Walk me through solving this problem step by step: [paste]. At each step, explain WHY that step, not just what it is. Then give me a similar problem to try on my own.

• Essay structure critique

  Here's my essay plan for [question]: [paste]. Does the structure actually answer the question asked? Where's the argument weak? What counter-argument have I ignored? Don't write it for me.

• Source evaluation

  I'm using these sources for [topic]: [list]. For each, what are the likely limitations — bias, age, methodology, scope? What type of source am I missing?

• Simplify then deepen

  Explain [difficult concept] at three levels: to a 12-year-old, to an undergraduate, and to a specialist. Then tell me which details the simpler versions had to leave out.

• Misconception check

  What are the most common misconceptions about [topic]? For each, explain why it's intuitive but wrong, and what the correct understanding is.

• Study plan

  I have [time] until my exam on [subject], covering [topics]. I'm weakest at [topics]. Build a study schedule that uses spaced repetition and prioritizes my weak areas, with specific activities per session — not just 'revise chapter 3'.

• Research question refinement

  My research question is: [paste]. Is it specific enough to answer in [length/scope]? Is it actually arguable? Suggest three sharper versions and tell me what each would require me to find.

• Paper comprehension

  Here's a paper abstract/section: [paste]. What's the core claim, what evidence supports it, and what's the main limitation the authors acknowledge? Then: what should I be skeptical about?

Common mistakes

• Having it write the assignment

  Beyond the academic-integrity problem, you don't learn the material — which surfaces in every closed-book assessment afterwards.

• Reading explanations instead of attempting them

  Reading a clear explanation feels like learning and mostly isn't. Attempt first, then check.

• Trusting factual claims blindly

  Models state incorrect facts fluently. For anything you'll be assessed on, verify against your course material or a primary source.

• Skipping the struggle

  Asking for the answer the moment something is hard removes the exact effort that builds memory. Sit with it, then ask for a hint rather than a solution.

Frequently asked questions