How to actually study lecture slides with AI (without just cheating yourself)

Lecture slides were never designed to be studied from. They're speaker notes — a visual aid for someone who was going to talk over them for 50 minutes. That's why they're full of half-sentences, unlabelled diagrams, and lines like “proof left as exercise” or “see seminar 4”. The explanation lived in the lecture theatre, and if you missed it (or it went too fast, or the lecturer mumbled), the slides alone won't give it back to you.

So when exam season arrives, most students end up staring at a slide that says “E(Rᵢ) = R𝒇 + βᵢ(E(Rₘ) − R𝒇)” with no idea what any of it means in words, and three bad options: rewatch the recording at 1× speed, trawl Google for an explanation that uses different notation, or just memorise the shape of the formula and hope. AI gives you a fourth option — but only if you use it properly.

Here's the failure mode. You paste your slides into a chatbot, get a beautifully clear summary, read it, nod, and feel like you understand. Cognitive scientists call this the illusion of competence: reading a fluent explanation produces a feeling of knowing that evaporates the moment you have to produce the answer yourself. Recognising an explanation and generating one are different skills, and exams only test the second.

This is also where the “AI is cheating yourself” worry is actually right. Not because getting an explanation is cheating — asking your lecturer to explain a slide again isn't cheating either. It's that outsourcing the effortful part — the retrieval, the self-testing, the struggle to put it in your own words — means the knowledge never becomes yours. You end up with a folder of perfect summaries and nothing in your head.

The fix is to use AI for what it's uniquely good at — instant, patient explanation of your specific material — while keeping the effortful, memory-building work for yourself. Three passes:

1. 01Explain. Go through the deck slide by slide and get every confusing slide explained in plain language. Don't skip ahead to summaries of the whole topic — work at the slide level, because that's the level your lecturer chose to emphasise things at.
2. 02Interrogate. For anything that still feels fuzzy, ask follow-ups until it clicks: “why does beta measure that?”, “what happens if this assumption breaks?”, “how does this connect to last week's lecture?”. This is the step most students skip, and it's where understanding actually forms.
3. 03Test. Close your notes and answer questions on the material — ideally questions generated from the slides themselves. Getting things wrong here is the point: every failed retrieval followed by a correction is worth more than three re-reads.

The order matters. Testing before you've understood produces frustration; explaining without testing produces the illusion of competence. Explain → interrogate → test is the loop that converts someone else's slides into your own knowledge.

Use AI for:

• Decoding dense or cryptic slides into plain language, in your module's own notation
• Filling in the steps a derivation or argument skipped
• Generating quiz questions and checking your answers
• Connecting a slide to earlier lectures ('where does this fit?')
• Translating terminology — especially valuable if English is your second language

Keep for yourself:

• Answering the quiz questions — from memory, before looking at the answer
• Writing the one-line summary of each lecture in your own words
• Deciding what's likely to be examined (your lecturer's emphasis beats any AI's guess)
• Anything that will be assessed directly — essays, problem sets, coursework. Understanding the material with AI is studying; submitting AI's words is not.

AI should remove the friction of studying, not the effort. Friction is decoding a slide written for someone else. Effort is retrieving it from your own memory — and that part has to stay yours.

1. 01Within 48 hours of the lecture (20 min): run the deck through the explain pass. Flag the slides you couldn't have explained yourself.
2. 02Same session (10 min): interrogate the flagged slides until they click. Write one sentence per flagged slide in your own words.
3. 03End of week (10 min): quiz yourself on the week's lectures, notes closed. Anything you fail goes on a list you re-test next week.

That's it. No four-hour Sunday note-rewriting session, no 1× recording marathons. Done weekly, this means exam season starts with material you already understand — and revision becomes practice instead of first contact. If you're already behind, see our guide to getting through dense lecture content before exams.