General Accounting Won't Stick? Let AI Run the Feynman Technique With You

It is the night before your accounting exam. You have read the chapter on adjusting entries three times. You can nod along to every example. Then you close the book and try to explain why a prepaid insurance payment becomes an expense at month end, and your explanation collapses after two sentences. That collapse is not a sign you are bad at accounting. It is the exact moment the Feynman technique is built to catch.

Why explaining out loud exposes what reading hides

Reading feels like learning because the sentences make sense while your eyes move across them. Explaining is different. To say a rule in your own words, your brain has to rebuild it from parts, and any missing part shows up as a stumble, a filler word, or a sudden jump to jargon. That stumble is useful information, not failure.

An AI assistant fits here because it can play the role of a patient listener who never gets tired and never pretends to understand. You explain, it asks one short question about the weakest phrase, you answer, and the loop continues. Short turns matter. A wall of text from the assistant lets you hide again; one question forces you to produce the next sentence yourself.

The naive attempt versus the calibrated one

The naive prompt

Explain adjusting entries in accounting.

That paragraph is accurate and it teaches you almost nothing. It is a summary you could have read in the textbook, and it lets you stay passive. You did not say a single word in your own voice, so no gap surfaced. The assistant did the explaining, which is the opposite of what the method needs.

The calibrated prompt

You are my accounting study partner running the Feynman technique. I will explain a concept in my own words. Your job: listen, then ask exactly one short question about the vaguest phrase I used. Do not explain the concept for me. Do not summarize. Do not praise me. If my explanation is correct but shallow, ask a question that forces a concrete example with numbers. One question per turn, then stop and wait for my reply. Start by asking me to explain adjusting entries.

Now the roles are reversed. You talk, the assistant listens, and the next turn depends on what you actually said. If you mumble that adjusting entries "fix the numbers at the end," the assistant can ask which numbers and why they were wrong before. That single question sends you back into the material with a specific target, which is how the gap closes.

A template you can paste right now

Replace the bracketed parts with your own topic and level. Keep the rules about one question and no praise; those are what stop the assistant from drifting back into lecture mode.

Act as my Feynman study partner for [ACCOUNTING TOPIC, e.g. bank reconciliation]. I am studying at [LEVEL, e.g. first-year university]. Rules: I explain first, in plain words. You ask one short question about the weakest part of my explanation. No lectures, no summaries, no compliments. If I use a technical term without defining it, ask me to define it. If my example has no numbers, ask for numbers. One question per turn, then wait. Begin by asking me to explain [TOPIC].

Quick fixes when the assistant slips

• If it starts lecturing again: "Stop. Ask me one question instead of explaining."
• If it accepts a vague answer: "My last sentence was fuzzy. Point at the fuzzy word and question it."
• If it piles on praise: "Drop the encouragement. Ask the next question only."

How you know it worked, and the voice trick

You know the loop worked when you can explain the rule to a friend who has never opened an accounting book, using a real number, without reaching for a term you cannot define. If you still need the textbook phrase to feel safe, the gap is still there.

Then try it out loud. The free voice mode in the mobile app lets you speak your explanation and hear the assistant's question back, which is closer to the pressure of a real exam than typing. Speaking also slows you down just enough that vague phrases become audible. Do this for ten minutes on one topic and you will feel the difference before the next study session.