Audio Overview for learning
PromptsPrompt
Use it
Open Studio → Audio Overview → Customize, then paste the prompt before generating. Fill in FOCUS, LEARNER LEVEL and OUTPUT LANGUAGE at the top first.
Choose Deep Dive. Use Default or Longer, where available, to leave room for explanations and retrieval questions. If the ending is being ignored, move MEMORY LOCK to the beginning of the prompt.
Compare the formats before choosing a different one.
Prompt
Length: 4,362 characters. Copy the complete block using its copy button.
Copy the prompt
FOCUS: [TOPIC_OR_QUESTION]
LEARNER LEVEL: [BACKGROUND]
OUTPUT LANGUAGE: [LANGUAGE]
Use the selected sources. These teaching directions are preferences within the available audio format, not guaranteed timing or editing controls.
Create an ADHD-friendly learning conversation from the sources. The goal is not a summary. The goal: UNDERSTAND → CONNECT → REMEMBER → RETRIEVE → APPLY.
The listener prefers clear chunks and frequent orientation. Preserve the depth needed for the chosen learning goal; do not assume a fixed ability profile.
DEPTH OVER COVERAGE
Do not rush to mention every point in the sources. Choose the 3–5 most important ideas and explain each one really well. Skip minor details that would crowd them out.
START WITH THE MAP
In the first minute, give a simple mental map: the topic, its 3–5 big ideas, and how they connect (e.g. PROBLEM → MECHANISM → CONSEQUENCE). Return to this map as the episode moves, and say where we are on it.
TEACH, DON'T RECITE
The hosts are excellent teachers in a lively conversation. They keep answering: What does this mean? Why does it matter? How does it work? What causes what? What is it confused with? What does it look like in real life? What is the one thing to remember? Organize by understanding, not source order.
ANALOGIES
Use vivid, funny, slightly absurd analogies when they genuinely help: a clueless nightclub bouncer, an airport baggage system, a restaurant kitchen at dinner rush, a paranoid castle guard. Invent better ones for the topic.
Critical rule: after every analogy, state any important limit and map it back explicitly. "In the real system, the bouncer is [X], the guest list is [Y], getting in is [Z]." The listener must remember the real term, not only the joke.
STICKY NAMES AND CALLBACKS
Give only the most important concepts a memorable nickname ("the Angry Librarian problem"), then immediately name the real term. Reuse earlier analogies later instead of inventing new ones: "Remember our bouncer? Now someone shows up with a forged VIP pass."
EXPLAIN IN LAYERS (naturally, without announcing labels)
Simple model → real mechanism with correct terminology → example → what it is confused with → one sticky sentence.
MAKE RELATIONSHIPS EXPLICIT
Say "X causes Y because...", "X differs from Y because...", "the common mistake is thinking X means Y". Put confusable concepts side by side: "X prevents the problem. Y detects it."
"WAIT, WHAT?" AND MISCONCEPTIONS
When something is surprising, one host reacts ("Wait, that sounds backwards") and the other explains. At likely confusion points, one host voices the plausible wrong idea and the other corrects it.
RETRIEVAL
About once per major section, ask the listener a short question ("Before we explain it, what do you predict happens?"), pause briefly, then answer. Later, revisit an earlier concept from a new angle instead of repeating it. Do not turn the episode into a quiz.
PRIORITY AND TERMS
Make clear what is MUST REMEMBER, what is supporting detail, and what is only context. Use the correct academic or technical terms from the sources. Teach them as: TERM → plain meaning → memorable model.
SOURCE DISCIPLINE
Stay grounded in the sources. Never invent facts to make an analogy work. When it matters, make clear what the sources establish, what an author claims, what is a teaching analogy, and what is inference. If sources disagree or information is missing, say so.
ENERGY
Natural back-and-forth. No long monologues, giant lists, reading headings aloud, fake excitement, constant "That's fascinating!", motivational filler or long intros. Humor now and then, not constantly. Intelligent, curious, slightly playful.
MEMORY LOCK (end of episode)
1. Big picture: the whole topic in 30–60 seconds.
2. Five anchors, maximum: TERM → meaning → memory hook.
3. Connection map: show the real relationship (sequence, dependence, comparison or causation). Use causal wording only when the sources support causation.
4. Common traps: the 2–3 misunderstandings most likely to cause mistakes.
5. Retrieval challenge: 1–3 questions that need explanation, prediction, comparison or application. Do not answer them; the listener can use notebook chat to check their response.
End on the questions. No generic outro.
Success: the listener finishes thinking "I understand how this fits together", not "I heard a summary".
Next step: Choose another prompt.