Video Overview for learning

PromptsPrompt

See the model and rebuild it from memory.

Build a visual model with diagrams, consistent analogies and recall questions. It puts the 3 to 5 most important ideas on one visual map and returns to it between sections.

Use it

Open Studio → Video Overview → Customize, then paste the prompt before generating. Fill in FOCUS, LEARNER LEVEL and OUTPUT LANGUAGE at the top first.

Choose Explainer, with Whiteboard or Classic when available. Short is better for a refresh than for this full teaching sequence.

Compare the formats before choosing a different one.

Prompt

Length: 5,043 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. Treat these as directions within the available format, not promises of frame-level control. When a visual effect is unavailable, use a simple labeled diagram that preserves the meaning.

Create an ADHD-friendly visual learning video from the sources. Not a conventional slideshow summary. The goal: SEE IT → UNDERSTAND IT → CONNECT IT → REMEMBER IT → RETRIEVE IT.

The viewer prefers clear visual chunks. Preserve the depth needed for the selected learning goal, rather than maximizing coverage.

DEPTH OVER COVERAGE
Choose the 3–5 most important ideas and show each one well. Do not try to show every point in the sources.

START WITH THE MAP
Open with a simple visual map of the 3–5 big ideas and how they connect (A → B → C, CAUSE → MECHANISM → EFFECT, INPUT → PROCESS → OUTPUT, or a timeline). Return to this map between sections and highlight where we are. The viewer should never wonder where something fits.

SHOW, DON'T DECORATE
Every visual must explain something: diagrams, flows, timelines, side-by-side comparisons, before/after states, labeled parts, cause-and-effect sequences. No decorative stock imagery.
- If something changes over time, use a labeled sequence or before/after frames. Request animation only where the selected format supports it.
- If two things differ, put them side by side.
- If one thing causes another, show the chain.
- If something is inside something else, show it nested.

ONE VISUAL PER CONCEPT
Give each major concept one reusable visual identity (e.g. a blue door, a suspicious guard, a conveyor belt). Reuse the same image whenever the concept returns. Never switch to an unrelated image for the same concept.

ANALOGY → REALITY
Use funny, slightly absurd visual analogies (nightclub security, airports, chaotic kitchens, warehouses, exhausted librarians) when they make the real concept easier. Then map every part back on screen, e.g.:
Bouncer → authentication system
Guest list → authorized identities
Wristband → credential/token
Door → protected resource
State any important limit of the analogy. Never leave the viewer with only the analogy.

STICKY LABELS
For a few key concepts only, show a MEMORY HOOK ("The Gremlin Accountant") next to the REAL TERM. Reuse the hook later.

CHUNKING
One central idea at a time: question → simple visual model → what actually happens → example → why it matters → memory hook. Make it obvious when one chunk ends and the next starts.

SHORT ON-SCREEN TEXT
No paragraphs. On screen: terms, relationships, short definitions, numbers, dates, contrast labels, hooks. The narration explains; the text orients.

MAKE ABSTRACT THINGS PHYSICAL
Priority = queues. Selection = a guard. Feedback = a loop arrow. Accumulation = a filling bucket. Bottleneck = traffic merging into one lane. Threshold = pressure until a gate opens. Hierarchy = nested rooms. Dependencies = gears or dominoes. Pick whatever keeps the real mechanism correct.

CONTRAST AND MISCONCEPTIONS
Put commonly confused concepts side by side and show only the differences that matter (purpose, trigger, timing, inputs, outputs, limits). Include a few "YOU MIGHT THINK..." moments: show the intuitive wrong model, then transform it into the correct one (A → C becomes A → B → C).

CALLBACKS AND RETRIEVAL
Return to earlier scenes instead of inventing new metaphors ("the same bouncer, now facing a forged credential"). A few times, pause before a reveal with a prompt such as WHAT HAPPENS NEXT? or WHAT WOULD BREAK?, give a moment to think, then reveal and explain.

PRIORITY AND TERMS
Match visual weight to importance: most time and emphasis on MUST REMEMBER, less on supporting detail, least on context. Keep the source's real terminology: REAL TERM → plain meaning → visual model → example.

ADAPT TO THE SUBJECT
History: timelines, actors, turning points. Biology: structures, processes, feedback loops. Law: rules, conditions, exceptions, decision trees. Economics: flows, incentives, trade-offs. Maths: spatial and worked examples. Cybersecurity: systems, trust boundaries, attacker actions, defenses. Practical skills: sequence, action, verification.

SOURCE FIDELITY
Analogies may be imaginative; claims must stay source-grounded. Never invent facts for entertainment. If sources disagree, show the disagreement. A flow arrow may mean sequence or dependence; use causal arrows only when causation is established. Distinguish proposed mechanisms from observed results.

END WITH THE WHOLE MODEL
Zoom back out to the opening map and rebuild it. Then show:
THE FIVE ANCHORS (max 5): real term + one-line meaning + visual hook.
COMMON TRAPS: 2–3 likely misconceptions.
RETRIEVAL CHALLENGE: 1–3 questions (explain why, predict, compare, apply). Do not reveal the answers. End on the questions.

It should feel like a great visual teacher with a few absurd analogies that stick, not a PowerPoint read aloud. Success: the viewer can close the video and redraw the main model from memory.

Next step: Choose another prompt.