ipIterPrompt

Explain at Five Levels

Understand any concept via five explanations, from a child's level to expert.

iterpromptUpdated 2026-06-01

Explains any concept at five escalating levels of depth — child, teenager, undergrad, grad student, and domain expert. The progressive structure is one of the fastest ways to genuinely learn a new topic because each level builds vocabulary for the next.

The prompt

Explain {{concept}} at five levels of depth. Each level must build on the previous one, introducing new vocabulary only after the intuition exists.

**Level 1 — Child (age 8):** one concrete analogy from everyday life. No jargon at all.
**Level 2 — Teenager:** the real mechanism in simple terms, correcting any oversimplification from level 1.
**Level 3 — Undergraduate:** proper terminology, the key equation or formal definition if one exists, and the standard example used in courses.
**Level 4 — Graduate student:** the edge cases, trade-offs, and what practitioners actually argue about.
**Level 5 — Expert:** the current frontier — open problems, recent developments, and where the standard textbook treatment is outdated.

End with: "The one misconception to unlearn:" — the most common wrong belief about this concept.

Run this prompt on a real model without leaving the page. Every run is saved to your history for this prompt.

Fill in the variables

How to use

  1. 1Works for anything: technical concepts, financial instruments, historical events, biology.
  2. 2Read only up to the level you need — level 3 is the sweet spot for working knowledge.
  3. 3Ask follow-ups at a specific level: "expand level 4's second trade-off."
  4. 4For interview prep, ask it to quiz you at level 3–4 after you've read the explanation.

Examples

Learning how LLM context windows work

Input

Concept: context windows in large language models

Output

**Level 1:** Imagine reading a book through a small window that only shows one page... **Level 4:** ...attention is O(n²) in sequence length, which is why naive long-context is expensive; practitioners argue about RoPE scaling vs. sliding-window attention... **The one misconception to unlearn:** the model doesn't 'remember' your whole conversation — everything must fit in the window or it's gone.

Pro tips

  • Save your favorite outputs — the level-1 analogies are gold for teaching others.

Frequently asked questions

How accurate is level 5?+

Level 5 depends on the model's training cutoff, so treat 'current frontier' claims as a starting point and verify recent developments. Levels 1–4 are generally reliable for established topics.

Can I get just one level?+

Yes — 'Explain X at level 3 of 5, where 1 is a child and 5 is an expert' works standalone, but you lose the vocabulary build-up that makes the full version effective.

Related

Expert Guidance for Acoustic and Deep Learning Research

Turn any AI chat into an expert guidance for acoustic and deep learning research with this community persona prompt.

Open ↗

Security Monitoring with Wazuh: A Comprehensive Research Project

Turn any AI chat into a security monitoring with wazuh: a comprehensive research project with this community persona prompt.

Open ↗

Sports Research Assistant

Turn any AI chat into a sports research assistant with this community persona prompt.

Open ↗

Clinical Research Presentation Guidance

Turn any AI chat into a clinical research presentation guidance with this community persona prompt.

Open ↗

Used in these guides