Uncle Sunny Academy
Prompt engineering and human+AI collaboration — a structured, verifiable methodology. Not trading, not investment content.
What will you learn here?
Uncle Sunny Academy consists of two modules. Module A (Fundamentals) shows you how to communicate effectively with AI systems — using a structured prompt template, the triple-check principle, and version discipline. Module B (SpaceX Mode, Advanced) demonstrates how a multi-AI workflow is built from first principles — with anonymised chain examples.
Module A
Fundamentals — Prompt Engineering Basics
Effective AI communication is not a talent — it is a learnable structure. This module covers the 5 core prompt elements, the triple-check principle, and version discipline.
Goal
What you want to achieve — stated precisely and measurably. Not 'write something', but 'write a 3-sentence summary for audience X in format Y'.
Bad: 'Explain prompt engineering.' Good: 'Write a 5-sentence, non-technical explanation of prompt engineering for a high school student, without bullet points.'
Context
What background information the AI receives — what it knows, what it does not. Context is always explicit, never assumed.
Bad: 'Continue the previous work.' Good: 'In the previous session we worked on topic X, the result was Y. Now we continue in direction Z.'
Constraints
What the AI must NOT do — exclusions, prohibited content, scope boundaries. Constraints are not distrust — they are precision.
Bad: (no constraints given). Good: 'Do not give financial advice. Do not use jargon. Do not exceed 200 words.'
Format
What format you want the answer in — list, table, prose, code, JSON. Format determines usability.
Bad: 'Give a summary.' Good: 'Give a 3-column table: Topic | Key point | Next step.'
Triple-Check Row
Ask the AI to verify its own answer from three angles: (1) does it meet the goal, (2) does it respect the constraints, (3) does it contain any assumption you have not confirmed.
Final prompt line: 'Before answering, verify: (1) do you meet the goal, (2) do you respect the constraints, (3) does your answer contain any assumption?'
Common Mistakes
| Mistake | Consequence | Fix |
|---|---|---|
| Vague goal | The AI generalises, the answer is unusable | Add measurability: who, what, in what format, what length |
| Implicit context | The AI assumes — and the assumption is usually wrong | Provide all relevant background information explicitly |
| No constraints | The AI exceeds its scope, unwanted content may appear | Always specify what the AI must NOT do |
| No triple-check | Errors pass unnoticed into the final output | For every critical output, request the AI's self-verification |
Practice Exercises
Write a 5-element prompt on a topic of your choice. Check: are all 5 elements present?
Take a previous prompt that did not get a good answer. Identify which element was missing.
Write a prompt whose last line is the triple-check request. Compare the answer with a version without triple-check.
Module B — Advanced
SpaceX Mode — First-Principles Prompt Engineering
Why does a multi-AI workflow work? Because we do not rely on a single AI's judgement. Just as SpaceX engineers do not rely on a single simulation — every critical decision is backed by an independent verification chain.
Core Principles
Every step has a measurable success criterion
No 'good faith' — only evidence
Every AI output goes through triple verification
Context is always explicit, never assumed
Errors must not be hidden — they must be documented immediately
Notify me at launch
Uncle Sunny Academy is opening soon. Subscribe and be the first to know.
Uncle Sunny Academy contains educational and informational content. The methodology presented is illustrative in nature. This is not investment, financial, legal, or medical advice. Data Analytic Investments assumes no responsibility for the results of applying the methodology presented here.