Module 1 of 5
The Logic
Lens
How AI actually works — from the inside out
FIRST PRINCIPLES · LLMs · TRAINING
A hands-on guide to understanding, using, and shaping artificial intelligence.
Each module ends with a Structured Expert Prompt (SEP) you'll try yourself.
A prompt is how you talk to AI. The quality of what you get back depends entirely on how well you ask.
"Help me with my essay"
Vague. No context. No role. No method.
AI gives you a generic, average answer.
"You are a [Expert].
Use [Method] to help me
with [Task]."
Specific. Gives AI a role and methodology.
Outputs are dramatically better.
How AI actually works — from the inside out
AI learns by finding patterns in massive amounts of data — like reading millions of books to understand language. Pattern recognition at unprecedented scale.
Tools like Claude and ChatGPT predict the most likely next word — billions of times, very fast. That's the "intelligence." It's not magic, it's math.
Training = learning from data (once, takes months). Inference = answering your question (real-time, takes seconds).
💡 First Principles Thinking: Break any big idea into its smallest true parts. That's what great scientists AND great prompts do.
By giving AI a specific persona (MIT scientist) and method (First Principles), you force it to THINK before answering — not just pattern-match a generic response. The quality difference is dramatic.
AI in real companies, real careers, real money
AI reads medical scans faster than doctors. Skills needed: data literacy, medical ethics, empathy.
Netflix uses AI to decide what shows get made. Skills: storytelling, analytics, cultural awareness.
AI reviews contracts in seconds vs. weeks. Skills: critical thinking, legal reasoning, persuasive writing.
AI predicts what you'll buy before you know you want it. Skills: consumer psychology, data analysis.
Jobs-to-be-Done stops AI from giving you a Wikipedia summary. It forces the AI to think like a consultant — identifying real pain points and what skills actually matter for YOUR future.
Using AI to actually learn — not just to cheat
→ Ask it to quiz you on your notes
→ Get it to explain concepts in simpler terms
→ Use it to brainstorm essay angles (not write it)
→ Have it critique your draft and suggest improvements
→ Ask for study schedules and accountability plans
→ Copy-paste AI answers as your own work
→ Skip thinking and just prompt for essays
→ Rely on it without checking facts
→ Use it as a crutch — your brain needs reps
The constraint "Do NOT give me the answer" is the most powerful phrase in prompting. It transforms AI from a shortcut machine into a thinking partner — which is how real learning happens.
Create things the world has never seen before
AI doesn't replace creativity — it amplifies it. The more specific and bold your vision, the better the output.
Generate mood boards, logo concepts, color palettes, visual stories
Brainstorm story angles, develop characters, write song lyric drafts
Script TikToks, YouTube videos, ad campaigns with hard creative constraints
Apply Friction Forensics — find where ideas get boring, then disrupt them
Generate chord progressions, lyrics, song titles, full album concepts
Brainstorm features, user flows, and zero-budget MVP concepts
Hard Edges = Better Ideas: "Zero-budget", "must fit in 15 seconds", "only 3 colors" — constraints spark creativity, they don't kill it.
Friction Forensics diagnoses WHY ideas fail before generating new ones. The Hard Edge constraints force AI away from the average — every idea must be distinctly different from the rest.
Who's responsible when AI gets it wrong?
AI learns from human data — so it inherits human biases. Facial recognition software has been shown to misidentify women of color at significantly higher rates. Real consequences in hiring, policing, and healthcare.
Every prompt you type is data. Who owns it? Who trains on it? Understanding data privacy is a critical 21st-century skill — one that will set you apart in any field you enter.
AI can generate convincing fake videos, audio, and images. How do we maintain truth in a world where seeing isn't believing? These are questions YOUR generation will need to answer.
You don't just use technology — you shape it. Your voice, your perspective matters.
Multi-agent prompting forces AI to argue with itself — you get multiple legitimate perspectives instead of one "safe" answer. The Board Resolution ensures you leave with actionable conclusions, not just debate.
Five expert prompts. Five superpowers. All yours.
The Formula: Give AI a role → Give it a method → Give it a constraint → Watch the quality skyrocket.
The tools are ready.
What will you build?
Consistency. Patience. Vision.
Try one SEP prompt on a school assignment
Research AI in one career you care about
Ask AI to debate itself on a topic you care about