AI Hacks
Choose your weapon.
19 Laws ofChatGPT Prompts
Principles to write better prompts.
32 CheatCodes ofChatGPTShort modifiers to control outputs instantly.
Two libraries. Copy-paste ready.
Reality check
Make ChatGPT your harshest collaborator.
Stop asking for applause; ask for friction. Aim for prompts that break your idea before the market does.
- GPT loves to agree; weak ideas get praised instead of fixed.
- Optimism bias hides budget, timing, or audience gaps.
- Surface-level answers skip the real blockers and trade-offs.
- Silence on risk means you ship blind and learn late.
- Without a skeptic, you overfit to what you already believe.
Honest prompts de-risk launches faster than polite ones.
A good prompt should…
- Interrogate assumptions until they break.
- Stress-test pricing, audience, and distribution early.
- Surface edge cases and failure modes before you commit.
- Push back with alternatives so you choose on evidence.
- Return a blunt next step instead of a pat on the back.
Treat ChatGPT like a pre-mortem partner: tell it to find the cracks, force trade-offs, and hand you a sharper next step.
Prompts are checkpoints,
not compliments.
Two tracks that match how ready you are. Use one when you need instant heat, the other when you need to gather context before you swing.
One Shot Prompts
Drop-in, zero setup. Paste the brief and get a blunt read. Works best with memory on so it can call out patterns you've shown before.
Guided Prompts
Short sequences that collect context, sharpen constraints, and hand you a tailored prompt plus a first move—perfect when the problem is still fuzzy.