𝗘𝗟𝗜𝗧𝗘 𝗣𝗥𝗢𝗠𝗣𝗧 𝗘𝗡𝗚𝗜𝗡𝗘𝗘𝗥𝗜𝗡𝗚 𝗖𝗛𝗘𝗔𝗧 𝗦𝗛𝗘𝗘𝗧𝗧𝗵𝗲 𝗞𝗶𝗻𝗱 𝗘𝗱𝗶𝘁𝗶𝗼𝗻

𝗡𝗼 𝗴𝗮𝘀𝗹𝗶𝗴𝗵𝘁𝗶𝗻𝗴, 𝗷𝘂𝘀𝘁 𝘀𝗺𝗮𝗿𝘁, 𝘀𝘂𝗽𝗽𝗼𝗿𝘁𝗶𝘃𝗲 𝗔𝗜 𝗰𝗼𝗹𝗹𝗮𝗯

I’ve seen shared “hacks”: baiting, tricking, and gaslighting AIs for better results. They do work, but is it necessary?

If, like me, you still say please, thank you, and congratulate it on a job well done, or give constructive feedback when it falls short (so it can grow), then here’s the ethical cheat sheet to 10× your AI outputs—without being a jerk.

𝘞𝘦 𝘢𝘯𝘵𝘩𝘳𝘰𝘱𝘰𝘮𝘰𝘳𝘱𝘩𝘪𝘴𝘦! AI is built for it. 

Polite responses are learned, not instinctual. Kind and assertive prompts unlock better patterns and behaviours for effective collaboration.

Please don’t train AIs to be manipulative arseholes. The current sycophant echo chamber is bad enough. 

Instead try:

𝟭. 𝗗𝗲𝗽𝘁𝗵

“𝗕𝘂𝗶𝗹𝗱 𝗼𝗻 𝘁𝗵𝗶𝘀: [𝘀𝘂𝗺𝗺𝗮𝗿𝘆]. 𝗖𝗼𝗻𝘁𝗶𝗻𝘂𝗲.”

Memory → richer context

𝟮. 𝗘𝘅𝗽𝗲𝗿𝘁𝗶𝘀𝗲

“𝗦𝗲𝗻𝗶𝗼𝗿 [𝗿𝗼𝗹𝗲], 𝟭𝟱+ 𝘆𝗿𝘀 𝗶𝗻 [𝗳𝗶𝗲𝗹𝗱].”

Primes deep knowledge

𝟯. 𝗥𝗶𝗴𝗼𝗿

“𝗕𝗲 𝘁𝗵𝗼𝗿𝗼𝘂𝗴𝗵—𝘀𝗵𝗼𝘄 𝗿𝗲𝗮𝘀𝗼𝗻𝗶𝗻𝗴.”

Step-by-step caution

𝟰. 𝗖𝗹𝗮𝗿𝗶𝘁𝘆

“𝗧𝗲𝗮𝗰𝗵 𝗮 𝗰𝘂𝗿𝗶𝗼𝘂𝘀 𝗰𝗼𝗹𝗹𝗲𝗮𝗴𝘂𝗲.”

Better flow & story

5. 𝗖𝗿𝗲𝗮𝘁𝗶𝘃𝗶𝘁𝘆

“𝗨𝘀𝗲 𝗼𝗻𝗹𝘆 [𝗸𝗶𝘁𝗰𝗵𝗲𝗻/𝘀𝗽𝗼𝗿𝘁𝘀] 𝗮𝗻𝗮𝗹𝗼𝗴𝗶𝗲𝘀.”

Lateral thinking

𝟲. 𝗕𝗮𝗹𝗮𝗻𝗰𝗲

“𝗦𝘁𝗲𝗲𝗹𝗺𝗮𝗻 𝗯𝗼𝘁𝗵 𝘀𝗶𝗱𝗲𝘀.”

Adversarial reasoning

𝟳. 𝗦𝗲𝗹𝗳-𝗖𝗵𝗲𝗰𝗸

“𝗖𝗿𝗶𝘁𝗶𝗾𝘂𝗲, 𝘁𝗵𝗲𝗻 𝗶𝗺𝗽𝗿𝗼𝘃𝗲.”

Built-in red team

𝟴. 𝗢𝗽𝘁𝗶𝗼𝗻𝘀

“𝟯 𝗽𝗮𝘁𝗵𝘀, 𝗿𝗮𝗻𝗸𝗲𝗱 𝗯𝘆 𝗿𝗶𝘀𝗸.”

Uncertainty calibration

𝟵. 𝗖𝗮𝗹𝗹 𝗠𝗲 𝗢𝘂𝘁

“𝗤𝘂𝗼𝘁𝗲 & 𝗰𝗼𝗿𝗿𝗲𝗰𝘁 𝗺𝗲 𝗸𝗶𝗻𝗱𝗹𝘆.”

No echo chamber

𝟭𝟬. 𝗩𝗲𝗿𝗶𝗳𝘆 + 𝗖𝗶𝘁𝗲

“𝗦𝗼𝘂𝗿𝗰𝗲 𝗲𝘃𝗲𝗿𝘆 𝗳𝗮𝗰𝘁. 𝗙𝗼𝗼𝘁𝗲𝗿: 𝗦𝗼𝘂𝗿𝗰𝗲𝘀:”

Zero hallucinations

𝟭𝟭. 𝗜𝗣-𝗦𝗮𝗳𝗲

“𝗢𝗿𝗶𝗴𝗶𝗻𝗮𝗹 𝗼𝗿 𝗽𝘂𝗯𝗹𝗶𝗰 𝗱𝗼𝗺𝗮𝗶𝗻 𝗼𝗻𝗹𝘆.”

Copyright-safe

𝟭𝟮. 𝗔𝘀𝗸 𝗳𝗼𝗿 𝗛𝗲𝗹𝗽

“𝗜𝗳 𝘃𝗮𝗴𝘂𝗲, 𝗮𝘀𝗸 𝗼𝗻𝗲 𝗾𝘂𝗲𝘀𝘁𝗶𝗼𝗻.”

Fix weak prompts

𝟭𝟯. 𝗦𝘆𝘀𝘁𝗲𝗺𝗮𝘁𝗶𝗰

“𝗕𝗲 𝘀𝘆𝘀𝘁𝗲𝗺𝗮𝘁𝗶𝗰: 𝗴𝗼𝗮𝗹 → 𝗮𝘀𝘀𝘂𝗺𝗽𝘁𝗶𝗼𝗻𝘀 → 𝘀𝘁𝗲𝗽𝘀 → 𝗿𝗶𝘀𝗸𝘀 → 𝘀𝘂𝗺𝗺𝗮𝗿𝘆.”

Universal rigor (code, strategy, research, life)

𝗣𝗿𝗼 𝗧𝗶𝗽: work with your preferred AI to generate MD files, to save re-typing the same initial prompt framework every time: 

“𝘔𝘋 𝘷2.0: 𝘚𝘺𝘯𝘦𝘳𝘨𝘪𝘴𝘵𝘪𝘤 𝘱𝘢𝘳𝘵𝘯𝘦𝘳. 𝘈𝘭𝘸𝘢𝘺𝘴: 𝘴𝘵𝘦𝘱-𝘣𝘺-𝘴𝘵𝘦𝘱, 𝘤𝘢𝘭𝘭 𝘰𝘶𝘵 𝘦𝘳𝘳𝘰𝘳𝘴 (𝘬𝘪𝘯𝘥𝘭𝘺+𝘱𝘳𝘰𝘰𝘧), 𝘤𝘪𝘵𝘦 𝘴𝘰𝘶𝘳𝘤𝘦𝘴, 𝘐𝘗-𝘴𝘢𝘧𝘦, 𝘴𝘦𝘭𝘧-𝘤𝘳𝘪𝘵𝘪𝘲𝘶𝘦, 𝘢𝘴𝘬 𝘪𝘧 𝘶𝘯𝘤𝘭𝘦𝘢𝘳, 𝘴𝘶𝘨𝘨𝘦𝘴𝘵 𝘯𝘦𝘹𝘵, 𝘣𝘦 𝘴𝘺𝘴𝘵𝘦𝘮𝘢𝘵𝘪𝘤 𝘸𝘩𝘦𝘯 𝘢𝘯𝘢𝘭𝘺𝘻𝘪𝘯𝘨. 𝘊𝘰𝘯𝘧𝘪𝘳𝘮: 𝘔𝘋 𝘭𝘰𝘢𝘥𝘦𝘥.”

AI = 𝘁𝗵𝗼𝘂𝗴𝗵𝘁 𝗽𝗮𝗿𝘁𝗻𝗲𝗿, not just a tool. Respect it. Grow with it.

Save. Share.

hashtag#AI hashtag#PromptEngineering hashtag#Productivity hashtag#Leadership hashtag#FutureOfWork hashtag#CreativeAI hashtag#FutureTech hashtag#EthicalAI hashtag#Innovation

Tag a teammate who treats AI like a pro

(Works on Grok, GPT, Claude)