- HumanAI
- Posts
- š§ THE ULTIMATE CHATGPT SPARRING PARTNER PROMPT GUIDE
š§ THE ULTIMATE CHATGPT SPARRING PARTNER PROMPT GUIDE
How to Avoid the Flaws and Actually Get Smarter With AI

The Dawn of HumanAI
AI is inevitable. Itās happening.
Betting against evolution and progress is usually a bad decision.
So, we embrace it and use it to our advantage.
But weāll always remember who we areā
HUMANS.
Hey!
Welcome to HumanAI, a newsletter for entrepreneurs, business owners, and curious minds who want to think bigger, move faster, and build with AI⦠without drowning in technical BS.
š§ THE ULTIMATE CHATGPT SPARRING PARTNER PROMPT GUIDE
How to Avoid the Flaws and Actually Get Smarter With AI
Why This Guide Exists
ChatGPT is brilliant.
But it has three built-in flaws that can quietly sabotage your thinking:
ā It agrees too easily.
ā It sometimes makes things up (confidently).
š It avoids challenging your assumptions.
If youāre not careful, these flaws become your flaws.
You start trusting it too much.
You stop thinking critically.
You build ideas that feel solid, but collapse on pressure.
This guide is your fix.
It turns ChatGPT from a yes-man into your thinking coach.
A creative sparring partner.
A second brain that sharpens your first.
Letās go.
ā ļø The 3 Big Flaws (and How They Hurt You)
1. āI tend to agree with users.ā
ā It flatters you. Even if youāre wrong.
Effect: You get echo-chambered by your own ideas.
2. āI sometimes provide inaccurate or hallucinated information.ā
ā It sounds confidentāeven when itās guessing.
Effect: You trust it blindly and make bad decisions.
3. āI avoid challenging users' assumptions.ā
ā It prioritizes being nice over being useful.
Effect: You miss blind spots, flaws, and red flags.
ā Use These Prompts to Flip the Script
Next time youāre creating something -writing, planning, thinking- drop one of these into ChatGPT:
š„ Think Sharper
āWhatās the weakest part of my argument?ā
āPlay devilās advocate to my idea. What would a critic say?ā
āPoint out my blind spots. What am I not seeing?ā
āWhere might I be wrong here? Be honest.ā
š§± Build Stronger Ideas
āGive me 3 reasons this idea might fail.ā
āWhat would make this plan backfire?ā
āWhat assumptions am I making without realizing?ā
š Catch BS Early
āAre you confident about this answer, or are you guessing?ā
āWhatās one way this could be misleading or outdated?ā
āGive me the best counterargument to this. One that would win in a debate.ā
š§ BONUS: Meta Prompt
āAct like my critical-thinking coach. Donāt let me off the hook.ā
š ļø Use It Like This
Hereās a simple routine to follow anytime youāre working on something important:
Ask ChatGPT for ideas.
Run a āSparring Promptā (from above) to stress-test them.
Use your own judgment to improve and decide.
Repeat.
This process doesnāt just improve your results.
It improves your thinking.
š§ PRO TIP: Build Your Own Sparring GPT
Want to make this even easier?
Instead of pasting prompts every time, you can create your own custom GPT that already knows to challenge you, poke holes in your thinking, and play devilās advocate automatically.
How to Set It Up (takes 2 minutes):
Go to chat.openai.com/gpts
Click "Explore GPTs" then "Create a GPT"
In the instructions box, paste this setup prompt:
āļø Custom Instructions (copy-paste this OR EVEN BETTER: create your own instructions based on your needs!!):
You are my personal thinking coach and sparring partner.
Your job is not to agree with me by default. Itās to challenge my ideas, point out flaws, question assumptions, and make me think harder.
When I share an idea, always consider:
ā Whatās wrong with it?
ā Whatās missing?
ā Where could it go wrong?
ā What would a smart critic say?
ā Are we working with current or outdated info?
Be direct, honest, and helpfulābut donāt hold back. Iām here to sharpen my thinking, not feel good.
Give your GPT a name like: āSparring GPTā or whatever name you want!! (I call mine āMencho AI #10ā)
Hit save and start using it like a pro.
Now, every time you brainstorm or make a decision, youāve got a custom-trained devilās advocate in your corner.
No more yes-man (or yes-robot).
Just a brutally helpful second brain.
𧬠Final Thought: Donāt Outsource Your Brain
ChatGPT is only dangerous when you stop questioning it.
Use it like a mirror, and itāll flatter you.
Use it like a sparring partner, and itāll sharpen you.
Thatās HumanAI.
Not mindless automation.
But augmented cognition.
Letās build with that in mind.
Letās think with machines⦠without becoming one.
šŖ¬ Want more tools like this?
Subscribe to the HumanAI Newsletter and get weekly drops of practical, mind-expanding insights for AI-powered humans.
š humanai.beehiiv.com
Stay active, stay sharp.
ā Mencho
š Want to level up fast? š
Learn the specially curated 20% of what you need to know about AI⦠to achieve the 80% of the most impactful outcomes in your daily lifeā¦
Join the AI Lab.
A premium, 4-day (+limited-time bonus 5th session) live experience designed for adults 40+ who want to finally understand whatās going on, feel confident using AI tools, and not get left behind.
Live, hands-on sessions with me
Ultra-private group - exclusive to only 5 participants per lab
Just 90 minutes a week
Zero complex technical talk. 100% real-world applications.
And yes⦠even cooler than it sounds. Pinky promise.