In partnership with
👋 Welcome to AI Central,
Yesterday we shared a guide that could make using AI easier in your day-to-day life. Today, we’re looking at another problem you might have noticed: AI often agrees with you, even when your idea has flaws or needs a second look.
The thing is, asking AI to be honest doesn’t always mean it will challenge your thinking. Sometimes, you need to ask the right questions to get useful feedback instead of a simple “that sounds great.”
So today, we’re sharing 4 simple ways to get AI to challenge your ideas (full guide here), along with 4 bonus prompts you can copy and paste to get more honest feedback on your plans and decisions.
Big hug and love you ❤️
Kris & The AI Central Team
P.S. Next time AI tells you your idea is great, try asking it what could go wrong. You might get a much more useful answer.
🔗 TOGETHER WITH BEEHIIV
Yes, that Arnold Schwarzenegger.
He has a newsletter on beehiiv. So do Codie Sanchez, Jay Shetty, David Begnaud, Colin & Samir, and Joanna Stern.
They could publish just about anywhere. They chose beehiiv to build a direct connection with the people who want to hear from them most. And you can do the same, whether you’re world-famous or just at the start of your plan for world domination.
You're seeing this ad because you're on the free plan - upgrade to remove ads & unlock 1200 AI Tutorials. Ads keep AI Central free for everyone
📚 Want even more resources?
Most AI content online is noise - too basic or too niche to help with the job you're actually paid to do.
If you’re looking for bonus content to the bootcamp, our AI library is the rest of syllabus: 3,000 hours of AI research, filtered into one place - searchable by topic, use case and author so you find exactly what you need in minutes.
There's a preview video on our homepage - watch it, then start a 1-month trial for $4.99 (less than a coffee).
👇 Get ahead before next week's guides
🔗 BROUGHT TO YOU BY SUPERHUMAN AI
200+ Proven Ways to Make Money With AI in 2026
The next wave of millionaires will be people who figured out how to make AI work for them.
The window to get ahead is still open. But not for long.
Here are 200+ proven ways to make money with AI in 2026.
Sign up for Superhuman AI, the free daily newsletter read by 1M+ professionals, and get instant access to all 200+ ways to profit from AI this year.
You're seeing this ad because you're on the free plan - upgrade to remove ads & unlock 1200 AI Tutorials. Ads keep AI Central free for everyone
📓 Get AI to argue with you
Inside this guide you will learn:
Section | What you can do by the end |
|---|---|
21.1 Agreement is trained in, not a personality it happens to have | Name the 4 shapes it takes, and stop reading them as politeness |
21.2 Telling it to be brutal gets you brutality, not the problem | Spot performed criticism on sight, and know why a harsher ask won't help |
21.3 Take yourself out of the question | Use 4 framings that give it no clue which answer you're hoping for |
21.4 A critique it survived is not a decision it approved | Say exactly what a surviving plan has proved, and what it hasn't |
You'll need: one real piece of your own thinking that hasn't been properly attacked yet. A plan, a recommendation, a hire, a draft you already like.
📝 Use these prompts to get honest feedback from AI
1️⃣ Get an honest opinion on your idea
Here are two options I'm considering: [Option A] and [Option B]. Make the strongest case for each, identify the biggest weakness in both, and tell me which one you'd choose based on the information provided. Explain what evidence would change your decision.
2️⃣ Find the biggest flaw in your plan
Review this plan: [Paste your plan]. Identify the single assumption it depends on most. Explain what could make that assumption false, what the consequences would be, and what I should verify before moving forward. Be specific to my plan, not generic.
3️⃣ Get feedback without bias
Review the following work as if a colleague had sent it to you for an honest assessment: [Paste your work]. Don't assume the author is right. Identify the weakest argument, any unsupported claims, and the one change that would improve it most. Quote the exact lines you're referring to.
4️⃣ Argue against your recommendation
My recommendation is: [Insert recommendation]. Build the strongest evidence-based argument against it, as if you had to convince the person approving it to reject my proposal. Use the information provided, identify specific weaknesses, and explain what evidence would prove your objections wrong.
One tip: run these prompts in a new chat, especially when you're reviewing work that AI has already praised. A fresh conversation gives it less previous context to lean on.
And remember, the goal isn't to make AI disagree with everything. It's to uncover a genuine weakness you can investigate before making a decision.
👉 Want the 4 framings that stop AI from guessing the answer you're hoping for? They're in section 21.3 of today's full guide.
