Every idea that’s born from our brains feels like a precious baby. We don’t want to give them up. So, most ideas feel stronger than they actually are especially when we are flush with it in the beginning. When something clicks, we are just too lazy to fully question it. ChatGPT doesn’t always help here either, since AI models tend to agree or stay neutral instead of challenging you directly. This AI echo-chamber is useful for searching or summarizing, but risky for decision-making.
So I started getting into the habit of using a Socratic-style master prompt to tease and pull apart any major idea. I want ChatGPT to question, challenge, and probe instead of keeping me in my bubble. And almost every time, it exposes blind spots I wouldn’t have noticed on my own.
- OS
-
Android, iOS, Web
- Developer
-
OpenAI
- Price model
-
Free with optional subscription
This prompt turns ChatGPT into a critic
It forces honest, step by step feedback
Let’s start with the Socratic prompt I use to stress-test any idea. In the latter sections we will break it down and see why this method of self-examination (thanks Socrates!) is so useful. Here again let me reiterate that giving ChatGPT context is the fundamental rule that works for every scenario. The quality of any prompt rides on the coattails of this rule.
Act as a skeptical expert and critical thinker. Analyze the following idea.
Identify the weakest assumptions. List the biggest risks and failure points. Explain what would have to be true for this to succeed.
Argue against the idea as a strong critic would. Ask me 3–5 questions that would expose flaws or force me to rethink it. Be direct and honest. Avoid being agreeable.
I’ll admit, the first time I used this, it felt like a productivity nightmare. After all, we use ChatGPT for quick answers. I wasn’t used to seeing my ideas challenged so directly, especially when I was excited about them.
But that’s exactly why it works. It pushes ChatGPT into a Socratic role where it needs to collaboratively question instead of agreeing automatically. This shift alone jolts us into critical thinking.
Use Socratic prompts whenever the cost of a confident wrong answer is high. For instance, career related research. It can be any situation where “sounds right” is not good enough. The time investment for Socratic prompting is less useful for simple, factual lookups where a direct command is faster and sufficient.
Most ideas fail due to our hidden assumptions
Ask what must be true for this to work
A big part of the prompt focuses on assumptions. You can isolate that into a Socratic question like that surfaces those hidden assumptions.
What assumptions am I making that must be true for this idea to succeed?
This question forces you to surface things you usually gloss over. For a writer like me, I can find out whether my audience actually cares or whether the idea is generic.
I used to skip this step because it felt like overthinking. If an idea felt right in my gut, I assumed it was worth pursuing without dissecting it too much.
But assumptions are where most ideas break. Questioning them early doesn’t actually slow you down but prevents you from wasting your effort or at least finetuning your idea. Socratic interrogation is similar to asking ChatGPT to play the Devil’s advocate.
It surfaces risks you tend to ignore
Ask what could realistically go wrong
Another part of the prompt focuses on risks. You can turn that into a simple and separate Socratic query:
If this fails, what are the most likely reasons why?
This shifts your thinking from optimism to realism. Instead of vague concerns, you start seeing specific issues like poor execution, fuzzy messaging, or lack of resources.
When I’m excited about an idea, I don’t want to focus on what could go wrong. But this isn’t about negativity, it’s about preempting the future and working backwards from there. Seeing risks early gives you a chance to fix them before they become real problems.
The questions matter more than the answers
Ask what a critic would challenge here
One of the most powerful parts of the prompt is the questions it generates. You can mimic that with:
What would a smart critic question or challenge about this idea?
This pushes you to step outside your own perspective and look at the idea from a more objective lens. You can also follow a basic rule of prompting and ask ChatGPT to be a specific kind of expert.
Questions felt less useful because they didn’t give immediate direction. But the right question can change everything. Want to widen your lens? Give ChatGPT more than one role. For instance, I can filter my article ideas with a prompt like: “How would an editor, a beginner, and technophobe each view this choice?”
This ChatGPT setting I skipped over ended up making my answers way better
A hidden ChatGPT setting made my responses sharper and more personal.
I use this before every big decision
Ask what would change your mind
Before committing to any idea, I often add one final Socratic question:
What evidence or outcome would make me reconsider this idea?
This helps you avoid getting emotionally attached. It creates a built-in check that keeps your thinking flexible and grounded. But that’s exactly when it’s most useful.
When you feel certain, you’re least likely to question yourself. Also, if the idea collapses under a single challenge from ChatGPT’s prompt, it wasn’t as solid as it felt. But we also shouldn’t stop there. I like to support the results from ChatGPT with more deep research on the web.
Allow ChatGPT to challenge your thinking
Socratic prompts widen our perspective. The biggest benefit for me has been the improvement in my critical thinking. Instead of quick answers, playing around with Socratic prompts keeps me thinking about my ideas even away from the screen. One of my favorite Socratic prompts (hat tip to an anonymous Reddit user) is: “What is the worst realistic outcome, and can I live with it?”
