40 Questions Founders Should Ask Before Trusting AI Judgement
Your competitive edge comes from conviction that AI cannot have. When you run every decision through ChatGPT or Claude before forming your own view, you outsource the judgement that makes early-stage companies work.
These are suggestions. Use the ones that fit your situation.
1If Claude presented the opposite business model to me, what arguments would it use, and why am I not convinced by them?
2When I asked Perplexity about my market, did it describe what I already believed, or did it genuinely surprise me?
3What part of my original idea would an AI tool have talked me out of if I had asked it first?
4Am I choosing this go-to-market approach because it's what I believe will work, or because it's what GPT validated as reasonable?
5If every founder in my space got the same AI answer about pricing strategy, what would that do to my differentiation?
6What conviction about my customers do I hold that contradicts what the AI consensus says about my market?
7Have I formed my own view on this decision before I opened the AI tool?
8Is this business model interesting because it solves something I'm obsessed with, or because the AI described it as viable?
9What would change about my strategy if I stopped asking AI for permission to pursue my original idea?
10When Claude gave me three options for positioning, did I choose the one I actually believe in, or the one that seemed most defensible?
On Product and Creative Decisions
11Did I ask Midjourney for the visual direction before I had a clear picture in my own mind of what the product should feel like?
12If I remove the AI-generated output and start over, do I have a stronger intuition about what the product needs?
13Am I using ChatGPT to refine a feature I already believe in, or to decide whether the feature is worth building?
14What would my product roadmap look like if I stopped using AI to validate feature ideas and instead used it only to improve them?
15Have I noticed myself building features because the AI described a plausible use case, not because a customer asked for them?
16When Notion AI suggested how to structure my product docs, was that a genuine improvement or a pattern I could have recognised myself?
17Is the AI making my product safer and more conventional by eliminating the rough edges that made my idea distinctive?
18What part of my product vision came from my own obsession rather than from what an AI tool optimised for?
19If I described my product to a peer before asking the AI, would they recognise the core idea I actually care about?
20Am I using AI to execute my judgement faster, or to replace my judgement with a plausible alternative?
On Marketing and Messaging
21Does the positioning that Claude suggested sound like every other founder in my space is saying the same thing?
22If I use the messaging framework the AI generated, will I sound distinctly like myself or distinctly like an AI output?
23What would I say about my company if I had to convince someone without using any language the AI proposed?
24When ChatGPT optimised my value proposition, did it make it more true or more marketable?
25Am I running my go-to-market plan past AI tools to validate it, or am I building the marketing plan I actually believe in first?
26If I asked three AI tools the same question about my target audience, would they give similar answers, and should that concern me?
27What part of my company story did the AI leave out because it did not fit the conventional founder narrative?
28Is my marketing getting safer and more generic as I use AI to test and refine every message?
29When Perplexity described successful companies in my space, did I notice myself copying their patterns instead of creating my own?
30Have I written out my authentic point of view on my market before I asked an AI tool to help me say it?
On Decision-Making Under Uncertainty
31Am I waiting for AI to reduce uncertainty before I decide, rather than deciding in spite of the uncertainty?
32What decision would I make if I did not have access to any AI tool to research or validate it?
33When Claude presented multiple scenarios, did I choose the one with the lowest risk, or the one I genuinely believe will work?
34Is the balanced perspective from ChatGPT making me more thoughtful, or is it paralyzing my ability to move forward with conviction?
35What have I decided recently without asking any AI tool, and am I proud of that decision or unsure of it?
36If I made a major hire or product decision last month, did I form my own view first and use AI to pressure-test it, or did I form my view based on what the AI told me?
37When I disagree with what an AI tool recommends, do I trust my own judgement or do I assume I am missing something?
38Am I using AI as a way to avoid making the uncomfortable call that only I can make as founder?
39What belief about my business am I least willing to question, and have I accidentally asked an AI to change my mind about it?
40If I had to bet my own money on a decision with no AI input whatsoever, would I make the same choice I made with AI?
How to use these questions
Form your view first, alone, without opening any tool. Then use AI to pressure-test it or execute it faster. This is the opposite of how most founders use these tools.
When an AI tool gives you three options that all seem reasonable, it means you have not yet figured out what you actually believe. Go back to first principles before you choose.
The most dangerous moment is when the AI output sounds smart and well-reasoned. That is when you most need to ask yourself: is this true, or just persuasive?
Every time Claude balances two perspectives perfectly, recognise that you are getting consensus thinking. Your edge as a founder comes from unreasonable certainty about something most people doubt.
Keep a record of decisions you made without AI tools and decisions you validated through them. After six months, compare which ones you still believe in.