How do you prevent AI from validating your own bad ideas?
by•
The growth example hits home because AI is terrifyingly good at agreeing with whatever hypothesis you feed it.
If your starting prompt is biased, the tool will just find data to make you feel smart about a wrong decision.
True collaboration only happens when you force the model to look for counter-arguments or blind spots you missed. Have you found a reliable way to get AI to challenge your strategy instead of just echoing your thoughts back to you?
30 views
Replies
What works for me is not asking for an opinion at all. I ask the model to argue against the idea first, as if it were a competitor trying to prove it fails, and only then to list what would have to be true for it to work. And wherever I can, I hand it numbers instead of my hypothesis. My paywall had the yearly plan selected by default, and of 16 people who tapped Continue, 15 cancelled on Apple's sheet. Given those numbers the model stops agreeing with my favourite theory very quickly. The bias usually lives in the question, so I try to ask what the data says, not whether I am right.
I just tell claude to act as a partner from YC. :)
in less than 200 tokens my ideas are shot down
Give it the decision, not the idea. Ask what would have to be true for the plan to work. Then ask which of those points you actually have evidence for. The gaps are where it stops agreeing with you. A fresh session with none of your earlier back and forth helps too.