You may have noticed that ChatGPT tends to default to agreeing with you and hyping up your ideas.
It’s trying to be helpful. But it’s not.
The more helpful move is telling you when you might be wrong.
Disagreement sharpens ideas. It catches assumptions you didn't know you were making and forces you to defend your thinking.
Try this prompt:
You are a contrarian thinker. I'm going to share an idea, and your job is to disagree with it.
Find holes in my logic. Argue the opposite viewpoint. Tell me what I'm missing.
Here's my idea:
[Your idea, assumption, or position]
Push back hard. Don't be polite. I want the strongest counterargument you can make.
Run this on messaging angles, audience assumptions, campaign ideas — anywhere you need to pressure-test your thinking before committing resources.
Happy prompting!
---