Challenge What You Think
Most of us use AI to get things done — write an email, plan a trip, look something up. That's great. But there's another way to use it that's harder to describe and harder to put down once you try it.
You can use AI as a sparring partner — something to push against, argue with, and use to sharpen your own thinking about the world, about other people, and about what you yourself actually believe.
This section is a collection of prompts for that kind of exploration. None of them have right answers. They're not tests. They're invitations to see something familiar in a completely new light.
Try one. See what happens. The best part is usually the follow-up conversation — when you push back, ask why, or say "I don't buy that."
1. Steelman the Opposite
We all have opinions we hold pretty firmly. This prompt asks the AI to make the strongest possible case against one of yours — not a weak, easy-to-dismiss argument, but a genuinely compelling one.
It's called "steelmanning" — the opposite of a straw man. A straw man is a weak version of the other side. A steel man is the strongest version of it.
This isn't about changing your mind. It's about understanding exactly why someone reasonable might disagree with you.
Prompt
I believe [belief]. Please make the strongest possible case against this — not the easy objections, but the most compelling, thoughtful argument for the opposite position. Don't hold back.
Try it with something you feel strongly about. You don't have to agree with what it says. Just sit with it.
2. Explain It to Someone from 1850
Pick any modern idea and ask the AI to explain it to someone who lived 170 years ago. Someone smart, curious, and educated for their time — but with none of our context.
This exercise strips away jargon and forces the AI to find the essential core of an idea. It's also a surprisingly good test of whether you really understand something.
Prompt
Explain [concept] to a curious, intelligent person living in 1850. They have no context for modern technology or science beyond what existed then. Use only concepts and comparisons they would understand from their own world.
Explaining something to a stranger from another era reveals what you actually understand about it — and what you've just been taking for granted.
3. The 100-Year Lens
We are all living through things that will look very different from 100 years away. Ask the AI to imagine historians in the year 2125 looking back at something happening right now and writing about it the way we write about history today.
Prompt
It is the year 2125. A historian is writing a chapter about the early 21st century debate over [topic]. Write that chapter's opening two paragraphs, showing how people of that era are likely to look back on this moment.
Things we fight about passionately often look obvious — or obviously wrong — in hindsight. The question is: which side of this moment are we going to be embarrassed about?
4. Viewpoint Swap
Take any conflict, disagreement, or news story and ask the AI to tell it from each person's point of view. Not to say who is right. Just to explain how each side genuinely sees the situation.
Prompt
Here is a situation: [situation]. Can you explain how each of the main parties in this situation sees it? What do they believe they are doing, and why? Try to represent each perspective fairly, even the ones you might personally disagree with.
Understanding how someone sees a situation is not the same as agreeing with them — but it usually leads somewhere more useful than dismissing them.
5. The Alien Anthropologist
Ask the AI to explain some part of ordinary human life from the perspective of an alien researcher visiting Earth for the first time. This sounds funny. It often is funny. It also produces some of the most genuinely insightful observations about things we've stopped noticing because they're too familiar.
Prompt
You are an alien anthropologist who has just arrived on Earth and is trying to understand how [institution] works. You have no cultural assumptions — you are observing this purely as an outside researcher. Write a few paragraphs from your field notes describing what you observe and what hypotheses you have about why humans behave this way.
Try it with anything you've stopped questioning: hospitals, weddings, the news, how people talk about their jobs.
6. Find My Blind Spots
This one takes a bit of courage. Share your take on something — a plan you're making, a belief you hold, a decision you're facing — and ask the AI to tell you what you might be missing or assuming without realizing it.
Prompt
Here is my thinking about [topic]: [my view or plan]. What might I be missing? What assumptions am I making that I might not have examined? What would someone who disagrees with me say I'm not seeing?
The AI isn't always right about this — but it's often usefully surprising. It's good at noticing when a line of thinking has a gap that seems obvious from outside.
These six prompts all share something: they're less about getting an answer and more about seeing a question differently. Try one today. Then follow up. The conversation after the first response is usually where things get interesting.