Take three minutes. Open your last meaningful AI conversation — not a one-off question, but one where you were actually working through something.
Read through it once. Then answer these ten questions.
TL;DR
This isn't a quiz with right answers. It's a mirror. The point isn't to score yourself — it's to notice patterns you couldn't see while you were inside the conversation.
The Questions
1. How many exchanges did you have?
Count them. If it was 1–2, you had a transaction. If it was 5+, you had a conversation. Neither is inherently wrong — but if most of your AI interactions are 1–2 exchanges, you're consistently stopping before the interesting part starts.
2. Who framed the problem?
Did you come in with a predefined question, or did you and AI arrive at the real question together? If you defined the problem and AI solved it — that's delegation. Powerful, but limited. The highest-leverage conversations are the ones where the problem itself gets refined.
3. Did you push back on anything?
At any point, did you say "I'm not sure about that," "what about the opposite," or "that doesn't match my experience"? Pushback is the single strongest predictor of insight quality. If you accepted everything AI said — ask yourself why.
4. Were you surprised?
Did AI show you something you hadn't considered? A connection you hadn't made? A perspective that genuinely shifted your thinking? If not, you may have kept the conversation too narrowly within your existing frame.
5. Did you ask AI to challenge you?
There's a difference between AI volunteering a caveat and you actively requesting a counterargument. The first is passive. The second is a deliberate choice to stress-test your thinking. Did you make that choice?
6. Did you bring in another domain?
Did you at any point ask AI to apply a concept from a different field? "How would a psychologist see this?" "What would a game designer do?" Cross-domain thinking is where the most novel insights come from — and it's the mode people use least.
7. Did AI change your mind about anything?
Not confirm what you already thought. Actually change it. If AI never changes your mind, one of two things is true: you're always right (unlikely), or you're not creating space for AI to challenge your views.
8. After the conversation, did you reflect on it?
When the conversation ended, did you think about how you interacted — or did you immediately move on to the next task? Metacognition — thinking about your thinking — is the skill that makes all other AI collaboration skills compound.
9. Would you approach it differently now?
With the benefit of hindsight, would you start the conversation differently? Frame the problem differently? Push harder on something you accepted? If yes — that gap between how you did approach it and how you would approach it is your growth edge.
10. What mode were you in?
Were you in execution mode (get this done) or exploration mode (help me understand)? Were you delegating, exploring, challenging, or synthesizing? Can you even name the mode?
What Your Answers Reveal
You don't need to score this. The patterns will be obvious.
If most of your answers point toward short, delegation-focused, no-pushback conversations — you're in the majority. That's where 80% of people are. It's not a failing. It's a starting point.
If you notice you never push back, never cross domains, and never reflect afterward — those are three specific levers you can pull tomorrow.
| Pattern | What it suggests | One thing to try |
|---|---|---|
| Short conversations (1–2 exchanges) | You're transacting, not collaborating | Add two more exchanges to your next important conversation |
| No pushback | You're consuming, not engaging | Challenge one thing AI says, even if you agree |
| No surprise | Your frame is too narrow | Ask: "What am I not seeing?" |
| No cross-domain | You're staying in your lane | Ask: "How would a [different field] approach this?" |
| No reflection | You're not compounding | Spend 60 seconds reviewing the conversation afterward |
The Meta-Point
The fact that you took this test means something. Most people never examine their AI interactions at all. They prompt, they receive, they move on.
You paused. You looked. That's the first step toward a different relationship with AI — one where you're not just using it, but growing through the way you use it.
If you want a deeper, automated version of this reflection — one that analyzes your actual conversation patterns across all eight dimensions of AI collaboration — that's what the AI Leverage Mirror does. But this 3-minute exercise, done regularly, will change your AI collaboration more than any tool.
The best mirror is the one you use.