Questions to Ask a Smart Person
For a short conversation with someone whose judgment you respect, in a hallway, over coffee, or in a scheduled half hour. These questions aim at how they think and decide rather than at facts you could look up afterwards.
The questions
Open any question for the note
What are you thinking about at the moment that you haven't finished working out?
Why ask it
Unresolved thinking is where someone is most interesting and least rehearsed, and it invites you into the problem rather than the conclusion. A polished answer here usually means you got the standard talk instead of the person.
What's a belief you held strongly and have since abandoned?
Why ask it
The useful part is not the belief but what changed their mind: an argument, a dataset, a bad experience. People who cannot name a single reversal are either not tracking their own errors or not telling you.
How did you actually learn what you know, as opposed to how you're supposed to have learned it?
Why ask it
The real path usually involves an odd job, a mentor, a period of obsession, or years of doing something badly. Credentials rarely explain competence, and the honest version is far more usable.
When you're new to a subject, what do you do in the first week?
Why ask it
This gets at concrete habits: skimming three overviews, finding the best critic, talking to a practitioner before reading anything. Vague answers about being curious mean there is no method to copy.
What's something widely believed in your field that you think is wrong?
Why ask it
You find out where they diverge from consensus and, more importantly, what evidence they think settles it. Someone who cannot disagree with their own field has probably stopped examining it.
How do you decide what to ignore?
Why ask it
Attention rather than intelligence is usually the scarce resource, and people who do good work tend to have explicit rules for what they refuse to read or attend. The rules are more transferable than their conclusions.
What's a mistake you made that you'd make again, given what you knew then?
Why ask it
This separates genuine misjudgment from bad luck, which most people conflate when telling the story afterwards. It also shows whether they evaluate decisions on process or on outcome.
When you disagree with someone you respect, how do you work out who's right?
Why ask it
The answer reveals a real procedure or the absence of one: seeking the strongest version of the other case, making a prediction, deferring on a domain basis. Watch for anyone whose procedure always ends with themselves.
What do you do when the evidence is thin and you still have to decide?
Why ask it
Most consequential decisions are made on incomplete information, and how someone handles that gap is the whole skill. Listen for whether they reduce the size of the bet rather than pretending to more certainty.
What's a question you've been asking yourself for years without resolving it?
Why ask it
Long-running open questions tend to be the ones organising someone's work quietly in the background. This is also where you are most likely to get something they have never said in a talk.
Who taught you the most, and what specifically did they do?
Why ask it
Naming the behaviour rather than the person gives you something you can imitate on Monday. If they cannot describe what the mentor actually did, the lesson probably has not been examined.
What did you spend a lot of time on that turned out not to matter?
Why ask it
The wasted effort is usually more instructive than the successful effort and less flattering, so it is rarely volunteered. It also tells you what to skip in your own version of their path.
How do you tell whether you're making progress on the things that matter to you?
Why ask it
People with clear answers here tend to have a small number of real measures rather than a dashboard. Vagueness often signals someone busy but drifting, which is worth knowing before you take their advice.
What's something you're deliberately bad at?
Why ask it
Chosen incompetence reveals priorities more sharply than a list of strengths, and the trade-off is usually explicit. It also tends to loosen the conversation, since the question is unexpected.
What kind of problem do you find yourself unusually good at?
Why ask it
Asking about problem type rather than skill produces a more specific answer, such as noticing when a question is badly framed. It also tells you what to bring back to them later.
What would change your mind about the thing you're most confident in?
Why ask it
A concrete falsifier shows the belief is held on evidence rather than identity. Someone who cannot name one is telling you something important without meaning to.
What are you reading or watching that has nothing to do with your work?
Why ask it
Cross-domain input is often where an unusual thinker's analogies come from, and the answer is specific and easy to follow up on. It also shifts the register away from interview mode.
What advice do you give that people usually ignore?
Why ask it
This surfaces the recommendation they believe most and have found hardest to transmit, along with their theory of why it does not land. Far better than asking for their best advice, which produces platitudes.
What should I be asking you that I haven't?
Why ask it
You hand over the agenda to the person who knows the terrain, and they will usually redirect you toward something you did not know existed. Save it for near the end, once they know what you are after.
Who else should I be talking to, and what should I ask them?
Why ask it
A specific name plus a specific question is the most valuable thing you can leave with, and it costs them almost nothing. A refusal to name anyone often means the conversation did not land as well as you thought.
Getting Value From a Short Conversation
Practical guidance for the conversation itself
How to Run It
Do the Reading First
Anything they have already written or said publicly is not worth their time to repeat. Arriving with one specific reaction to something they published changes the conversation immediately.
Ask About Process, Not Verdicts
How they decided is portable to your own situation, whereas what they concluded usually is not. When you get a verdict, ask what led to it.
Follow Where They Slow Down
People speed up on rehearsed material and slow down when they are actually thinking. The pause is the signal to ask one more question rather than move on.
Write It Up Within the Hour
Detail decays fast, and the specific phrasing they used is often the part worth keeping. Note what you disagreed with too, since that is where your own thinking is.
Adjust for the Setting
You Have Five Minutes
- 1What are you thinking about at the moment that you haven't finished working out?
- 2What advice do you give that people usually ignore?
- 3Who else should I be talking to?
They Are an Expert in Something You Are Entering
- 1What did you spend a lot of time on that turned out not to matter?
- 2What's something widely believed in your field that you think is wrong?
- 3When you're new to a subject, what do you do in the first week?
It Is a Recurring Relationship
- 1What's a question you've been asking yourself for years without resolving it?
- 2What would change your mind about the thing you're most confident in?
- 3What should I be asking you that I haven't?
Common Pitfalls
Turning It Into a Debate
Trying to win a point costs you the rest of the conversation, and you learn nothing you did not already believe. Disagree once, briefly, and mostly to test their reasoning.
Asking Questions With a Public Answer
Anything findable in ten minutes of searching wastes the scarce part, which is their unpublished judgment. Bring the searchable part with you as context.
Stacking Three Questions Into One
Compound questions get answered on whichever clause is easiest, and the interesting part gets dropped. Ask one thing and wait.
Performing Rather Than Listening
Long preambles establishing your own credentials eat the time you asked for. State the context in two sentences and stop.