You have probably had this evening. You leave a dinner, a party, or a long work lunch, and on the drive home you feel oddly empty. Nothing went wrong. Everyone was pleasant. You talked for three hours, and you could not say one true thing you learned about anyone, or one true thing anyone learned about you. You were surrounded by people and you came home a little lonelier than you left.

Here is the uncomfortable part. The person across the table probably drove home feeling the same way. Each of you wanted to say something real, and each of you held back because you assumed the other one didn't want to hear it.

That assumption is the subject of this article. It is common, it is measurable, and it is usually wrong.

The quiet bet we make in every conversation

Every time you choose between asking So, what do you do? and asking What's something you've been thinking about a lot lately?, you are making a prediction. You are guessing how the other person will react. Will they find the deeper question intrusive? Will the silence afterward be awkward? Will you seem strange?

Most of us make that bet the same way. We pick the safe question, and we stay in the shallows.

In a 2022 set of experiments published in the Journal of Personality and Social Psychology, psychologists Michael Kardas, Amit Kumar, and Nicholas Epley tested whether that bet is a good one. They paired strangers and had them talk about either shallow topics or deep ones, the kind of questions that invite real disclosure, like what someone is grateful for or a time they cried in front of another person. Before the conversations, participants predicted how awkward they would be, how connected they would feel, and how much they would enjoy it. Afterward, they reported what actually happened.

The pattern held across their studies. People expected deep conversations to be more awkward and less enjoyable than they turned out to be. They underestimated how connected they would feel. And, crucially, they underestimated how interested their partner would be in what they had to say. The researchers called these miscalibrated expectations, and argued they create a barrier to the very conversations people would enjoy most.

In other words: the wall between you and a real conversation is mostly made of a forecast, and the forecast is pessimistic.

Why we underestimate other people

If deep conversation is so rewarding, why do we keep predicting otherwise? A few well-documented tendencies stack on top of each other.

We assume others are less curious about us than they are. In the Kardas, Kumar, and Epley work, one of the most striking gaps was not about awkwardness but about interest. People seemed to believe their inner life was a burden to hand over, something the listener would tolerate rather than welcome. It connects to a broader finding in the same research tradition: we tend to think other people care less about the substantive parts of us than they actually do.

We overweight the first few seconds. The moment of discomfort when you ask a real question is vivid and immediate. The connection that follows ten minutes later is harder to imagine in advance. So the forecast is dominated by the part that stings.

We think we're the only one who wants more. This is a form of what psychologists call pluralistic ignorance: everyone privately holds a preference, everyone assumes the others don't share it, and so everyone publicly behaves in a way nobody actually wants. A whole room of people quietly hoping someone will go first.

This isn't limited to deep questions. In a well-known 2014 study, Nicholas Epley and Juliana Schroeder asked commuters on trains and buses in Chicago to either strike up a conversation with a stranger, sit in solitude, or do whatever they normally did. Commuters predicted that talking to a stranger would be the least pleasant option. Those who actually did it reported a more positive commute. The pessimism was in the prediction, not the experience.

What depth actually does for you

It would be fair to ask whether any of this matters beyond a nicer evening. There's suggestive evidence that it does.

In a 2010 study in Psychological Science, Matthias Mehl and colleagues had participants wear a small device that recorded snippets of ambient sound throughout their days. Researchers coded the conversations they captured as small talk or substantive. People who reported higher well-being spent less of their time in small talk and more in substantive conversation. The study was correlational, so it cannot say depth causes happiness. Happier people may simply have more meaningful conversations. But it fits with what you already know from the drive home: shallow talk can fill a whole evening without feeding anything.

There's also older experimental work on how closeness forms. In the late 1990s, Arthur Aron and colleagues developed a procedure, now widely known as the 36 questions, in which pairs of strangers took turns answering questions that gradually escalated in personal depth. Compared to pairs who exchanged small-talk questions, those who went through the escalating disclosure reported feeling closer. The mechanism wasn't magic. It was reciprocal, gradual self-disclosure: each person revealing a little more, and being met when they did.

That word, gradual, matters. Depth is not a cliff you jump off. It is a staircase you climb together.

Curiosity is the bridge

Here is the reframe that makes this easier. Going deeper does not require you to confess something dramatic or to interrogate anyone. It requires you to get genuinely curious about the person in front of you, and then to ask the question your curiosity actually wants answered.

Most small talk is not curiosity at all. It is ritual. Where are you from? What do you do? How was your weekend? These questions have expected answers, and both people are reciting lines. Deep questions are different because they have no script. When you ask someone what they've changed their mind about recently, neither of you knows what's coming. That uncertainty is exactly what makes it feel risky, and exactly what makes it feel alive.

You can also lower the risk by going first. Disclosure tends to invite disclosure. If you offer something honest about yourself, a small worry, an odd fascination, a question you can't stop turning over, you give the other person permission to do the same. You are not demanding vulnerability from them. You are showing them the door is open.

And when the forecast in your head says they won't care, you can now answer it with evidence: people consistently underestimate how much others want to hear this. Your nervousness is real. It just isn't a reliable guide.

Your next moves

  • Replace one ritual question today. Next time you'd ask How was your weekend?, try What was the best part of your weekend, and why that? The follow-up why is where the real answer lives.
  • Write down three deeper questions and keep them on your phone. For example: What's something you're looking forward to that most people don't know about? What's a belief you've changed in the last few years? What's been on your mind more than usual lately? Having them ready removes the in-the-moment hesitation.
  • Go first with one honest sentence. Before asking someone something personal, offer a small, true disclosure of your own. It doesn't have to be heavy. I've been weirdly anxious about turning forty is enough to change the temperature of a conversation.
  • Run a prediction check. Before your next coffee or call, jot down how awkward you expect a deeper question to feel on a scale of 1 to 10. Afterward, rate how it actually felt. Do this three times and look at the gap. You are recalibrating your forecast with your own data.
  • Stay one beat longer. When someone gives a real answer, resist the urge to immediately relate it back to yourself. Ask one more question about what they just said. That single extra beat is often the difference between a nice exchange and a memorable one.

Where the questions come from

The hardest part of a deeper conversation is usually the first question. Not because you lack curiosity, but because in the moment, your mind reaches for the safe script. That's the quiet problem curiokit was built around: it gives you a steady supply of genuinely interesting questions, the kind that open people up rather than close them down, and helps you notice and save the ones that make you curious yourself. If you'd like a little help climbing that first step, you can explore it at curiokit. Either way, try one real question this week. The person across from you is probably hoping someone will.