By
·
6 minute read
For the most part, we’ve agreed as a society that stereotyping people is painting the complexity of humans with super broad brushstrokes and is a bit of an ignorant move. But somehow, dress up that same move with a little therapy speak and now it’s acceptable.
Your boss didn’t answer an email, so that’s an avoidant attachment style. A coworker asks a few too many follow-up questions, so he’s gaslighting you. Someone is difficult and self-centered, so now they’re a narcissist.
We love these labels because they give us instant certainty about a confusing situation.
I’m not telling you to stop. Partly because you won’t, and partly because pattern recognition is an asset. Your brain is built to recognize what it has seen before and predict what comes next. The problem isn’t the first judgment. It’s what happens after it.
Once you have the theory, you start using it to explain everything that happens afterward. What started as a useful shortcut becomes the lens you read everything else through.
The point isn’t to stop making first reads. It’s to hold an evolving model.
What a mental model actually is
A mental model is a working theory about how something works. You have them for people, your industry, your company, food, money, and yourself. They give your brain a starting point instead of making it work everything out from scratch every time.
And they’re supposed to be incomplete.
The problem is that we forget that part. The theory does its job on day one and never gets revisited.
An evolving model is a working theory you expect to get more refined as you gather information. You can act confidently from what you currently believe while knowing your picture is still crude. You don’t have to suspend judgment until you have perfect information. You make the best call you can from what you have, then stay willing to take in what comes next.
That’s what it means to stand in confident uncertainty: to act from the model you have while knowing it isn’t finished. You don’t wait for new information to change your mind. You go looking for information that could change your read.
You can still make decisions. You can decide that a coworker is difficult to work with. You can decide not to trust someone with a particular responsibility. You can decide that a strategy isn’t working. The question is whether you can make that decision without getting invested in the explanation you built around it.
Research on judgment has found that people who are more willing to actively seek out information that could challenge their existing views tend to make more accurate judgments. Being smart doesn’t mean you will do this. In fact, the research found that being open-minded was associated with greater accuracy even when intelligence was taken into account. [1][2]
Why we close so fast
Not knowing is uncomfortable. A confusing situation sits there without an explanation, and a label ends it. “Narcissist” gives a chaotic interaction a name. “Avoidant” gives an unanswered email a reason.
And there’s a little relief in it. If he’s the narcissist, I’m the reasonable one.
Once you’ve decided what someone is, you don’t have to keep figuring them out. You can read whatever happens next through the model you already built. That’s efficient. It’s also how a useful model becomes a distorted one.
There’s another reason it’s hard to reopen a read. The read becomes part of the story you’ve built about yourself in relation to that person.
If I decide my boss is avoidant, I’m not only forming an opinion about my boss. I’m forming a story about what happens between us. I’m the person who communicates clearly. He’s the person who doesn’t.
Now anything that challenges his label also challenges how I understand the relationship. It asks me to reconsider him and the role I gave myself in the story.
That’s harder to give up.
And the read costs him something too. He might genuinely be avoidant. But if I decided that from one unanswered email, that’s what he becomes. To me, and to anyone I tell.
A provisional interpretation has quietly become a reputation.
What evolving models look like
The person who is good at this doesn’t necessarily look uncertain. They often look more grounded than everyone else in the room.
They’ll tell you clearly where they stand. Then, when new information comes in, they adjust their position without treating the adjustment as a loss of status. You can watch them refine their thinking in real time without the flinching that usually happens when someone feels their position moving in public.
They aren’t trying to preserve the appearance of having known all along. They want the better understanding more than they want the credit.
That’s confident uncertainty. It comes from being comfortable acting on a model that you know may need to change.
Why this is getting harder
We aren’t taught much about how to do this. We’re taught to make decisions, defend our positions, communicate with confidence, and project authority. We’re taught much less about what to do when the information changes.
Culturally, we also mistake certainty for authority. The leader with the clean answer sounds stronger than the leader who says, “Here’s what I think based on what we know. Here’s what would change my mind.” When a leader updates their position in light of new information, we read it as weakness or indecision rather than as evidence they’re paying attention.
Then there are the confirmation loops around us. Algorithms give us more of what we already engage with. Feeds compress complicated situations into short clips, declarations, and hot takes. Even our conversations can be self-reinforcing. You tell someone your interpretation, they agree, and now the interpretation feels more true.
AI makes this particularly easy. You give a system your framing and it hands that framing back to you in a cleaner, more coherent form. It can feel like independent confirmation when you’re really just hearing your own premise expressed more convincingly.
Disconfirming information used to arrive by accident. Increasingly, you have to go find it.
If everything around you reinforces the models you already have, evolving them becomes something you have to do on purpose.
Start with the read you’re most certain about
People are good material for this because they’re complicated and the feedback keeps coming. They’re also hard because you have something at stake, which makes them a better test than a market analysis.
Take the boss you decided is avoidant. Don’t try to talk yourself out of it. That isn’t the exercise. Go looking for what the read is missing.
Maybe he doesn’t answer anyone’s email, not just yours. Maybe he answers when a decision is needed and goes quiet when it’s an update. Maybe he’s still hard to get a straight answer from, but for a different reason than you assumed.
The original read is still useful. You’ve simply added enough information to make it more precise. You’re increasing the resolution of the model rather than swapping one conclusion for another.
Write the original read down. Record what you expected to happen and what actually happened. Your brain will edit the history otherwise. Once you know how something turned out, it’s remarkably easy to remember that you always knew. A written record makes that harder. Over time it shows you where your models were too broad, where they held up, and where reality made you get more precise.
You’re giving your brain repeated experiences of one thing: my model was incomplete, I refined it, and nothing bad happened.
Start with situations where you don’t have much at stake. Pay attention to the moments when your first interpretation turns out to be incomplete, and practice revising it. With enough repetition, you get better at doing the same thing when the stakes are higher.
Someone shuts down your idea in a meeting. Someone questions your expertise. A customer reacts differently than you expected. A strategy you believed in stops producing what you predicted. The emotional charge makes the original read harder to loosen, which is exactly why those moments matter. Instead of defending the model, ask what you haven’t accounted for.
People are just where you start. The same ability applies to a strategy, a business, a market, a decision, or your own read on yourself.
Where certainty should live
When you get good at evolving your models, you don’t lose certainty. You move it off the model and onto your ability to refine it.
You can know what you think, make a decision, and stand behind it without needing the explanation to stay fixed forever. You can run into information that changes the picture without feeling it as a threat to your judgment or your authority.
That confidence comes from trusting that you can keep learning, adapting, and refining your understanding as you gather more information.
Your first read will always be there. That’s part of being human. The difference is knowing that your first read is where the model starts, not where it ends.
Bibliography
1. Haran, U., Ritov, I., & Mellers, B. A. (2013). “The Role of Actively Open-Minded Thinking in Information Acquisition, Accuracy, and Calibration.” Judgment and Decision Making, 8(3), 188–201.
Cambridge University Press
2. Stanovich, K. E., & West, R. F. (1997). “Reasoning Independently of Prior Belief and Individual Differences in Actively Open-Minded Thinking.” Journal of Educational Psychology, 89(2), 342–357.
Original paper
AI Disclaimer: The ideas are mine. AI did a line edit, which means it may have improved my writing, ruined my voice, or added a little AI wisdom where none was requested. If something sounds suspiciously like a robot, blame the robot.