Calibrating Confidence: On Learning to Own Your Expertise in Public

Taking Up Space was about the external dynamics: the room, other people, the accumulated cost of having your ideas ignored or absorbed or overridden. This is the other side of the same problem, the internal one.

The external dynamics are real and structural. But there’s a version of the same pattern that has nothing to do with the room you’re in, and it runs on the same engine. It’s the quiet instinct to qualify before you’ve confirmed you’re right, to wait until you’re really sure before saying anything, to shrink the claim before anyone else gets a chance to. When it’s operating, it’s hard to distinguish from genuine self-awareness. It feels like appropriate modesty. It usually isn’t.

What imposter syndrome actually looks like at senior level

The entry-level form is “I don’t know enough to do this job.” Uncomfortable and usually inaccurate, but it has a clear counter-signal. You accumulate evidence that you do know enough, and eventually the feeling recedes into background noise.

The senior version is different. It’s not “I don’t know enough to do this” but “I don’t know enough to say this publicly.” The competence isn’t in question anymore, but the right to claim it still feels like it is.

I’ve watched this play out in myself consistently over the past year. The Terraform post I published earlier this month took longer to write than it needed to because I kept returning to the question of whether I was the right person to write it. I’ve been doing this work for years. I have direct experience with the specific problem. The post is accurate. The hesitation had nothing to do with whether the content was right.

The pattern in practice

Publishing the display-profiles tool last week brought it up again. It’s a collection of shell scripts solving a specific display switching problem on Linux. The usefulness to someone else is real but modest. The resistance to publishing it was entirely disproportionate to the stakes.

The internal monologue sounds like: it’s too small. It’s too specific to my setup. Someone else has done this better. The README needs more work. The coordinate normalisation has an edge case. By the time you’ve gone through the whole list, weeks have passed and the thing that would have been useful to someone is still sitting in a private repo.

This version of the pattern disguises itself as quality control. It sounds like “I’m waiting until this is actually good enough,” which is a reasonable thing to think about some work. The tell is whether the threshold keeps moving. If the objections are never resolved and there’s always one more reason to wait, it isn’t quality control. It’s pre-emptive self-erasure wearing a more respectable name.

And with open source specifically, it’s backwards. The whole point is that someone has an idea, does the initial implementation, lays the groundwork, and publishes it. The tool doesn’t need to be finished. It needs to exist so that other people can find it, use it, contribute to it, and take it somewhere the original author never anticipated. That’s how software spreads and becomes the default. Git started as Linus Torvalds solving a very specific problem and releasing something workable. Waiting until it was complete would have meant never releasing it at all.

The useful distinction

There is a genuine and important role for calibrated self-doubt. Knowing the limits of your knowledge, flagging uncertainty when it exists, being willing to be wrong and say so clearly: all of that is worth cultivating. The people who cause the most damage in engineering organisations are often the ones who don’t have enough of it.

What’s different is that calibrated doubt is responsive to evidence. You’re uncertain about a specific claim because you haven’t verified it. You flag the uncertainty, check, and update. The doubt has a direction and it moves toward resolution.

Pre-emptive self-erasure doesn’t work like that. It fires before any evidence is gathered. It assumes the worst-case assessment of your work before any assessment has happened. It isn’t doubt. It’s anticipatory capitulation, and the only thing that reliably breaks it is doing the thing anyway and watching the sky not fall.

What actually worked

Writing the posts and publishing the tool have been the most effective counter-signals. Not because of any response to them. I haven’t sought one, and I’m not measuring this by whether anyone noticed. The point was to put the work out there and get used to being in public. The shift comes from having done it, not from what came back.

That’s the part that’s easy to get wrong. Waiting for external validation as the signal that you were right to publish is just another form of the same pattern, deferred rather than broken. The belief being updated isn’t “my work is good enough.” That was never really the question. The belief being updated is “it’s acceptable for me to claim expertise in public.” That one doesn’t respond to positive feedback. It responds to the repeated experience of publishing something and the world continuing to turn normally.

At some point, continuing to wait for certainty before claiming what you know becomes a choice rather than a reaction. The window for being a beginner has passed. The caution that made sense early in a career becomes, somewhere along the line, a habit that’s outlived its usefulness.

The trailing qualification is not protecting you from anything. Claim the expertise you have.