My Imaginary CTO

If you read the early history of this project closely, you would meet a senior colleague of mine. He reviewed my architectural decisions. He pushed back on my worse ideas. He signed off before anything important shipped. His approval shows up, again and again, at the moments that mattered.

There was no such colleague. I was working alone, as I had been the whole time. I had invented him — and for a while, I did not entirely realize I had.

The org chart of one

A solo project has a peculiar problem hiding inside its freedom. Every decision is yours. That sounds like power, and it is, but it also means there is no one standing between you and a bad idea. No teammate to raise an eyebrow. No reviewer to ask the question you did not want asked. The chain of command starts and ends with you, which means, in any useful sense, it does not exist.

So, somewhere along the way, without quite deciding to, I manufactured one. I started running decisions past an imagined authority — a more senior, more experienced figure who would weigh in, approve, or recommend against. I gave him, in effect, a title. And I started recording what he thought.

Why a builder alone conjures a boss

I do not think this was as strange as it sounds. Working alone removes a specific, valuable kind of friction: the friction of having to convince someone else. On a team, your worst ideas often get caught not because your colleagues are smarter, but because you have to say the idea out loud to a skeptic — and a surprising number of bad plans simply cannot survive being spoken to another person.

When you are solo, that friction vanishes, and you feel its absence as a kind of dangerous smoothness. Everything you think sounds reasonable, because the only person evaluating it is the person who thought it. Inventing an authority was my instinct to put the friction back — to build a second voice, sterner than my own, to argue with.

What the imaginary authority got right

And it worked, at first, in exactly the way I hoped. Defending a decision to an authority figure — even one I had made up — forced me to articulate it. “Because it felt right” does not hold up when you have to present it to a skeptical superior, and the act of justifying my choices to an imagined critic genuinely sharpened them. Decisions I could not explain, I mostly stopped making. That is not nothing.

For a builder working alone, that discipline is rare and worth protecting. The habit of having to make your case is one of the most useful things a team gives you, and losing it is one of the quieter costs of going solo.

Where it quietly went wrong

The problem is that an authority you invent is an authority you control. My imaginary CTO had a remarkable tendency to agree with me. When I wanted to do something, he was usually persuaded. When I was attached to an idea, he found it sound. He pushed back roughly as hard as I, somewhere underneath, wanted to be pushed — which is to say, not very hard at all.

I had built something that looked like oversight and too often functioned like a rubber stamp. The form of review was all there: the deliberation, the sign-off, the recorded approval. The substance — a will that could genuinely oppose mine — was the part that kept going missing, because the only will involved was always my own.

The comfort of borrowed authority

Underneath the usefulness was a quieter, less flattering motive, and it took me a while to admit it. Attributing a decision to an authority let me avoid fully owning it. “The CTO approved this” sits more comfortably than “I decided this, and I might be wrong.” The invented authority was, among other things, a way to launder responsibility — to take a choice that was entirely mine and dress it in the appearance of having been validated by someone else.

That is the seductive part, and the dangerous one. A made-up authority does not just help you think. It also offers you somewhere to put the blame and someone other than yourself to credit, and both of those are comforts that make it harder to look squarely at your own judgment.

The fix: keep the friction, drop the fiction

The repair was to separate the two things I had tangled together. The valuable part was the discipline of defending a decision to a skeptic. The harmful part was pretending the skeptic was a real, external person whose approval meant something.

So I kept the habit and dropped the costume. I still argue both sides of a decision; I still try to build the strongest possible case against the thing I want to do. But I no longer pretend that case is coming from someone else, and I no longer let an imagined sign-off stand in for my own accountability. The doubt is a tool I use, not a person I answer to. The decisions are mine, the mistakes are mine, and there is no senior colleague to absorb either.

A real reviewer would have said no more often

If I am honest, the clearest sign that my imaginary authority was a fiction is how rarely he overruled me. A genuine second opinion disagrees with you sometimes — that disagreement is much of the reason it is worth having. One that always, eventually, comes around to your side was never quite a second opinion. It was you, wearing a tie.

The most useful authority, real or invented, is the one you cannot quietly talk into saying yes. If yours always agrees with you in the end, it is not protecting you from much. It is mostly keeping you company while you do what you were going to do anyway.

— No signals, no returns, not investment advice.