Why I Write the Curse Instead of the Cure

I have written four novels across three universes, and somewhere around the third I noticed I had been writing the same book the whole time. Different gods, different centuries, different crimes. The same fear underneath all of it: power showing up before the wisdom to hold it. Capability arriving ahead of conscience. I did not set out to write that. It is just where my hand goes when I am not watching it.

The clearest place it surfaces is immortality. In the mythology, a man is granted endless life as a reward, and I could not make myself write it as a gift. It kept turning into a sentence. Not the kind you serve and finish — the kind that keeps running after everyone who could pardon you is gone. He outlives the people who knew why he did what he did. He outlives the meaning of his own choices. By the fourth century he is not blessed; he is the last witness to crimes nobody else remembers, carrying a context no living person can verify. That is a man who can no longer be forgiven because there is no one left with standing to forgive him.

Most stories about more — more years, more reach, more power — are written from the moment of acquisition. The coronation. The breakthrough. I am bored by that moment. It is the easy frame. The honest frame is forty years later, when the thing you built is still running and you are still answerable for it and the original reasons have eroded out from under you. I write the cost because the cost is where the truth is. The gift is just the marketing.

This is the same instinct I bring to engineering, and I stopped pretending the two were separate a long time ago. The thing I have learned running systems behind real revenue is that consequences outlive the act that caused them by an order of magnitude. A decision takes an afternoon. The schema it commits to lives for years. A shortcut taken at 2 a.m. under pressure becomes the load-bearing assumption nobody is allowed to question by the time the company is large. The act is brief. The consequence is patient. It waits.

I have sat on the incident bridge at 2 a.m. and watched a single line of code, written months earlier by someone who has since moved on, take down a service that thousands of people depend on. The author is gone. The intent is gone. What remains is the consequence, fully alive, indifferent to the fact that nobody meant it. That is the closest thing I know to my immortal man — a choice that survives its maker and keeps acting in the world long after the reasons have left the building.

So when I build, I build for the conscience to keep pace with the capability. That is what recovery docs are. What incident reviews are. What on-call rotations are — a way of making sure someone is still standing in the room with the consequence when the cost finally comes due. We cut customer-impacting incidents by forty percent not by getting faster but by refusing to let capability run ahead of accountability. The tooling can outrun us. The judgment cannot be allowed to.

The current moment makes this less of a metaphor than I would like. We are shipping capability faster than at any point in my career — AI writing code, generating systems, compressing months into weeks. I have measured the velocity gain and I will defend it. But velocity is exactly the condition under which capability outpaces conscience, and that gap is the whole subject of every book I have written. The machine can now produce the act in seconds. The consequence still takes years. Nothing has closed that distance; if anything we have widened it and called the widening progress.

That is why I write the curse instead of the cure. Not because I am pessimistic — I run teams that ship, I bet my time on the upside every day. But the cure is the part everyone already wants to hear, and the curse is the part that keeps people honest. My job, in fiction and in engineering, is to stand at the far end of the decision — years downstream, where the cost has fully arrived and the reasons have gone quiet — and report back what is actually there. Someone has to write from forty years later. I would rather it be me, on the page, than the system teaching it to all of us the expensive way.