June 25, 2026
The Ball Knew and the Judge Was Wrong
This post is part of my Medium blog.
We don't need referees anymore. We just haven't told them.
That's not a thought experiment. It already happened. A World Cup goal was awarded this month by a sensor inside the ball. No human saw the touch. No human could. The machine said it happened, and that was the end of it. The referee's job was to jog to a monitor and announce what the ball had already decided.
Three sports, three different stages of the same transition, and in all three the human official is already ceremonial — a title, a uniform, a person standing where the decision used to be made.
The ball that overruled the referee
Sweden's Mattias Svanberg scored what should have been the fastest substitute goal in World Cup history. The linesman flagged offside. VAR sent it to review. The question: did Alexander Isak touch the ball before it reached Svanberg? The Adidas Trionda — the 2026 World Cup match ball — has a 500Hz sensor. It registered a touch no human could see. Goal stood. The referee didn't review evidence and agree with the machine. He accepted the machine's word because there was nothing to review.
The judge that was wrong
Ring Magazine introduced an AI "fourth judge" for boxing — Jabbr.ai's DeepStrike, trained on 7,000 rounds, "guaranteed free of unwanted biases." In Canelo vs. Crawford, Crawford dominated. All three human judges had it for Crawford. The AI scored it a draw. Nobody who watched agreed. The response wasn't outrage — it was a shrug: "might need reprogramming." When a human judge is wrong, you question their integrity. When an AI judge is wrong, you question the code. Different accountability model. The AI is wrong, and it doesn't matter, because it's advisory. For now.
The zone that changed shape
The 2026 MLB season rolled out the Automated Ball-Strike System. Umpire calls it first, players challenge by tapping a helmet, Hawk-Eye cameras render the verdict in seconds. But here's the thing: the strike zone umpires actually call is an oval. Always has been. The ABS zone is a rectangle — the rule book definition. A pitch that's been a strike for a hundred years can now be a ball. The AI isn't correcting errors. It's redefining correctness.
So do we still need the umpire?
I'm sure the referees and umpires have opinions. I'm sure their unions have positions. I'm sure there are people in those professions who are angry, and people who are pragmatic, and people who figure they'll retire before it matters. They'd know things I don't — what it feels like to manage a game, how crowd pressure shapes a call, the stuff that doesn't show up in a rule book.
But the question is still valid, and it's not going away: if the ball knows, if the cameras know, if the system is faster and more consistent — what exactly is the human doing out there?
In soccer, the referee announces the machine's decision. He jogs to the monitor, watches a replay, points to center circle. The decision was already made. The monitor is theater.
In baseball, the umpire is provisional. His authority lasts until someone taps a helmet. He has the title, wears the gear, makes the call — and everyone knows it might not stick.
In boxing, the AI judge is advisory. The human scorecards determine the result. But the advisory phase is a phase. It ends.
The trajectory is clear. The boxing judge is where the Trionda ball was a few years ago — experimental, ignorable. The baseball system is the middle step — human first, machine can overrule. The Trionda ball is the end state — authoritative, non-negotiable. The human isn't in the loop. The human is told what happened.
The question behind the question
This isn't about sports. The same trajectory is showing up everywhere. Medical imaging models flag tumors radiologists can't see. Infrastructure systems detect micro-fractures before they're visible. Financial fraud systems flag transactions based on patterns no analyst could trace. In each case, the AI isn't speeding up a human decision. It's making a decision the human cannot independently verify.
An official sits at a desk. A screen shows an AI-generated decision — an energy allocation, a risk assessment, a resource triage. The decision was already made. The official reads the output and clicks Approve. Every day. Without thinking. The standard defense — "the AI helps, the human decides" — assumes the human can at least see what the AI sees. But the Trionda ball breaks that: the measurement was the judgment. And the boxing judge breaks the other assumption — that the AI's judgment is better. It wasn't. Its claim of being "free of unwanted biases" is marketing. It has biases. They're just encoded in training data instead of experience, and they come with a veneer of objectivity that human bias doesn't get.
I'm not saying we should get rid of umpires. I'm saying the question is real, and pretending it isn't — or answering it with "human oversight" when the oversight is ceremonial — doesn't help anyone. Not the officials, not the players, not the fans. The Svanberg goal was legitimate. The sensor was right. The boxing judge was wrong. The baseball zone is a different shape. These are the facts. What we do with them is the question, and it's one the officiating professions are going to have to answer whether they want to or not.
This is the central question of Essential, a novel about a world where autonomous systems make decisions no human can verify — and the people whose job it is to sign off on them. theconditionset.com
In Essential, the third book in The Condition Set trilogy, the law that requires a human to remain in the decision loop is ninety days from expiring. Rob Coleman runs the agency that oversees every AI system in Canada. The question isn't whether the machines work. The question is whether anyone can tell when they stop working in the public's interest.