Judgment Call

“We take option B,” he said.

He was my boss.

Option B meant continuing a project that had been largely unsuccessful so far, had already consumed considerable resources, and had nevertheless been extended several times.

The decision came at the end of a long meeting attended by senior managers, project leads, and subject-matter experts. There had been other options.

We could have stopped the project and cut our losses. Of course, that would also have meant finding a different solution to the original problem, and nobody really knew what that solution would have looked like.

We could also have pushed the project toward some kind of minimal consensus version with the existing resources. No masterpiece. Guaranteed disappointment on the business side. But at least another IT tool that nobody would have considered remarkable.

Instead: option B. Full speed ahead. More money. More people. More commitment.

I did not understand the decision. And from the naïve position of someone who was there to observe and learn from my boss, I asked him about it later. Why? That was essentially my question. Why put even more resources into this?

He was already moving on to the next topic, perhaps the one after that, and only glanced up briefly from his phone.

“I made a judgment call,” he said.

That was all I got. It was obvious that no further explanation would follow. I was not satisfied.

Then again, it was not my money. It was not my decision. I would not be the one held responsible for the consequences. And unlike him, I had not spent years in different companies managing budgets, leading people, approving projects, and making thousands of other decisions.

So if anyone was qualified to make that call, it was him.

In the end, the project turned out “okay” as far as I can tell. Whether the users were genuinely happy with it, I do not really know. But there was no major backlash either. The project was completed. The board seemed content with the outcome. And with a little glitter dust, it was certainly possible to treat it as a success.

Perhaps that was the lesson. Perhaps this is simply what it looks like when experienced managers make decisions.

Over the years I heard the phrase again and again.

A decision about the next board member.
A merger that could easily have failed.
A strategic investment where the data supported two options equally well.

“Judgment call.”

Judgment was the asset of executives. Something only they possessed, and something they had earned the right to use. Trained and authorized by thousands of situations like this one.

After the spreadsheet comes experience.

That is probably why judgment is so often framed as intuition. That mysterious feeling in the gut. Something you either have or do not have. Something that makes some people right more often than others and separates the successful from the merely competent.

But intuition does not fall from the sky.

It is built the hard way. Through work and training. Through mistakes and learning. Repetition stabilizes patterns. Subtle warning signs become easier to detect and quicker to process. A decision made from intuition is often made fast and feels right.

So what we call judgment may be nothing more than experience compressed to its essence. If that is true, then it is not mysterious at all. It is the result of decades of lessons learned. The residue of past decisions and their consequences. It is formed in uncertain situations where no answer is ever fully available.

For centuries, perhaps for millennia, every discipline offered beginners some kind of learning space in which they could try, fail, repeat what was known to work, and slowly turn learning into experience and experience into intuition.

This is a convincing story about how judgment is formed. It may not be the only one. There is another way to look at a judgment call.

Reduced to its core, the phrase tells us only that a decision had to be made without clarity about which decision was correct - if a correct option existed at all. The defining feature of a judgment call is uncertainty.

And yet we often talk about decisions as if all of them were the final step in some fully mature process. Executives appear to carry an internal model of reality. They see the situation, understand its dynamics, and therefore know which path is the right one.

In practice, it may be something far less sophisticated than that.

It may simply be the application of lessons learned from the past. Decision-makers select the option that promises the greatest chance of success. They do not need a full model of the world to do that. And even if all the data required for such a model existed, there is often not enough time to build it, test it, and rely on it.

What remains is a decision that uses past patterns to move toward the most plausible next step in an uncertain situation.

It has worked for a long time and perhaps still does. What makes the thought uncomfortable today is that modern AI systems operate in a surprisingly similar way.

Large language models do not understand the world in terms of cause and effect. They do not derive a full theory of the situation in front of them. They have learned which patterns were successful in the past, which combinations tended to fit, and which responses were more likely to be accepted as coherent, useful, or correct. All of that happens not through causal understanding, but through statistics at scale.

This is not unique to language models. Many modern AI systems operate under the same constraint: they move through uncertainty by applying learned patterns to a present case and selecting the most probable continuation.

None of this proves that machines possess judgment.

It merely makes the concept feel less settled than we might prefer.

We tend to speak about judgment as if it belonged to a different category altogether. As if it were the opposite of probabilistic guessing. As if a judgment call began where data-driven systems necessarily fail.

But perhaps that contrast was always a little too flattering.

If what we call judgment is, in many practical situations, the fast application of compressed experience under uncertainty, then it may not be as far from probabilistic inference as we would like to believe.

The difference may be real. But it may no longer be where we once thought it was.

Even now, years later, I still cannot say whether the decision my boss made back then was the right one or if it produced the best outcome. But what was never in question was his authority to make the call. When he said “judgment call,” the matter seemed closed.

Today I am less certain about the phrase than I used to be.

Not because I think judgment does not exist. But because I am no longer sure that we have been precise about what separates it from inference under uncertainty. “Judgment call” often sounds like the end of a discussion. It may increasingly be the point at which a more precise one should begin.