Attribution Failure Is the Most
Expensive Misdiagnosis.

Your organization judges AI on reasoning and a leader on confidence

Your organization already pays for more judgment than it hears. Your leadership team is reading this quarter’s competitor analysis. One of the analysts you trust most drafted it with AI. Everyone at the table knows AI was involved, and that is the reservation. AI sounds sure of itself and never tells you what it does not know. So the team sends the analysis back for its sources before acting on it, which is good discipline. At the same time the team forms a read of the analyst, and the trust the analyst has earned does not stop it. The word for that read around the table is offloading. In that read the analyst handed the judgment to AI and kept the execution. That read has the division backward. The analyst handed AI the execution and kept the judgment, and can tell you which claims they checked, which they threw out, and why they are comfortable with the final output. Scaffolding the work that way is the exact productivity increase you want from AI use. Nobody at the table asks the analyst for the reasoning they are holding ready.

Later that week the same team takes a strategic direction call. The leader is sure of the move and says nothing about what it rests on. One person asks, and the answer is “trust me, I have been around the block once or twice,” and the team accepts it. The title and the years carry the call. The team hears the question as a challenge when it was the team doing its job. The team does not send the call back for its reasoning the way it sent the analysis back for its sources. Every leadership team does this. Acting on a leader’s call is how work gets done, and a leader’s confidence passed for judgment for as long as the leader knew more than anyone else in the room. As I argued in my essay What Leading Becomes, that stopped being true once the people below held more of the information. The habit outlived its reason. Notice what the team did with the call. It took an answer as written and acted on it, the exact thing it accused the analyst of doing with AI. On the leader’s call nobody uses the word.

The subject of both scenes is your team’s skill, not either person. On my framework page I named this failure mode attribution failure: communication breakdown diagnosed as individual deficit rather than interface failure, the most common and most expensive misdiagnosis in organizational life. In both scenes the misdiagnosis is the team’s, down onto the analyst’s work, where it read a deficit, and up onto the leader’s call, where it read judgment, and the reasoning is what got lost. Telling confidence from reasoning is a skill, and your team used it on the analysis without anyone having to prompt it. One of the oldest norms in leadership stopped it on the call. Under that norm a leader’s confidence counts as judgment, and asking what it rests on counts as a challenge. That norm keeps the team from using the skill your organization now needs at every level. The people at your table reason differently from one another, and that variation is now priced as capital. AI has made confidence cheap, so reasoning is what your team has left to go on. Ask people what their answer rests on, the way you already ask AI, before you decide how far to trust it. Where your organization aims the question of what an answer rests on is a design decision about your organization in The Age of AI, at the intersection of The Neurodiversity Movement, which showed us cognitive variation is a good thing, and The Brain Economy, which priced it.

The Operating Brief: ask for the reasoning before the read goes into the record

Reasoning is what judgment gets calibrated on now

A leadership team decides better when it hears what a recommendation rests on. The same team receives a director’s finished recommendation to exit a product line and decides how far to act on it. It decides on two things: how sure the director sounds, and how the director’s earlier calls went. I call that decision judgment calibration. It is how an organization decides whose judgment to act on, and how far, before the result is in. Every leadership team runs it, usually without a name for it.

Two things changed underneath it. Judgment is now the working skill a role is measured by at every level, so your team is deciding how far to trust many more judgments, most of them from people with no track record yet. And AI now produces a fluent, confident answer to any question in seconds, so confidence no longer separates one judgment from another. Reasoning is what is left, and it is in the room in the hour the recommendation is heard. It is the only part of a judgment another person can inspect and add to before the outcome.

Your people reason differently from one another. Different nervous systems produce different cognitive architectures, shaping thinking, sensing, and processing across mind and body, not broken brains. At your table that shows up in how people reason to a conclusion and how they put it into words. The Brain Economy priced that variation. Brain capital is brain health plus brain skills, quantified as infrastructure with a measurable return, and the reasoning your people produce is the skill half of that capital. Hearing it is how the capital gets used. The Age of AI makes reasoning the core input to judgment calibration, because confidence now comes from AI as well as from the people in the room. Together those three facts make judgment the skill your organization runs on at every level, and hearing reasoning is how a team builds it.

A team that hears reasoning protects you, because the decision rests on judgment the people in it can understand and add to. A team that listens for confidence cannot protect you on the day your delivery is good and your reasoning is wrong.

Many communication misunderstandings are reasoning differences

When two people at your table disagree, the team often has two reasonings to work with. Take the misunderstandings your organization filed under communication this year, and the sentences written afterward in good faith about how one of the people communicates. Many of those sentences came from a reasoning difference. Two directors have the same pilot pack, and the meeting has to decide whether to expand the pilot to every region now or hold it a quarter. The first director reasons from the expansion plan. The pilot met its targets and the budget is approved, so expand now. The second reasons from what has already happened. Support tickets in the pilot region doubled last month, and the last two expansions slipped after numbers like these, so hold a quarter. Each conclusion holds on the evidence its director weighted, and neither says which. The team hears a disagreement and then it reads the people. The second director gets taken as reluctant about the plan, and the meeting closes on the first with “align offline” as the action item beside the two names. Every leadership team has written that action item, and it is the right call when the difference between two people is a working relationship. Here it was two reasonings, and the team heard neither. Two reasonings on one question is more judgment than the room had when it heard only the two conclusions.

Many of the other sentences came from how the reasoning was said. I call that an articulation difference. The same reasoning can be said conclusion first or evidence first, plain or with qualifiers, out loud or in writing. Said two ways, it gets heard as two people, because the listeners take the way it was said as a sign of who the person is and write that down.

Anyone who has led people has read a delivery and written the read down as a finding about the person, and anyone who communicates differently has been read that way. From the receiving end, what I could see was the channel changing the read. The same reasoning a room had heard as hesitation when I gave it aloud went through without comment when I sent it in writing. Many neurodivergent professionals can show you the pair, the sentence a room agreed about them and the written document from the same months that nobody flagged.

One forecast error, caught and reported twice, shows that read made by a constructed room and then by AI.

Check the reply the way your team checks any AI-drafted work, and ask what its “ready now” rests on. By the reply’s own account it rests on how the message was said, an answer you would send back. A room with no AI in it split the same two authors the same way, and nobody called that offloading. When the sentence about a person changes with the channel, the room or the manager, and the person has not changed, it came from how they said it. Have your team hear the reasoning under what was said: what people observed, what they concluded, and what they recommended. When the difference is in the reasoning, what each person weighed has to be said out loud. When the difference is in how it was said, the reasoning was there and the team missed it. Either way the skill is the team’s, the same one it already uses on AI-drafted work.

Hearing and understanding reasoning becomes the organization’s skill

Your organization spends real money teaching people to be heard. Open a review that lists communication skills as a development goal, and the workshop invitation behind it. The training is worth what you spend on it, and nearly all of it works on the speaker. The hearing side has no budget line and no box on the review. Everyone assumes that once a reasoning is said well, the team has heard it.

The nine interventions of my Cognitive Translation Protocol, in three lanes for my six failure dimensions, are simple by design. They are changes an organization can make this quarter to documents and meetings it already keeps. The interpretation lane has the team write what it observed before what it concluded. The process lane holds a place for reasoning that finishes after the meeting and keeps it with the decision. The environment lane gives reasoning a channel and a scaffold that do not depend on the pace of the meeting or the room. Together the three lanes begin to open the floor for reasoning to be what the organization listens for and truly hears going forward. Once the lanes are in place, the team still has to ask.

Before the team agrees a sentence about a person, it asks what the sentence rests on and what reasoning it heard. I call that the reasoning question. It runs in both directions, on the confident call and the “ready now” as much as on the hedged flag. In practice it is one question the person running the calibration meeting asks before a sentence is agreed, a line ahead of the sentence a manager writes, and a line in the meeting norm your leadership team already keeps.

In The Age of AI judgment is needed at every level, and that need flips who the onus lies on, from the speaker to the room. Hearing and understanding reasoning becomes the skill. The people in your rooms have earned more of that hearing than AI’s answers get, never less, because a person’s reasoning develops when it is asked for. Sometimes the failure is in the person: the reasoning does not hold once it is asked for and heard, and then the coaching aims at the reasoning and the team can say why. Rebuilding the assessment of people on the reasoning question comes later in this arc.

Building the habit takes one question and two lines on documents you already keep, and running it takes minutes per case. The first cycles take longer, because leaders whose calls have been taken on confidence for years get asked by their own teams what a call rests on, and some may hear it as a challenge. After that the team asks as a matter of course, and nobody who asks is read as difficult. A hedge or a written answer stops being written down as a finding about the readiness of the person who gave it.

Why this works, and where it comes from

Your organization may already ask this question in two other places. Quality management, in the form W. Edwards Deming gave it, asks whether a result belongs to the system before attributing it to the person who produced it. Leadership programs that follow Argyris and Senge teach a leader to state their own reasoning and ask for the reasoning of others before concluding anything about them. Where both exist, the calibration meeting and the leadership meeting have borrowed neither.

Anderson and colleagues found that status goes to overconfidence over and above ability, and Tetlock found confidence a poor guide to accuracy. So a team that goes by confidence rewards overconfidence and leaves the reasoning unchecked. Ross found that people explain what someone did by who they are before they look at the situation, so asking has to be a deliberate step. Deming’s first question is the system, and my Cognitive Translation Protocol puts that system at the interface between two cognitive architectures. Argyris and Senge taught the hearing as a personal discipline, and this essay makes it the team’s skill, with a set moment to run it. Ask people what their answer rests on, the way you already ask AI, before you decide how far to trust it.