← All Series

Complete Series · 7 essays · 2026

On Judgment

What do we actually mean when we say someone has good judgment?

I kept using the word judgment across product, data, AI, strategy and leadership.

Eventually I realized I was asking it to describe too many different things.

Reasoning. Evaluation. Taste. Decision-making. Authority. Accountability. The ability to interpret incomplete evidence. Even some of what I thought would remain distinctively human as AI became more capable.

So I decided to take the word apart.

On Judgment is a seven-part inquiry into what judgment actually is, what makes an assessment reliable, how people develop that capability, and what happens when organizations and machines connect assessments to consequential actions.

The inquiry did not end where I expected. I started with the idea that organizations scale judgment by encoding it in products, rules, metrics and software. I ended with a narrower claim: organizations distribute assessments, standards, authority and feedback across decision systems. The question is whether those systems remain valid, challengeable and correctable.

This was not a finished framework presented one piece at a time. The argument changed as the questions became sharper.

Start with Part 1 ↓

The complete inquiry

Seven essays, from defining judgment to designing correctable decision systems

Part 1Definition

I Keep Talking About Judgment. What Do I Actually Mean?

I had argued repeatedly that organizations need to preserve, improve and scale judgment. Before making that claim any further, I needed to understand what I was actually talking about.

The first essay traces how judgment entered my thinking through product decisions, data, AI and the Semantic Supply Chain, then starts separating assessment from reasoning, preference, decision, authority and accountability. It ends with a working definition rather than a finished answer.

Read Part 1 →
Part 2Boundaries

Judgment Is Smaller Than I Thought

The counterexamples made the definition thinner.

This essay treats judgment operationally as an assessment about something, then asks what follows once we stop using the word as a container for every valuable part of a decision. Judgment is separated from reasoning, preference, generation, decision and authority.

The practical question becomes much more precise: What exactly is the assessment inside this decision?

Read Part 2 →
Part 3Outcomes

Was It Good Judgment, or Did It Just Work?

Once judgment is separated from decision, another problem appears: how do we know whether an assessment was actually good?

A successful outcome does not prove that the original judgment was sound. A bad outcome does not necessarily prove that it was poor. The essay treats outcome as evidence about an earlier assessment rather than the verdict on it, and asks what a decision system needs in order to remain correctable.

Read Part 3 →
Part 4Formation

Where Does Good Judgment Come From?

Experience does not automatically create expertise.

This essay looks at informative experience, expert intuition, domain transfer and the organizational conditions that shape judgment: what evidence people see, whose expertise counts, whether disagreement survives long enough to surface unique information, which standards define “good”, and how consequences are interpreted.

It ends by moving the problem from the individual to the organization: What happens when a company tries to make an assessment or standard durable after the person who formed it is gone?

Read Part 4 →
Part 5Scaling

What Happens When a Company Tries to Scale Judgment?

The inquiry moves from how individual judgment develops to what happens when assessments and standards become durable inside organizations.

But not every rule, metric or process encodes someone's judgment, and making a decision durable is not necessarily a loss of context. The more useful questions concern operational force: what the system preserves, what it can cause, where it remains valid, and how it can be challenged or updated.

Read Part 5 →
Part 6Human role

If AI Can Judge Better Than Us, What Should Humans Still Do?

If AI can form better assessments than humans in a defined class of cases, keeping a person in the loop does not automatically improve the decision or provide meaningful control.

This essay separates assessment capability from authority and accountability. It examines which human roles remain useful, what effective oversight requires, and how automation changes the experiences through which people develop the capabilities they may still need.

Read Part 6 →
Part 7Diagnostic

What Remains of Judgment?

The final essay returns to the original claim that organizations scale judgment by encoding it in rules, metrics and software. That explanation no longer survives intact.

What remains is a narrower view of judgment as assessment, and a more practical focus on consequential decision systems. The essay proposes the decision class as the unit of analysis and consolidates the inquiry into four diagnostic questions: what assessment matters, what it can cause, who can challenge it, and what evidence should change it.

The result is not a comprehensive or validated new framework, but a smaller diagnostic whose value still needs to be tested against existing approaches.

Read Part 7 →

From scaling judgment to correctable decision systems

I started this series with the idea that organizations needed to preserve and scale judgment. Seven essays later, I no longer think that is quite the right problem.

Judgment, in the narrower sense I now use, is the act of forming an assessment. But an assessment is not a decision, and neither is the same as having authority to act. Organizations can scale particular assessment capabilities, make standards durable and connect them to action. Those are different operations, with different implications.

The more useful question is what happens when an assessment becomes consequential: what standard it serves, where it remains valid, what it can cause, who can challenge it and what evidence should change it.

That leaves a smaller practical diagnostic, not a universal theory of judgment. Whether it adds value beyond existing approaches still needs to be tested.

The claim I would now defend is narrower than the one I started with: make the assessments that shape consequential action visible, challengeable and correctable.