We cannot judge things well when our words hide important differences.
A simple word can sometimes create a complicated mistake.
We give something a name, then believe we understand it. The name helps us handle complexity. Yet it can also hide differences that matter.
This connects two ideas I have been exploring. One comes from Complexity Is Not Confusion. The other comes from Emma Boudreau’s We Should Stop Using The Term AI. Her argument about AI points toward a much larger lesson.
Before we can judge something well, we must distinguish what we are judging.
Complexity Is Not Confusion
Complexity and confusion are different problems.
A forest is complex because many living systems interact within it. An economy is complex because millions of decisions affect one another. A community is complex because people share resources, interests, rules, and relationships.
Complexity exists in the reality itself. Confusion exists in our understanding of that reality.
Clear thinking does not remove complexity. It reveals the relationships that matter. It helps us see the parts without losing sight of the whole.
That means simplification has limits.
Every simplification removes information. Good simplification removes details that do not matter. Bad simplification removes distinctions that do.
This is where language becomes important.
The Trouble With Calling Everything AI
Boudreau argues that “AI” has become a catch-all term. Very different technologies now live beneath the same label. These include language models, image classifiers, forecasting systems, and facial recognition.
The problem is not simply the word AI.
The problem begins when the word replaces examination.
Consider two applications. One system helps scientists predict protein structures. Another generates pictures from written instructions.
Both may be called AI.
Yet they have different purposes and consequences. Their risks can also differ greatly. Judging them simply as “AI” hides those distinctions.
This produces a poor question.
Is AI good or bad?
The category is too broad for the question.
Better questions quickly appear.
What system are we discussing. What does it actually do. What problem does it solve. Who benefits from it. What resources does it consume.
Then we can ask what happens when it fails.
Those distinctions give judgment something solid to work with.
Economics Has the Same Problem
Economics suffers from the same tendency.
Consider the word “market.”
We sometimes speak about markets like they are single mechanisms. Yet markets depend upon laws, property rights, contracts, information, institutions, customs, and power relationships.
Different markets can therefore behave very differently.
The same problem appears with growth.
We commonly hear that economic growth is good. But the statement hides the most important question.
What exactly is growing?
An economy might grow because useful production increased. Spending might also increase because society must repair damage. Financial activity can expand while household security weakens.
The same numerical direction can describe very different realities.
This does not make economic measurement useless. It means measurement requires interpretation.
The word “growth” cannot make the judgment for us.
Neither can words like efficiency, productivity, capital, debt, or ownership.
Each word compresses a complicated reality.
The question is whether that compression preserves what matters.
Ostrom Looked Beneath the Word Commons
Elinor Ostrom provides perhaps the strongest example.
The old argument about shared resources often offered two choices. Resources should be privately owned or centrally regulated. Otherwise, people would supposedly overuse them.
Ostrom studied what people actually did.
She examined forests, fisheries, pastures, lakes, groundwater, and other shared resources. She found communities that developed institutions for managing common resources successfully. The Nobel committee recognized this work for showing how common property could be managed by user associations.
This required an important distinction.
A commons is not simply something everybody can use however they want.
A functioning commons can have boundaries. It can have rules about access and use. Users can participate in changing those rules.
There can also be monitoring and graduated sanctions. Communities can create ways to settle conflicts. Larger systems can organize governance across several connected levels.
Calling everything “common property” could hide these differences.
Ostrom opened the category and looked inside.
That changed the judgment.
The Hidden Lesson From Ostrom
There is something deeper here.
Ostrom did not merely provide another economic answer. She changed the distinctions used to understand the problem.
The important question was no longer simply this.
Private or government?
A different question became possible.
What institutional arrangements allow people to govern shared resources successfully?
That is a much richer question.
It moves our attention from labels toward relationships. We begin looking at users, resources, boundaries, rules, incentives, monitoring, trust, conflict, and authority.
Ostrom herself warned against oversimplification. Her Nobel lecture described design principles as underlying regularities among durable systems. They were not instructions blindly followed by every successful community.
Reality remained diverse.
The principles helped us understand that diversity without erasing it.
That is intelligent simplification.
AI, Economics, and Commons Share the Same Problem
AI, economics, and commons seem like different subjects.
One concerns technology. Another concerns economic organization. The third concerns shared resources and governance.
Yet the same thinking problem appears in all three.
“AI” can hide differences between technologies.
“Growth” can hide differences between economic outcomes.
“Market” can hide differences between institutional arrangements.
“Commons” can hide differences between open access and organized governance.
The danger appears when the category becomes the conclusion.
Once that happens, we stop looking.
We begin reasoning downward from the label. AI must behave this way. Markets must behave that way. Commons must inevitably fail.
Reality becomes forced into our categories.
Good judgment works in the opposite direction.
It keeps returning to reality.
Distinction Comes Before Judgment
We can describe this as a simple thinking cycle.
Name → Observe → Distinguish → Understand → Judge → Act → Learn
Naming comes first because we need words.
But observation must follow.
We examine what actually exists. Then we identify differences that matter. Those distinctions improve our understanding.
Only then are we ready to judge.
Action eventually tests that judgment against reality. Results give us feedback. We learn, and our distinctions can improve again.
This makes language corrigible.
Our words remain tools rather than prisons.
We Still Need Simple Words
None of this means broad categories are bad.
We need them.
Imagine discussing technology without saying AI. Imagine economics without words like markets or capital. Imagine discussing shared resources without using the word commons.
Language would become exhausting.
The goal is not maximum detail.
The goal is sufficient distinction.
A useful map leaves many things out. Otherwise, it would become as complicated as the territory itself.
But imagine a road map that removed bridges.
The map would certainly become simpler.
It would also become less useful.
Good language works the same way.
Simplify what can be simplified. Preserve what must be distinguished.
A Better Test for Our Words
This gives us a practical test.
Whenever an important word appears, ask what differences it might be hiding.
When someone says AI, ask which system.
When someone says growth, ask what grew.
When someone says market, ask which rules shape it.
When someone says commons, ask how people govern it.
When someone says ownership, ask what rights ownership provides.
These questions seem small.
Yet each one forces us beneath the label.
And once we see the differences, our judgment may change.
Good Judgment Begins With Better Distinctions
Complexity cannot be defeated by giving complicated things simple names.
Sometimes a simple name helps us understand.
Sometimes it merely makes our confusion easier to say.
Boudreau’s criticism of the term AI shows this problem clearly. One label now covers technologies with very different purposes and consequences.
Economics shows the same problem. Words like markets and growth can become containers for very different realities.
Ostrom shows us another path.
She did not accept the broad category as the final explanation. She studied how real people governed real resources. Those observations revealed distinctions that simpler theories had missed.
That may be the larger lesson.
We need simple language because reality is complex.
But simplicity should help us see.
It should never require us to become blind.
Good judgment begins with better distinctions because we cannot judge what we have failed to distinguish.
Credits
Inspired by Emma Boudreau’s We Should Stop Using The Term AI, published in chifi on Medium, June 8, 2026. Read the original article on Medium
This essay also draws from Elinor Ostrom’s work on economic governance and the commons. Her research challenged simple assumptions about shared-resource governance. Elinor Ostrom’s Nobel Prize Lecture
Developed alongside the ideas explored in Complexity Is Not Confusion.
Tags
#Systems_Thinking #Artificial_Intelligence #Economics #Commons #Critical_Thinking
No comments:
Post a Comment