From Writing to Reading: Why AI Makes Critical Thinking More Important Than Ever

To Chris

Thesis

The arrival of AI has not made writing obsolete. It has changed what education should value. If drafting becomes inexpensive, then the central academic skill is no longer producing a first draft but critically reading, questioning, revising, and improving one. In this sense, philosophers such as Derrida become suddenly timely, not because students must adopt their conclusions, but because they exemplify habits of reading that the age of AI demands.

The Wrong Question

Since the arrival of powerful AI systems, schools and universities have found themselves asking a familiar question: Should students be allowed to use artificial intelligence to write essays? Some teachers advocate banning these tools altogether, arguing that they encourage plagiarism, weaken critical thinking, or reproduce the assumptions found in the vast collections of human texts on which they were trained.

These concerns are understandable, but they may also be asking the wrong question.

Whether we approve of AI or not, it has already become part of the educational landscape. Students can now generate a competent first draft on almost any topic within seconds. Like the calculator, the search engine, or the spell checker before it, AI is unlikely to disappear simply because educators wish it would.

The real challenge, then, is not deciding whether students should use AI. It is deciding what kind of education the age of AI requires.

For centuries, education devoted enormous attention to teaching students how to produce texts. Today, however, the production of a first draft has become easier than ever. If drafting is no longer the scarce resource, perhaps education should shift its emphasis toward a different skill: the ability to evaluate, question, revise, and improve what has been written.

This change does not diminish the importance of writing. On the contrary, it transforms writing into a more reflective activity. The first draft may increasingly be produced with technological assistance, but understanding still depends upon human judgment.

The central educational question is therefore no longer Can AI write? It is Can students read critically enough to know what deserves to be kept, questioned, revised, or rejected?

Every Text Deserves Close Reading

Public discussions about AI often revolve around one concern: its biases. Because AI systems learn from texts written by human beings, they inevitably reproduce many of the assumptions, omissions, and perspectives found in those texts.

This observation is correct, but it is also incomplete.

The real issue is not that AI-generated texts contain assumptions. All texts do.

A philosophical dialogue by Plato contains assumptions. A scientific paper contains assumptions. A newspaper editorial contains assumptions. Wikipedia contains assumptions. ChatGPT contains assumptions. Even Derrida's own writings contain assumptions.

Some assumptions are explicit. Others remain hidden. Some prove fruitful and open new ways of thinking. Others deserve to be questioned or rejected.

The task of education, therefore, is not to divide texts into two camps—those written by humans and those written by machines. The task is to cultivate readers capable of examining any text critically, regardless of its origin.

This is where philosophy becomes unexpectedly relevant to the age of AI. Philosophers have spent centuries developing methods for reading beyond the surface of a text. They ask not only what an author says but also what the argument quietly presupposes, what distinctions it relies upon, and what remains unquestioned.

The educational challenge of the twenty-first century may therefore be surprisingly traditional. Students do not merely need more information. They need better habits of reading.

Reading Like Derrida #1

Every text deserves the same question: What assumptions make this argument possible?

The New Literacy

The arrival of AI suggests that we may be witnessing the emergence of a new kind of literacy.

For centuries, literacy largely meant two complementary skills:

Reading

↓

Writing

Students read books, articles, and essays in order to learn how to produce texts of their own.

Today, however, that sequence is changing.

Increasingly, literacy looks more like this:

Reading

↓

Editing

↓

Rewriting

↓

Questioning

↓

Reading again

The difference is profound.

When producing a first draft required considerable time and effort, writing naturally occupied the center of education. But if AI can generate a coherent draft in a matter of seconds, then the truly valuable skill is no longer producing language. It is exercising judgment.

Students must decide whether an argument is convincing, whether a definition is adequate, whether important evidence has been omitted, whether key concepts remain ambiguous, or whether hidden assumptions have shaped the conclusion.

In other words, the scarce resource is no longer information.

Nor is it language.

The scarce resource is judgment.

Ironically, this shift may elevate rather than diminish academic standards. Instead of evaluating only a student's ability to produce text, teachers can increasingly evaluate higher-order intellectual abilities: conceptual precision, analytical discrimination, revision, interpretation, and intellectual independence.

Writing does not disappear.

It becomes an act of thoughtful editing.

Reading Like Derrida #2

Slow down. Before asking whether a text is correct, ask how it is constructed.

What Derrida Teaches AI Users

At first glance, Jacques Derrida may seem like an unlikely guide for the age of artificial intelligence. His writings long predate ChatGPT, and he never imagined a world in which machines could produce fluent essays on demand.

Yet the habits of reading that Derrida exemplifies have become remarkably contemporary.

Consider a typical interaction with an AI system. A student asks for an explanation of Roman Jakobson's theory of translation. Within seconds, the AI produces a clear and competent summary. It correctly explains the distinction between intralingual, interlingual, and intersemiotic translation. For most readers, that would be the end of the matter.

Derrida, however, reads differently.

He slows down.

He notices that Jakobson explains intralingual translation by calling it rewording. He explains intersemiotic translation by calling it transmutation. But when he reaches interlingual translation, he simply calls it translation proper.

Why?

Why does this category apparently require no explanation?

Why is the adjective proper introduced without argument?

Those questions transform an ordinary reading into a philosophical one.

Derrida is not looking for mistakes. Nor is he trying to prove that Jakobson is wrong. He is paying attention to what most readers overlook. A single adjective quietly reveals assumptions about language, meaning, and translation that the text itself leaves unexamined.

This is precisely the habit of mind that students increasingly need when reading AI-generated texts.

An AI may produce an accurate summary.

It may even produce an excellent one.

But students should still learn to ask:

What distinctions does this explanation take for granted?

What concepts remain undefined?

What alternatives have been excluded?

What assumptions quietly organize the argument?

These questions are not directed against AI.

They are directed toward every text worthy of careful reading.

That is why Derrida remains precisely useful today. His greatest lesson is not a philosophical doctrine but an intellectual practice: slow down, notice asymmetries, question definitions, examine hierarchies, and ask why one word appears instead of another.

These are not merely the skills of a literary critic. They are becoming the essential habits of an educated reader in the age of artificial intelligence.

Reading Like Derrida #3

Notice asymmetries before looking for errors.

Ask why one word was chosen instead of another.

Small textual details often reveal the deepest assumptions.

The Teacher's Role Changes

If the central academic skill is shifting from producing texts to critically evaluating them, then the teacher's role must also evolve.

For generations, the typical sequence of writing instruction looked something like this:

Student writes

↓

Teacher evaluates

The teacher's task was to assess the quality of a text that the student had produced independently.

AI changes this sequence without eliminating the teacher's role. If students increasingly use intelligent tools to generate a first draft, the educational process may instead look like this:

AI produces a first draft

↓

Student evaluates

↓

Student revises

↓

Teacher evaluates the student's editorial judgment

Notice what has changed.

The object of evaluation is no longer simply the finished essay. It is the student's ability to improve an imperfect text.

Did the student recognize weak arguments?

Did they identify vague definitions?

Did they detect unsupported claims?

Did they strengthen the evidence?

Did they question hidden assumptions?

These are not secondary skills. They are among the highest intellectual abilities that education seeks to cultivate.

Seen in this light, AI does not necessarily reduce academic standards. It may raise them. Producing a competent first draft has become easier, but exercising sound judgment remains difficult. Teachers can therefore devote less attention to measuring the ability to produce language and more attention to evaluating conceptual precision, analytical discrimination, intellectual independence, and the quality of revision.

The teacher becomes less concerned with asking, "Did you write every sentence yourself?" and more interested in asking, "Can you explain why this version is better than the previous one?"

Reading Like Derrida #4

Editing is not a mechanical correction of mistakes.

It is a form of thinking.

A Classroom Exercise

How might these ideas be translated into classroom practice?

One possibility is to use AI not as a substitute for learning but as the starting point for it.

Imagine a lesson on Roman Jakobson's theory of translation.

In the first stage, students ask an AI system to explain Jakobson's distinction between intralingual, interlingual, and intersemiotic translation. The result will probably be accurate, coherent, and well organized. At this point, there is no discussion. Students simply read the explanation.

In the second stage, they are given Derrida's discussion of Des Tours de Babel. Their task is not to decide whether Derrida is right or wrong. Instead, they ask a different question:

What did the AI overlook?

Did it notice Jakobson's expression translation proper?

Did it ask why only one category appears to explain itself?

Did it examine the philosophical work performed by the adjective proper?

Did it question the assumptions that organize Jakobson's taxonomy?

Most likely, the AI produced a competent summary without pursuing these questions.

Finally, students return to the AI-generated explanation and revise it. They do not merely make it longer or more elegant. They make it more careful. They identify missing assumptions, clarify distinctions, and incorporate insights gained through close reading.

The assignment has subtly changed.

Students are no longer asked simply to write an essay.

They are asked to improve one.

This seemingly small change transforms the purpose of the exercise. Writing becomes inseparable from reading, and revision becomes an act of interpretation rather than mere correction.

Reading Like Derrida #5

Do not ask only whether a text is correct.

Ask how it could become more accurate, more precise, and more thoughtful.

Conclusion

Artificial intelligence has undoubtedly changed the economics of writing. Producing a coherent first draft is now faster and easier than ever before. What has not changed is the nature of understanding.

Understanding still requires attention. It still requires judgment. It still requires the ability to distinguish between what a text explicitly says and what it quietly presupposes.

For this reason, the central educational challenge is not deciding whether students should use AI. Nor is it protecting them from machine-generated texts. The challenge is cultivating readers capable of examining any text—whether written by Plato, Jakobson, Derrida, a classmate, or ChatGPT—with the same habits of careful scrutiny.

In this respect, philosophers such as Derrida have acquired unexpected importance. Their value lies not primarily in the conclusions they reach but in the way they read. They slow down. They notice asymmetries. They question what appears self-evident. They uncover assumptions hidden in ordinary words.

These habits are no longer merely philosophical virtues.

They are becoming essential educational skills.

The age of AI has not diminished the importance of critical reading.

It has moved it to the very center of education.

Related Post

Reading Like Derrida: Jakobson, Babel, and the Meaning of "Translation Proper"

https://derridaforlinguists.blogspot.com/2026/07/blog-post_09.html

References

Adler, M. J., & Van Doren, C. (1972). How to Read a Book (Rev. ed.). Simon & Schuster.

Derrida, J. (1985). Des tours de Babel. In J. F. Graham (Ed. & Trans.), Difference in Translation (pp. 165–207). Cornell University Press.

Elder, L., & Paul, R. (2020). The Thinker's Guide to Analytic Thinking: How to Take Thinking Apart and What to Look for When You Do (2nd ed.). Foundation for Critical Thinking.

Fisher, A. (2011). Critical Thinking: An Introduction (2nd ed.). Cambridge University Press.

Jakobson, R. (1959). On linguistic aspects of translation. In R. A. Brower (Ed.), On Translation (pp. 232–239). Harvard University Press.

Peirce, C. S. (1998). The Essential Peirce: Selected Philosophical Writings (Vol. 2, 1893–1913). Indiana University Press.

Richards, I. A. (1929). Practical Criticism: A Study of Literary Judgment. Routledge & Kegan Paul.

Russell, B. (1940). An Inquiry into Meaning and Truth. George Allen & Unwin.

Saussure, F. de. (2011). Course in General Linguistics (W. Baskin, Trans.). Columbia University Press. (Original work published 1916)

Thomas, F. (2011). How to Read Like a Professor: A Lively and Entertaining Guide to Reading Between the Lines (Revised ed.). HarperCollins.

Wineburg, S. (2021). Why Learn History (When It's Already on Your Phone). University of Chicago Press.