Why Strange Thinking Beats Predictable AI Writing Every Time

The Odd Over the Obvious

Allowing A.I. to handle our writing diminishes our ability to engage in the kind of complex thinking that defines human creativity. This issue couldn’t be more urgent.

Recently, many educators have been alarmed by how often students rely on A.I. to write their papers. Some even wonder, why not? But history and literature offer a clear answer.

Choosing the Strange Over the Predictable

Seventeenth-century scholars followed a principle called lectio difficilior potior: when faced with differing versions of a text, the more unusual reading is often the original. This is because copyists and editors tend to replace odd or challenging words with more familiar ones. The lesson? Don’t let the easy choice erase the unique, the complex, or the thought-provoking.

Take Shakespeare’s Romeo and Juliet. After Romeo kills Juliet’s cousin Tybalt, a speech that appears in all early editions is given to Juliet’s father, Lord Capulet, who surprisingly argues for Romeo’s forgiveness. Editors and directors, however, have long reassigned this speech to Romeo’s father, Lord Montague, because that makes more sense at first glance. But this predictable choice erases a richer possibility: that Juliet’s father might have accepted Romeo if they had communicated. This subtle complexity is lost when we choose what seems obvious over what challenges us.

When we settle for the most probable outcome, we ignore a multiverse of better possibilities. This narrowing of vision threatens our ability to imagine alternatives in politics, economics, and the environment.

A.I. Threatens Human Complexity

Text-generating A.I. programs operate by predicting the most likely next word based on vast amounts of existing text. They favor the familiar and the average, the safe and the common. Over time, as these models train on A.I.-generated content itself, this tendency compounds.

What A.I. produces is repetition, not originality. It can’t invent new words or ideas the way Shakespeare did. Increasing unpredictability in A.I. models only adds randomness, not meaningful complexity. This leads to confident but false statements—errors that, like the tragic misunderstanding in Romeo and Juliet, have real consequences.

Writing for Readers

Recent shifts in early education away from phonics toward guesswork mirror how A.I. predicts text. Instead of decoding, students learned to anticipate words based on context. This approach has been widely criticized for weakening reading skills and diminishing the joy of discovery.

Despite academic trends that downplay authorial intent, readers continue to seek connection with unique human voices. We want to understand what a writer truly means and feels. This desire fuels our rapid brain development and social nature.

It’s not just about catching plagiarism. Students recently protested having their names read by A.I. at graduation, valuing the human presence and meaning behind those moments. Whether reading or speaking, we crave intention, not hollow repetition.

Many A.I.-generated papers share a dull, uniform style that lacks personality. Predictable writing can never replace the distinct voices of poets like Percy Bysshe Shelley or storytellers like Robin Wall Kimmerer. Writers should embrace the odd and unexpected; that’s where real value lies.

Writing for Freedom

True liberty depends on unpredictability. As Timothy Snyder highlights in his book On Freedom, freedom requires creating futures that tyrants and machines cannot foresee or control. Accepting the “normal” means surrendering those futures.

A.I. writing enforces the status quo. It produces content that echoes existing ideas, making it a tool of conformity rather than creativity. Worse, when prompted to argue a certain point, A.I. can produce polished but unexamined rhetoric, feeding confirmation bias and deepening political divides.

Social media’s algorithms amplify this effect by showing us more of what provokes fear or anger, pushing us toward predictability and away from genuine human complexity.

Both left-wing and right-wing speech codes limit language diversity, leading to euphemisms and blandness that signal rising control, as George Orwell warned. The antidote is to explore less-filtered histories and embrace uncertainty. John Keats called this "Negative Capability"—the ability to sit comfortably with doubt and mystery, a trait Shakespeare mastered.

Developing the skill to read unexpected voices and discern meaning is essential. Relying on A.I. to write or argue short-circuits this growth.

Human Minds Need Diversity

New research shows that working with large language models makes people’s opinions more conventional and less innovative. Diversity in thought is not a buzzword but a survival factor, as epidemiologists note for biological populations.

Human genetic diversity has already shrunk drastically, threatening our species’ long-term survival. Our mental diversity matters just as much.

A.I. encourages sameness, dulling curiosity and resistance to harmful trends. Reading and writing, as social immune systems, foster individuality and connection across time and space.

Like true lovers separated by distance, writers must keep communicating from the heart and mind. The quick, technical fixes that A.I. offers often lead to tragic misunderstandings—a lesson we would do well to remember.

For writers interested in understanding how to balance A.I. tools with maintaining creative integrity, exploring thoughtfully designed courses can help. Find practical guidance on Complete AI Training’s resources for writers.