Thinking Without Rushing to the End
In engineering culture, there is a long-standing joke about the "holy war" surrounding text editors: Vim versus Emacs. It is often spoken of as a relic of the late 20th century, easily dismissed as a story from a time when personal computers were smaller, more intimate, and somehow more ideological. Much of the joking reduces the intensity to issues of muscle memory or keyboard shortcuts, as if that passion could be explained by mere habit.
However, the very fact that this conflict has survived for so long points to something deeper. These editors are designed to center on plain text, possess minimal interfaces, and deliberately reject visual comfort. In theory, they should invite calm. Yet, in practice, they elicit strong emotions. What appears to be a battle over software might actually be better understood as a battle over how thought itself should move.
There is a paradox here that we must first address: the closer a tool gets to the raw act of thinking through language, the more fiercely people will defend it. Simplicity does not make one let go of attachment; rather, it can sometimes make it even stronger.
Plain text as an ideal, a reality that is not neutral
Plain text occupies a special place in modern intellectual culture. It is often spoken of not just as a format, but as a stance. Plain text is durable, portable, and resistant to obsolescence. It carries the promise of keeping a distance from specific platforms or vendors. It is also said that choosing plain text means choosing continuity over novelty.
That promise is indeed real. Text files written decades ago can still be read without issue today. They can calmly survive migration processes that would cause more elaborate formats to be lost. In that sense, plain text feels like a small safety device against oblivion.
However, the ideal changes its appearance little by little once it is lived as a daily practice. As soon as plain text as a principle becomes a daily environment, habits, customs, and dependencies are born. While Org mode files are technically text, the thoughts supported there are shaped by elements like folding, agenda views, capture flows, and internal syntax—elements that truly come to life only within Emacs. Vim users emphasize neutrality, but the speed of their thinking cannot be separated from modal editing and the memory ingrained in their bodies.
The file itself is neutral. But cognition is not.
This is not a failure of plain text. It simply makes us realize that neutrality at the level of storage does not mean neutrality at the level of experience. Tools shape the direction of our attention, and that attention eventually shapes the way we think.
Open surfaces and protecting the state of being unfinished
Rather than the distinction between simple and complex, the distinction between "open and closed" is actually far more useful. It is because of this difference that plain text continues to attract thinkers even now, when richer tools are appearing one after another.
An open surface does not demand a conclusion. It accepts remaining unfinished. Pen and paper remain attractive because they do not interrogate where a thought is headed. A sentence can trail off. Notes in the margins can contradict the main text. Even words crossed out with a line continue to exist there.
Plain text editors preserve this quality to a significant degree. They do not impose correctness. They allow repetition, hesitation, and contradiction to exist simultaneously. Thoughts that have not yet taken shape can remain in that unformed state without being penalized.
This tolerance is more important than it looks. Thoughts rarely begin in a clear form from the start. They begin with fragments and intuitions that do not yet know their own shape. A surface that demands consistency too early carries the risk of distorting the shape of a thought before it has had time to settle.
From holy war to default
For a while, it seemed as if these conflicts had already become a thing of the past. Editors like Atom, and later VS Code, appeared with more accessible interfaces, well-organized default settings, and extensible ecosystems. The center of gravity shifted from ideology to productivity. Many engineers stopped "choosing" an editor altogether and simply accepted what was provided.
It felt like a resolution, but in reality, it was closer to "normalization." The complexity did not disappear; it was merely hidden behind extensions and customs. VS Code succeeded not because it was neutral, but because it allowed users to avoid the need for philosophical commitment. It allowed them to think without having to think about their tools.
That quietude was important. It created a generation that did not carry an editor as an identity. But at the same time, it also set the stage for the next disruption to occur.
Closed tools and the pressure to "finish"
Resistance to tools like Word, Excel, and PowerPoint is often misunderstood as elitism or nostalgia. But in reality, it has almost nothing to do with the issue of quality. These tools are extremely effective for the purposes for which they were designed.
The problem lies in the premises. Word assumes a finished document. PowerPoint assumes a message with key points extracted. Excel assumes a model that produces interpretable output. These premises are not wrong, but they belong to the "latter half" of the thinking process.
If you use them in the early stages, pressure to reach a conclusion creeps in. You have to choose a font. You have to align the layout. You have to put valid values in the cells. Slides must be understandable to others. Attention shifts from the meaning itself to the presentation. Judgments about appearance end up taking the helm of your thinking before you know it.
As a result, many people intuitively separate their work. They secure a separate space for thinking and translate it into Office formats later. This is not inefficient. It is an intuitive defense mechanism to protect the space for cognition.
Markdown and the layering of simplicity
Markdown was supposed to be the solution to this tension. Readability that doesn't depend on tools, and a structure that doesn't become rigid. It promised to offer both at the same time, providing a middle path between raw text and heavy decoration.
However, almost simultaneously, various differences emerged. iA Writer pushed the tranquility of focus and typography to the forefront. Bear introduced a philosophy of organization centered on tagging. Typora tried to make the formatting itself disappear from consciousness through live preview. While each tool remained technically simple, each clearly expressed a different philosophy of writing.
Markdown itself remained minimal. Human preferences did not.
This recurring pattern reveals something important. Simplicity at the level of form does not erase complexity at the level of experience. It only shifts where that complexity resides.
PKM that reflects values
Personal Knowledge Management (PKM) systems have made these tensions visible once again. Tools like Obsidian and Notion are not just note-taking apps. Each reflects different premises about what thinking is.
Obsidian emphasizes files, links, and local ownership. Even when layers like plugins and graphs are added, the lineage of plain text is deeply inherited at its foundation. On the other hand, Notion is based on structure, databases, and collaboration. It prioritizes readability and ease of adjustment over raw flexibility.
Both can be used thoughtfully. And both can become rigid. The debate surrounding these two is not about features. It is a question of where thinking should live, and when it should become something "to be shown."
Here, too, a foundation that should be neutral becomes a place where meaning resides.
Dialogue returns in unexpected ways
Against this backdrop, the emergence of AI-powered editors like Cursor, Antigravity, and Zed raises the question again at a deeper level. These are not just editors with more features. They shake the very premises of what a writer is and where agency lies.
When AI appears not as a mechanism for optimization but as a conversation partner, an "open state" is unexpectedly restored. Half-formed thoughts are welcomed, errors are tolerated, and rewriting is accepted as a premise. Meaning emerges not from execution, but from the exchange.
The reason this feels natural is that thinking has always worked that way in conversation. In this sense, a truly conversational AI behaves more like a notebook than Word. Language remains the interface, and conclusions are not rushed.
However, this open state is surprisingly fragile. When AI is integrated as autocomplete or instant correction, it actually reinforces closed behavior. What makes the difference is not the level of intelligence, but how you engage with it. Is it a dialogue partner or a finisher? That distinction continues to hold meaning.
How do people reassemble complexity around simplicity?
At this point, another paradox becomes clear. No matter how open and minimal a tool may be, people will reassemble complexity around it. Orthodox practices emerge for Markdown, hierarchies emerge in PKM systems, and editors inevitably become burdened with moral weight.
Conflict does not arise because tools are complex. It arises because tools become symbols. Open systems invite projection, projection invites identity, and identity invites defense.
Holy wars do not return because simplicity has failed. They rise again because meaning has been added to it.
To remain open without becoming rigid.
Noticing this tendency does not mean letting go of open tools. What is needed is humility regarding what tools can and cannot do.
Plain text does not guarantee freedom. It only delays enclosure. Open surfaces do not erase dogma either. They only make it easier to forget that it is taking shape.
The reason plain text has been attractive was not because of its purity. The reason plain text has been attractive was not because of its purity. It was because it protected a space where uncertainty could remain as uncertainty. Pen and paper have quietly achieved this. Plain text extended that into the digital world. And now, dialogue with AI offers another form of that promise.
However, none of these methods exempt us from our own habits. The temptation to turn freedom into factions and simplicity into ideology always remains.
Perhaps the most sustainable stance is not enthusiasm or rejection, but remembering the positioning of our tools. Tools are scaffolding for thought, not proof of it. Open systems are invitations, not guarantees.
Thought requires time before it is given form. Allowing that grace period, and noticing when you have stopped allowing it—that may be the most important practice of all.
The original English article is here