The Importance of Long-Form

We live in an extraordinary age. More information is available to more people, at a lower cost, than at any point in human history. The collective knowledge of civilization is accessible from devices we carry in our pockets, and artificial intelligence has made it possible to retrieve, analyze, and synthesize that information almost instantaneously. Yet amid this unprecedented abundance, we appear to be losing something essential: the ability and willingness to slow down long enough to think deeply about what we're consuming.

Most of us recognize that social media presents dangers, particularly for young people. However, our concerns typically revolve around screen time, addictive behavior, unrealistic expectations, or the psychological consequences of constantly comparing ourselves to others. While these concerns are legitimate, they overlook a potentially more consequential problem: the medium through which we consume information fundamentally influences how we learn to think.

More specifically, short-form media disproportionately rewards emotional manipulation and cognitive laziness, whereas serious long-form content demands intellectual rigor and develops our capacity for independent thought. The distinction is not simply a matter of how much information we consume, but rather the intellectual faculties required to consume it. What succeeds in one medium frequently fails in the other, and the habits we develop through each increasingly influence how we understand the world.

This distinction has become particularly important in our home. As parents, when we express frustration with our children's social media consumption, we're not simply concerned about how much time they're spending on their devices. We're concerned about the intellectual habits they're developing during some of the most formative years of their lives. A steady diet of short-form entertainment conditions us to expect constant stimulation, instant gratification, and conclusions requiring very little intellectual effort. A steady diet of books, particularly those that challenge our existing understanding, exercises a different set of faculties: concentration, patience, comprehension, inference, and the ability to evaluate an argument without being emotionally captured by the person making it.

These concerns extend far beyond childhood. The decline of long-form reading may be contributing to some of the most consequential intellectual and cultural problems of our time, including political polarization, the deterioration of public discourse, and our growing difficulty distinguishing sound reasoning from emotionally compelling nonsense. And with artificial intelligence rapidly changing how we acquire and produce knowledge, the intellectual habits we're abandoning may soon become more valuable than ever.

A Postliterate America

A recent informal poll of my extended family revealed how few books many of us read. Shortly thereafter, The Atlantic published a fascinating article by Rose Horowitch titled The End of Reading Is Here, demonstrating that this phenomenon extends far beyond one family's reading habits.

A study analyzing approximately 236,000 responses to the American Time Use Survey found that the proportion of Americans who read for pleasure on any given day declined from 28 percent in 2004 to just 16 percent in 2023. Perhaps most astonishingly, the definition of reading was remarkably inclusive. It counted books, magazines, newspapers, e-books, and even audiobooks! In other words, roughly 84 percent of Americans did not engage in any of those activities on a typical day.

And the deterioration is not limited to how frequently we read. The complexity of what we read appears to be declining as well. Sentences in New York Times best sellers are approximately one-third shorter than they were a century ago. While sentence length alone is hardly a measure of intellectual quality, it offers an interesting indication of how expectations placed upon readers have changed. More concerning, nearly 30 percent of American adults cannot paraphrase or draw inferences from a multipage text, compared with fewer than 20 percent in 2017. Among high-school seniors, just 35 percent demonstrated proficiency in skills such as evaluating arguments and interpreting complex literary themes.

The obvious conclusion might be that Americans are becoming illiterate, but that isn't quite right. We encounter written words constantly through texts, emails, headlines, social media, and the endless stream of information delivered to our phones. The problem is that we're increasingly consuming information in fragments rather than developing the ability to comprehend and synthesize complicated ideas across sustained arguments.

As Horowitch puts it, America is becoming postliterate. We have largely retained the mechanical ability to read while increasingly abandoning the more sophisticated intellectual activities that reading makes possible.

This is a critical distinction because literacy is not merely the ability to recognize words. Its greater value comes from the ability to reason through what those words mean, identify relationships between ideas, recognize contradictions, and formulate independent conclusions. Reading a book from beginning to end is one activity; understanding its argument, challenging its premises, and incorporating what survives that scrutiny into our broader understanding of reality is something considerably more demanding.

Some of the most troubling evidence in Horowitch's article comes from America's educational institutions. Professors at elite universities are increasingly encountering students who struggle with assigned books and resist the effort required to read them. At Harvard, one professor reportedly needed assistance explaining to students why excerpts and summaries cannot substitute for complete primary texts. The irony is difficult to overstate: students admitted to one of the world's most prestigious universities increasingly need to be persuaded that reading is an essential part of learning.

It would be easy to blame the students, the schools, or the phones. But doing so risks missing the underlying mechanism. If the environment in which we consume information increasingly rewards speed, brevity, and emotional reaction, we shouldn't be surprised when the habits required for sustained comprehension begin to deteriorate. The more interesting question is why that environment rewards those characteristics in the first place.

Emotional Manipulators

To understand what is happening to our information environment, we need to examine the incentives embedded in the medium itself. Consider how information competes for attention on platforms such as TikTok, Instagram, Facebook, and X. Their business models depend heavily on engagement. The longer we remain engaged, the more opportunities these platforms have to monetize our attention. Content is therefore rewarded principally for its ability to attract and retain that attention, not for its accuracy, sophistication, or contribution to our intellectual development.

And what attracts attention most efficiently? Emotion. Anger, fear, outrage, tribal loyalty, and moral indignation can be communicated almost instantaneously. Intellectual rigor, by contrast, requires patience, context, evidence, and often a willingness to acknowledge uncertainty. It is extraordinarily difficult to communicate the nuances of monetary policy, geopolitics, theology, or complex scientific questions in a thirty-second video. It is considerably easier to identify a villain, assert a conclusion, and provoke an emotional response.

Chris Voss understood the extraordinary influence of emotion on human behavior, as he explains in Never Split the Difference. Recognizing and understanding emotion can make us better listeners, negotiators, and communicators. But that same understanding can be used to bypass careful reasoning and manipulate an audience. This is hardly a revelation to anyone who has spent time around a marketing department!

Consequently, short-form media creates a selection mechanism that disproportionately rewards those skilled at emotional manipulation. This does not mean everyone who succeeds on social media is intentionally dishonest, or that useful information cannot be communicated concisely. It means the incentives favor certain forms of communication regardless of their underlying truthfulness. A carefully researched argument can lose an attention contest to a provocative falsehood simply because the latter is easier to understand, quicker to consume, and more emotionally satisfying.

The most effective emotional manipulators need not construct rigorous arguments because their audiences are seldom required to evaluate them. We're encouraged to react rather than reason, affirm rather than investigate, and share rather than understand. The creator receives immediate feedback through views, likes, and engagement, reinforcing whatever behavior produced the desired response. The audience receives its own reward: the satisfaction of having an existing belief reinforced, often without the discomfort of having to examine it.

Over time, this creates a feedback loop between the incentives of the platform, the behavior of its creators, and the intellectual habits of its consumers. Creators become increasingly adept at communicating in ways that provoke emotional reactions, while audiences become increasingly accustomed to receiving information in a form that requires little independent intellectual effort. As consumers, we're not passive victims of this process. Every time we choose emotionally gratifying content over substantive analysis, we help reinforce the very incentives we ought to be questioning.

Emotional manipulation is an intellectually lazy strategy, but it can be an extraordinarily effective one. The problem becomes self-reinforcing when audiences lose the patience or ability to distinguish manipulation from reasoning. An argument's emotional appeal increasingly substitutes for the evidence necessary to support it, and confidence begins to look suspiciously like competence.

This is not unique to any political ideology. The same mechanisms can reward a progressive activist, a conservative commentator, a religious provocateur, or a financial influencer. The audiences and emotional triggers may differ, but the underlying incentives remain remarkably similar. One of the more interesting characteristics of our political environment is how effectively otherwise opposing ideological groups have adapted to the same communications model. We may disagree about virtually everything while becoming remarkably alike in how we communicate those disagreements.

Long-form content operates under a different set of constraints. An author seeking to persuade us over thousands of words must develop an argument capable of surviving extended scrutiny. Claims require supporting evidence, premises must connect logically to conclusions, and contradictions become increasingly difficult to conceal. As readers, we have the opportunity to consider the author's reasoning, revisit earlier statements, and evaluate competing explanations without being carried along by the emotional momentum of a rapidly changing video.

Of course, long-form content can be manipulative, dishonest, or intellectually bankrupt. History is full of lengthy books defending terrible ideas. Length does not establish truth, just as brevity does not establish falsehood. Some of the greatest wisdom ever written appears in remarkably concise forms, including the biblical Proverbs. The distinction is not that short ideas are inherently inferior, but that a communications environment dominated by fragments increasingly discourages the sustained intellectual work necessary to evaluate them.

A compelling thirty-second argument may disintegrate when subjected to thirty minutes of careful questioning. Conversely, a sound argument should become clearer and stronger as we examine its assumptions, evidence, and implications. This is where the concept of antifragility becomes relevant. A rigorous intellectual framework should not merely withstand criticism; it should improve through exposure to challenges, contradictions, and attempts at falsification. Serious reading creates opportunities for that process, particularly when we deliberately engage with ideas that challenge our existing beliefs.

There is an important difference between consuming information that makes us feel intelligent and engaging with information that actually makes us more intelligent. The former rewards certainty and affirmation; the latter requires humility, effort, and the willingness to discover that something we believed was wrong. If our objective is to understand reality rather than simply defend our preferred interpretation of it, we should be willing to invest in the intellectual habits that make such understanding possible.

The Compounding Mind

One of the more fascinating observations in Horowitch's article concerns the fundamental difference between written and spoken language. Writing separates information from the person communicating it, allowing ideas to be examined independently of the speaker's charisma, social standing, emotional delivery, or persuasive abilities. Because written arguments endure, we can revisit them repeatedly, identify inconsistencies, and discover meaning we may have initially overlooked.

This was among the great intellectual advances of civilization. Written language made it possible to preserve knowledge across generations, build upon discoveries, and develop increasingly sophisticated systems of reasoning. Walter Ong, the historian and author of Orality and Literacy, argued that writing transformed human consciousness by creating the conditions for sustained concentration, logical deduction, and analytical thought. The written word did considerably more than preserve what humanity already knew; it changed what humanity was capable of discovering.

Consider what this means in practical terms. When listening to a persuasive speaker, we're influenced by tone, emotion, confidence, delivery, and the social context in which the message is presented. These attributes can be useful, but they can also obscure weaknesses in the underlying argument. When reading, especially when the subject is complex, we have an opportunity to separate the message from the messenger and evaluate the ideas themselves. We can stop, question, investigate, reread, and consider whether the conclusion actually follows from the evidence provided.

Writing provides similar advantages when we're the ones constructing the argument. It requires us to make our reasoning explicit and produces a record of what was actually claimed, which can subsequently be compared against evidence, challenged, or falsified. A disagreement conducted through careful writing can become a genuine intellectual exercise rather than a contest of rhetorical agility. The point should not be to win an argument through superior delivery, but to improve our understanding of whatever truth the argument is attempting to uncover.

The distinction becomes particularly consequential when examining complicated subjects. Understanding an economic system, for example, requires more than memorizing a collection of isolated facts. It requires recognizing relationships between incentives, capital allocation, human behavior, historical precedent, and unintended consequences. Each additional layer of understanding informs how we interpret subsequent information, and those relationships eventually produce insights that would be difficult to arrive at through disconnected fragments of information.

Knowledge compounds, but only when we develop the intellectual capacity to connect what we know. This is one of the great benefits of reading across disciplines that initially appear unrelated. Economics, history, philosophy, psychology, theology, and technology each provide a different perspective on human behavior and the world we inhabit. As our understanding deepens, ideas that previously appeared disconnected begin to reveal relationships. What might seem like an enormous intellectual leap to someone encountering a conclusion for the first time may simply be the result of connecting ideas accumulated through years of reading, contemplation, and experience.

And there is a particularly important reason to read books written long before our own time. Contemporary commentary tends to reflect contemporary assumptions, controversies, and intellectual fashions. The perspectives of earlier generations offer something different: an opportunity to examine our own era through minds that were not shaped by its prevailing narratives. Reading Marcus Aurelius, Adam Smith, or Karl Popper does not require us to accept their conclusions. It gives us the opportunity to understand their reasoning, examine the worlds in which their ideas developed, and evaluate whether those ideas survive scrutiny today.

When we restrict our intellectual diet to contemporary information, we risk mistaking the assumptions of our own moment for universal truths. History provides a much larger sample of human experience. It reveals recurring patterns, demonstrates the consequences of ideas that have already been tested, and reminds us that many supposedly novel problems have confronted humanity before. It can also expose weaknesses in our preferred historical analogies, which is precisely why we should examine the original arguments rather than rely exclusively on someone else's summary.

Reading should therefore be considerably more than a passive exercise in accumulating information. The objective is not to collect the greatest number of books completed or the largest library of quotations. It is to engage with ideas deeply enough that they become useful instruments for understanding reality. Sometimes that requires taking notes, returning to difficult passages, researching an unfamiliar concept, or deliberately seeking out the strongest argument against a conclusion we find compelling. None of this requires extraordinary intelligence. It requires curiosity, humility, discipline, and a willingness to do the work.

Writing reinforces the same process. We frequently discover what we think only when we attempt to explain it coherently. The act of constructing an argument exposes assumptions we did not realize we were making, inconsistencies we had overlooked, and relationships we had not previously recognized. It forces us to distinguish what we believe from what we can actually demonstrate. In that sense, writing is not simply the presentation of intellectual work; it is part of the work itself.

The implications of developing these capabilities have always been substantial. But the emergence of artificial intelligence introduces an extraordinary new dimension. As machines become increasingly capable of retrieving information, generating explanations, and producing sophisticated written arguments, the ability to independently evaluate, connect, and challenge ideas may become more consequential than ever.

Intelligence Inequality in the Age of AI

Perhaps the most consequential finding in The Atlantic is that reading has become increasingly concentrated among a relatively small segment of the population. According to the article, just 20 percent of American adults accounted for more than 80 percent of all books read in the preceding year. While the broader population is reading less, those who remain committed to reading are often reading more.

This suggests something more significant than a general cultural decline. We may be witnessing an increasingly pronounced divergence between those who continue developing the intellectual faculties associated with sustained reading and those who increasingly rely upon simplified, emotionally compelling information. The distinction is not necessarily one of innate intelligence. It is a difference in how consistently we exercise our capacity to think, reason, and understand. And because those capabilities can be developed through deliberate practice, this is an inequality that remains substantially within our power to address.

Throughout most of human history, access to information represented an enormous competitive advantage. Books were expensive, libraries were limited, literacy was far from universal, and acquiring specialized knowledge was often restricted to a relatively small portion of the population. Today, virtually the entire accumulated written knowledge of human civilization is accessible to anyone with an internet connection. Yet our willingness to engage deeply with that knowledge appears to be declining.

Artificial intelligence introduces a fascinating paradox. We have created tools capable of synthesizing enormous quantities of information, examining competing explanations, and helping us reason through problems that previously required extraordinary time and effort. At the same time, we're increasingly abandoning the habits that would allow us to use those tools most effectively.

The obvious assumption is that AI will make intellectual capability more equal by giving everyone access to an extraordinary research and reasoning assistant. There is considerable merit to that argument. AI can help novices perform tasks that previously required specialized knowledge, explain difficult concepts, and overcome barriers that once prevented people from participating in sophisticated intellectual work. But improving access to answers is not the same thing as developing the judgment necessary to evaluate them.

Consider two people using the same AI system to investigate a complicated economic question. One asks for an explanation, receives a persuasive answer, and accepts it. The other understands enough about economics, history, and human behavior to recognize that the question itself may contain faulty assumptions. Rather than accepting the initial response, that person asks the system to examine alternative explanations, identify contradictory evidence, distinguish correlation from causation, reconsider the historical analogies, and explain what would have to be true for its conclusion to be wrong.

Both have access to the same technology. Their results may be profoundly different, not because one possesses a higher innate intelligence, but because one has developed a broader intellectual foundation and the habits necessary to make the technology more useful.

The distinction extends beyond knowing how to write an effective prompt. Those of us who have spent years reading across history, philosophy, science, economics, and other disciplines possess a larger collection of concepts, relationships, and historical precedents against which to evaluate AI-generated explanations. We are better positioned to recognize when something is missing, identify a questionable assumption, or connect an unfamiliar idea to something learned years earlier. In effect, the accumulated knowledge and reasoning habits developed through reading can become force multipliers for the capabilities of AI.

Research is beginning to illuminate this distinction. A 2025 study by researchers at Microsoft and Carnegie Mellon surveyed 319 knowledge workers about their use of generative AI. Greater confidence in AI was associated with less reported critical-thinking effort, while greater confidence in one's own ability was associated with more. The researchers also found that AI was changing the nature of intellectual work, shifting effort away from gathering information and toward verifying outputs, integrating responses, and overseeing the quality of the finished product. These were self-reported associations, not proof that AI causes intellectual decline, but they raise an important question about which habits our growing reliance on technology will reinforce.

The more capable AI becomes, the more tempting it will be to substitute its conclusions for our own reasoning. Yet the apparent sophistication of an answer tells us remarkably little about its truthfulness. AI can produce elegant explanations built upon flawed premises, incomplete evidence, or outright factual errors. Without the knowledge or intellectual discipline necessary to recognize those failures, we risk confusing the ability to generate a persuasive answer with the ability to understand whether it is correct.

That is why the relationship between AI and intelligence inequality may be more complicated than it first appears. AI can raise the floor of what less experienced people are capable of accomplishing while simultaneously raising the ceiling for those who have developed the judgment to direct it effectively. Whether the overall gap narrows or widens remains an empirical question. But a technology that makes information easier to obtain does not automatically make wisdom, discernment, or independent thought equally abundant.

In fact, the combination of declining reading habits and increasingly powerful AI introduces an uncomfortable possibility: we could become extraordinarily capable of producing intellectual work while becoming progressively less capable of evaluating it. A world overflowing with convincing arguments is not necessarily a world that understands more of what is true.

The Work We Cannot Outsource

The value of long-form reading therefore extends far beyond an appreciation for literature or a preference for books over social media. It is a deliberate investment in the capabilities that allow us to make sense of the world, recognize manipulation, challenge our own assumptions, and reach conclusions independent of whatever narrative happens to be most popular.

Fortunately, these capabilities are not reserved for a small collection of unusually gifted people. We develop them through the same basic mechanism by which we develop most worthwhile abilities: consistent, deliberate effort. Just as physical strength improves when we repeatedly challenge our muscles, our intellectual capabilities improve when we regularly undertake work that requires sustained concentration, comprehension, and reasoning. The objective is not to demonstrate intellectual superiority over anyone else. It is to become more capable than we were yesterday.

That begins with a choice about how we spend our attention. We can devote another hour to a stream of provocative headlines, thirty-second videos, and conclusions packaged for effortless consumption, or we can spend some of that time working through a serious book. The immediate reward may be less exciting, but the cumulative benefit can be considerably greater. We gain access not merely to information, but to the reasoning of people who may have spent decades investigating questions we can begin exploring in a single afternoon.

The choice of what we read matters as well. We should seek out books that explain unfamiliar subjects, develop arguments over many chapters, and challenge ideas we already hold. We should read authors whose conclusions we disagree with, and we should venture well beyond contemporary commentary into history, philosophy, science, and the works that shaped earlier civilizations. Encountering an argument we cannot easily dismiss is not an inconvenience. It is an opportunity to discover where our own understanding might improve.

Nor should we regard AI as the enemy of this process. Properly used, it can be an extraordinary intellectual companion, capable of explaining unfamiliar concepts, testing arguments, identifying counterexamples, and helping us explore relationships that might otherwise remain undiscovered. But there is an enormous difference between using AI to challenge and improve our reasoning and using it to avoid reasoning altogether. We should welcome tools that make intellectual work more productive without surrendering the responsibility to determine whether that work is sound.

These distinctions are especially important for our children, whose habits of attention and thought are still developing. The objective should not simply be to restrict access to social media or demand that they read more books. It should be to help them understand why their intellectual independence matters and give them opportunities to experience the satisfaction of working through something difficult. Rather than teaching them to depend upon others for answers, we should help them develop the curiosity, discipline, and confidence necessary to investigate questions for themselves.

The same challenge applies to all of us. We may be unable to control the incentives of social media platforms, the intellectual fashions of our institutions, or the extraordinary pace at which artificial intelligence is advancing. But we retain considerable control over the habits we cultivate, the information we consume, and the effort we're willing to invest in developing our own understanding.

This is part of the purpose behind The Grain. Serious ideas deserve serious consideration, and the pursuit of truth should not be constrained by the attention span of an algorithm. The articles published here will often require more than a few minutes to read, not because length is an achievement in itself, but because some arguments deserve the context, evidence, and reasoning necessary to evaluate them properly. The ambition is not to persuade readers to accept a particular worldview without question, but to help us all become clearer thinkers, more capable of recognizing error, and more willing to follow the evidence wherever it leads.

The importance of long-form is ultimately not that it allows us to consume more information. It is that it helps us develop the intellectual capacity to understand what we consume, challenge what we are told, and connect what we learn into a coherent understanding of reality.

As information becomes abundant and answers become effortless, our willingness to think may become one of the most consequential choices we make.

The opportunity is available to all of us. But we still have to do the reading.