RK Kuppurao · Archive

The Cost of Being Captured

Blogs killed long-form writing. Tweets weakened blogs. Shorts disrupted long-form video. Reels accelerated the same cycle.Comments weakened conversation. Likes weakened comments.

That story is simplified, but it points to something real: each new format tends to reward speed, compression, repetition, and emotional reaction. Over time, the incentives shift away from reflection and toward whatever can capture attention most reliably.

We call this the attention economy: a system in which human attention is scarce, measurable, and valuable. Platforms compete for it. Creators compete for it. Advertisers buy it. Data companies track it. And users — often without realizing it — give it away in small increments throughout the day.

The consequence is not merely that we spend too much time online. It is that the structure of online media changes how we think, what we encounter, and what we come to believe deserves our attention.

The Incentives Behind the Feed

Most platforms do not primarily optimize for whether content is useful, true, thoughtful, or memorable. They optimize for signals that are easier to measure: clicks, watch time, completion rate, comments, shares, likes, return visits, and total time spent on the platform.

That is why so much online content is designed around retention.

On YouTube, for example, creators have strong incentives to keep viewers watching. A 20-minute video that loses most of its audience after two minutes performs poorly compared with one that keeps viewers engaged longer. This encourages a familiar structure: an intriguing promise at the beginning, a series of partial answers in the middle, and the most valuable information delayed until the end.

“There is one more thing you need to see.”

“Wait until you hear the final point.”

“The answer may surprise you.”

Sometimes this is simply good storytelling. But often it is not. It is a strategy for preventing the viewer from leaving.

Creators are not necessarily to blame for responding to these incentives. If platforms reward watch time, creators will design for watch time. If they reward comments, creators will ask for comments. If they reward engagement, creators will find ways to manufacture engagement.

This is why viewers routinely hear requests such as: “Please like this video to help the algorithm.”

The phrasing is almost absurd when examined closely. The viewer is being asked to perform unpaid labor for the distribution system of a platform that already profits from the viewer’s presence. But it works because creators depend on visibility, and visibility depends on the platform’s metrics.

YouTube benefits when viewers remain on YouTube. The longer people remain there, the more recommendations they receive, the more advertisements can be shown, and the more data can be collected about what holds their attention.

The growth of extremely long videos — eight-hour ambience tracks, twelve-hour compilations, twenty-four-hour livestreams, and endless background content — is an expression of the same logic. In many cases, the goal is no longer to deliver a clearly bounded piece of information or entertainment. The goal is to occupy time.

Engagement Is Not Conversation

Comments are often described as engagement, but engagement is not the same as conversation.

A real conversation involves attention, context, responsiveness, and some willingness to revise one’s view. Most comment sections offer very little of that. They are optimized for visibility, speed, performance, outrage, and agreement from strangers.

Many comments are harmless. Some are funny or insightful. But a large share are reflexive, repetitive, hostile, or designed to attract attention themselves. Increasingly, they are also generated, amplified, or shaped by artificial intelligence.

A comment section can therefore create the appearance of public discussion without delivering much actual understanding.

Likes are even more limited. A like is a low-effort signal: “I saw this,” “I agree,” “I want to be seen agreeing,” “I support this person,” or simply “I want the platform to remember that I interacted with this.” It does not require explanation, disagreement, thought, or dialogue.

The transition from conversation to comments and from comments to likes is not automatically bad. It makes communication faster and easier. But it also reduces the amount of depth required to participate.

We increasingly confuse activity with engagement, engagement with understanding, and popularity with value.

The Architecture of Doomscrolling

Instagram, TikTok, YouTube, X, and similar platforms have become highly effective at identifying what makes people pause.

They do not merely track whether you liked a post. They can infer interest from a much wider set of behaviors: how long you watch, whether you replay, whether you pause for a moment, whether you read comments, whether you share, whether you search for related content, and whether you return to the app later.

A brief pause can become a signal.

The system does not need to know why you paused. It may have been curiosity, disgust, fear, admiration, anger, confusion, or accidental attention. But if the signal looks like interest, the platform may show you more of the same.

This is the engine of doomscrolling. The feed does not need to persuade you that every piece of content is important. It only needs to find one more thing that makes you hesitate before you close the app.

The result can be a strange form of captivity: content that you do not particularly enjoy, do not consider useful, and do not even remember later can still keep you engaged for an hour.

The deeper concern is not simply distraction. It is narrowing.

Recommendation systems learn from what you consume and then offer more of it. That can be useful when the subject is practical and low-stakes. If you are researching running shoes, backpacks, hiking boots, recipes, or a new laptop, recommendations can save time.

But the same mechanism becomes more troubling when it shapes political, cultural, ideological, or moral beliefs.

If a platform learns that you respond to a particular political framing, grievance, conspiracy, fear, or social identity, it may keep reinforcing that pattern. Over time, opposing views appear less often. Contradictions become easier to dismiss. The most emotionally satisfying explanation begins to feel like the only explanation.

An echo chamber does not necessarily require censorship. It can emerge simply because the system keeps serving you what you are most likely to consume.

That weakens critical thinking. It reduces exposure to disagreement. It makes contrarian views feel rare, extreme, or illegitimate. And it can gradually replace inquiry with confirmation.

Journalism Has Its Own Incentive Problem

Social media is not the only institution shaped by attention incentives. Journalism has been dealing with them for much longer.

The ideal of journalism is straightforward: gather evidence, verify claims, distinguish facts from speculation, present relevant context, and hold power accountable. In practice, journalism has always been influenced by owners, advertisers, political pressures, audience preferences, editorial judgment, and commercial competition.

Today, many consumers can identify the broad ideological tendencies of major news organizations almost immediately. That does not mean every report from those organizations is false or biased beyond use. It does mean readers increasingly approach news with an awareness that framing matters.

Which facts are emphasized? Which are minimized? Who is quoted? Which stories become national controversies? Which stories receive little attention? What language is used to describe one side versus another?

The expansion of opinion journalism makes this even more complicated.

Opinion is not inherently bad. Strong arguments, criticism, interpretation, and moral judgment all have a legitimate role in public life. But opinion is governed by different standards than reporting. It is not always required to offer balance. It may select evidence to support a thesis. It may be intentionally provocative. Its purpose may be persuasion rather than truth-seeking.

The problem arises when opinion is packaged to look like reporting, or when consumers stop distinguishing between the two.

A healthy media diet requires recognizing that a headline, a news report, an editorial, a column, a podcast monologue, a viral clip, and a social-media post are not equivalent forms of information. They should not be consumed with the same degree of trust.

The Practical Response

The situation is not hopeless, but it does require intentional habits. The platforms are designed to reduce friction between impulse and consumption. The only reliable response is to add some friction back.

1. Stop doomscrolling

The simplest intervention is to reduce exposure.

You do not need to abandon every online platform. Some platforms maintain real human relationships, help people coordinate events, provide useful communities, or make work easier. The goal is not technological purity. It is to distinguish tools that serve your life from tools that consume it.

Consider which apps genuinely help you:

Then consider which apps or feeds mainly leave you distracted, agitated, envious, politically inflamed, or vaguely dissatisfied.

The right number of social-media platforms is not the same for everyone. But it is worth asking whether each one earns its place in your life.

2. Resist the recommendation system

Recommendation systems are built from your behavior. You cannot fully escape them, but you can make your signals less predictable.

Use “Not interested” when a platform offers material you do not want to reinforce. Unfollow accounts that reliably trigger low-value consumption. Clear search or watch history when that is useful. Avoid lingering on content that you do not actually want more of.

You can also become more deliberate in how you discover information.

Search for topics directly instead of accepting whatever appears in the feed. Visit sources intentionally. Subscribe to a limited number of writers, researchers, publications, creators, or newsletters whose work you value. Read outside your usual ideological lane — not to adopt every opposing view, but to understand how intelligent people can interpret the same facts differently.

A feed learns from passive behavior. Deliberate searching is one way to reclaim agency.

3. Do not reward clickbait

Clickbait works because it exploits a basic human instinct: the desire to close an information gap.

A headline suggests that you are missing an important fact, a hidden danger, a life-changing habit, an overlooked opportunity, or the explanation for a problem you have felt but could not name. It promises resolution, but often withholds enough information to make you click.

Examples include:

These headlines are not always dishonest. Sometimes the article or video delivers real value. But the structure is designed to activate uncertainty and promise a payoff that is vague enough to apply to almost anyone.

The key question is simple:

Does this headline tell me what I will learn, or does it make me feel incomplete until I click?

A useful headline is specific. It identifies the topic, the claim, and often the evidence or scope.

Compare these two titles:

The second title may be less emotionally potent, but it respects the reader. It tells you what the article is about and lets you decide whether it is worth your time.

Avoiding clickbait is not a trivial skill. Even disciplined people are vulnerable to it because clickbait is designed around curiosity, anxiety, hope, and social comparison. Developing the habit of ignoring it is a form of attention training.

For the next decade or several decades, that may be one of the most valuable forms of self-protection available: the ability to decide what deserves your attention before someone else decides for you.