There is a moment, somewhere around the fourth or fifth minute of thumbing down a social-media feed, when the thumb begins to move on its own. The eyes are still scanning, the brain is still half-registering images and headlines, but the decision to keep going has quietly left the building. The scroll has become automatic. That moment — that precise handoff from intention to reflex — is, I would argue, one of the more consequential small events in modern daily life, and we have barely begun to reckon with it.
The infinite scroll was patented by Aza Raskin in 2006. Raskin has since said publicly that he regrets the invention, estimating that it costs humanity roughly 200,000 collective hours per day. That figure is impossible to verify and is probably not meant to be precise. But the direction of the claim is right, and the mechanism behind it is straightforward enough: remove the natural stopping point — the bottom of the page, the end of the column, the click to the next page — and consumption expands to fill whatever time is available.
Print newspapers had stopping points. The evening news had a closing signature. Even the early web, with its paginated results and deliberate click-through, imposed a small friction that made you conscious of moving from one thing to the next. Infinite scroll eliminated that friction entirely. What followed was not simply more reading; it was a different kind of reading, or perhaps a different kind of not-reading: a state of continuous partial attention that psychologists have been studying with increasing concern for a decade and a half.
The mechanics of capture
Understanding why the scroll works so well requires a brief detour into behavioral psychology. Variable-ratio reinforcement — the same schedule that makes slot machines so difficult to walk away from — is built into every feed. Most posts are unremarkable. Occasionally, one catches you off guard: something funny, something infuriating, something that touches a nerve. The unpredictability of the reward is precisely what makes the behavior persistent. If every post were equally interesting, or equally dull, it would be far easier to stop. The randomness keeps you in.
Feed algorithms amplify this effect. They are not neutral curators; they are optimized, in the most literal engineering sense, for engagement. Engagement means time on platform. Time on platform means more ad impressions. The algorithm does not care whether the content that holds your attention is illuminating or infuriating, accurate or false, good for you or bad for you. It cares only that you stay. This is not a conspiracy; it is a business model, written in code, executing at scale.
What gets crowded out
The practical costs are worth naming plainly. The first is time. Americans reportedly spend several hours per day on their phones, a large share of that on social feeds. Whatever those hours displace — sleep, conversation, exercise, sustained reading — is a real loss, even if it is largely invisible because it is distributed across millions of individual choices that each feel minor.
The second cost is attentional. There is a reasonable body of research suggesting that frequent task-switching — the constant micro-interruptions of a scrolling feed — degrades the capacity for sustained concentration. Whether that degradation is permanent, partial, or fully reversible with changed habits is still genuinely contested among researchers. But the directional concern seems warranted.
The third cost is epistemic. A feed that surfaces content based on engagement rather than accuracy or importance creates a systematically distorted picture of the world. Events that generate strong emotion — outrage, fear, tribalism — travel further and faster than events that are merely important. Over time, a person whose primary news diet is a social feed may develop a sense of the world shaped more by what is emotionally resonant than by what is actually significant.
None of this is unique to the scroll. Sensationalism in media long predates the algorithm. But the scale and the personalization are new. A newspaper’s front page was the same for every reader. Your feed is yours alone, tuned to your particular set of triggers.
What a responsible alternative might look like
Criticizing the scroll is easier than proposing something better, so it is worth being specific. Several platforms have experimented with chronological feeds, which at least remove the engagement-optimization layer, even if they do not restore a stopping point. Some researchers and designers have advocated for explicit “session endings” — prompts that ask whether you intended to keep going. A few apps have introduced time limits, though these are trivially easy to dismiss. None of these feel like adequate answers.
What might actually matter is a change in the underlying metric that platforms optimize for. If regulators or advertisers or users themselves began demanding that platforms optimize for something other than raw time-on-platform — comprehension, satisfaction, accuracy of the information consumed — the design incentives would shift. That is a large structural change, and it would be naive to expect it to happen quickly or voluntarily. But it is the right problem to name.
In the meantime, the more modest interventions are not nothing. Turning off autoplay. Using RSS readers, which present a finite list of articles and stop when the list is done. Reading long-form pieces on a single page rather than through a feed. Choosing, deliberately, when to begin and when to stop. These are small decisions. They do not fix the underlying architecture. But they return agency to the reader, and that, at the level of individual experience, is not a small thing.
A note on balance
It would be unfair not to acknowledge what the infinite scroll gives back. Feeds have connected people across distances, surfaced voices that would never have reached a mainstream platform, and delivered genuinely valuable information to people who might not have sought it out. The young activist whose cause reaches an audience of thousands through a shared post; the person in a rural area who finds a community of people with the same rare diagnosis — these are real benefits, and they should not be waved away.
The question is not whether feeds have any value. Of course they do. The question is whether the specific design choice of the infinite scroll — the removal of the stopping point, the optimization for engagement above all else — is the right way to deliver that value. I think the honest answer is that it probably is not, and that we are only slowly developing the social and regulatory vocabulary to say so clearly.
Raskin built a clever thing. It is worth asking, at last, whether clever was enough.
More from Daily Watch Reports: read more in Opinion columns, explore internet culture in Tech, or browse the slower reads in Culture.
