Think about the last time you finished a book. At some point you turned the final page and there was nothing under it. The story ended, and your decision to stop was made for you. Now think about the last time you stopped scrolling. You probably can't remember it, because it likely didn't happen as a decision. Something else interrupted you: a bladder, a person, a dying battery, a flash of disgust at the time. You didn't finish the feed, because the feed doesn't finish.

Most of us blame ourselves for this. We call it weak willpower. But look at the structure of the thing before you indict the person. Almost every activity in human life arrives with an ending built in, and the feed is one of the few that doesn't.

The ending you never noticed you relied on

A meal ends when the plate is empty. A TV episode ends when the credits roll. A newspaper ends at the last page, and a conversation ends when the other person says they have to go. These are stopping cues: external signals that tell you a unit of activity is complete. They convert a vague question ("Have I had enough?") into an easy one ("Is it done?").

That conversion matters because "have I had enough?" is a hard question for a human brain. Satisfaction is slow and fuzzy, so we lean on the environment to answer for us. Psychologists have documented this with food. In their work on unit bias, Andrew Geier, Paul Rozin, and Gheorghe Doros found that people treat a single unit, one pretzel, one portion, one package, as an appropriate amount to consume, largely regardless of its size. The unit supplies the answer to "how much is enough."

A stopping cue is a small piece of cognitive labor that the world does for you. Remove it and the labor falls back on you, at exactly the moment your judgment is weakest.

What infinite scroll actually removes

An infinite feed has no last page, no unit, and no credits. Aza Raskin, who designed the infinite scroll mechanic, has said publicly that he regrets how much time it consumes. The design deletes the natural pause where a page loads and you're asked, implicitly, whether you want more. Autoplay does the same thing for video: the next item begins before the question can form.

Notice what this does to the structure of your choices. With a finite activity, continuing requires a decision, and stopping is the default. With an endless feed, continuing is the default, and stopping requires a decision you must generate from scratch. Defaults are powerful because deciding costs effort. The feed has quietly flipped which side has to pay it.

That is why "I'll just check for a minute" so often turns into forty. You didn't decide to stay. You failed to produce a decision to leave, and nothing in the design asked you to.

Why "just a bit more" always feels reasonable

There's a second mechanism worth knowing about, from information foraging theory, developed by Peter Pirolli and Stuart Card. They proposed that humans seek information the way animals seek food, moving through "patches" of content and weighing the payoff of staying against the cost of moving on. In classic foraging models, an animal leaves a patch when the rate of return drops below what it could expect elsewhere.

A feed is a patch designed to hide that drop. Each swipe has a small, uncertain chance of something good: a funny post, a useful link, a message from someone you like. Because the good items are scattered and unpredictable, your estimate of the patch's value stays propped up by the last good thing you saw. The average return may have been falling for twenty minutes, but the most recent reward feels like proof that the next one is close.

So every individual moment of continuing is locally defensible. One more swipe costs almost nothing. The cost arrives only in aggregate, which is a scale your in-the-moment judgment doesn't track well.

Why the tiredness doesn't tell you either

You might expect fatigue to act as a natural brake. But scrolling asks for very little. It needs a thumb and passive reception, so it stays easy long after you've lost the energy for anything demanding. That's part of why it tends to expand at the edges of the day, when you're depleted and every other option feels heavy by comparison. The activity that's easiest to continue is also the one hardest to end.

The lesson isn't that you're defective. It's that you were handed an activity with the brakes removed and told to rely on judgment alone.

The fix: build the cue the feed took away

If the problem is a missing external cue, the solution isn't more resolve. It's to manufacture the cue yourself, ideally before you start, when your thinking is clear. This is the same logic behind any good precommitment: shift the decision to the calm moment and let the environment enforce it later.

Two features make an artificial stopping cue work. It has to be defined in advance, so you aren't negotiating mid-feed. And it has to be external, so it fires whether or not you're paying attention. A private intention like "I'll stop soon" fails both tests.

Your next moves

  • Name the unit before you open the app. Say it out loud or type it in your notes: "I'm checking messages and one feed, then done," or "ten minutes." A defined unit gives your brain what unit bias runs on: a thing that can be completed.
  • Set a physical timer, not a mental one. Use a kitchen timer or a phone alarm set to a different app than the one you're scrolling. When it rings, the cue arrives from outside the feed, which is the whole point.
  • Swap one endless source for a finite one today. Replace a feed you scroll with something that ends: a newsletter, a single article, a podcast episode with a known runtime, a daily digest. Finished is a feeling the feed will never give you.
  • Add one step of friction in front of your worst app. Log out after each session, move it off the home screen, or use the browser version instead of the app. You don't need to make it impossible, only to create a pause where a decision can happen.
  • Ask one question at each phone pickup. "What am I here for?" If you can answer, do that thing and leave. If you can't, you're about to forage in a patch with no goal, and you'll have no way to know when it's exhausted.

What to expect

The first few times, the cue will feel annoying. The timer will ring while something half-interesting is on screen, and you'll feel a small tug to override it. That tug is the missing decision, finally showing up. You don't have to win a battle; you only have to let the cue do its job and notice that the feeling of "more" passes without being fed.

Over time the finish line becomes part of how you use the thing, the way a closed book does. It also changes what the time feels like. A session with an end is something you chose. One without an end is something that happened to you.

Where Reclaim fits

What the feed removed was a stopping cue, and Reclaim is built to put one back. It lets you decide in advance, while you're clear-headed, where a session should end, and then holds that line when the next swipe is pulling at you. You don't need it to apply anything above; a kitchen timer and a defined unit will carry you a long way. But if you'd rather not rely on remembering the plan at the moment you're least likely to, it's here at reclaim.lumenlabs.works.