Infinite feeds remove endings. A pause puts one back.

A circular stream of video tiles interrupted by a golden pause and an opening toward a calm landscape
A visual metaphor for changing “next” from an automatic event into a deliberate choice.

The most useful way to describe doomscrolling is not “a dopamine loop.” That phrase sounds scientific, but it overstates what a browser extension can claim to change. A more accurate description is an automatic interaction loop:

content → swipe → next content → swipe → next content

Autoplay, instant loading, and infinite scroll remove the natural endings that make stopping easy. A pause, explicit play button, or “load more” choice creates a discontinuity:

content → boundary → choose → continue or leave

Researchers often call this design friction. It does not force a decision; it stops treating continuation as the default.

The evidence for a small interruption

The one sec research is especially relevant. In a six-week field study with 280 participants, the intervention added a brief delay before selected apps opened; target-app openings ultimately fell by 57%. A separate preregistered experiment with 500 people found that the behavioral interruption reduced consumption, while adding an explicit option to quit was more effective still. The lesson is not “a delay fixes everything.” It is that a short interruption can create a real point at which an automatic action ends. Read the original PNAS paper.

Other studies point in the same direction, with important variation:

InterventionWhat researchers foundWhat it means for NoLoop
Brief friction before opening an appThe one sec studies found fewer target-app openings and stronger effects when people could actively quit.A voluntary decision point can be more useful than a passive reminder.
Personalized full-screen boundaryIn a 3-week randomized trial of 70 iPhone users, participants with tailored reminders and an option to quit or continue reduced time in their most problematic app by about 29 minutes a day.A clear “continue or leave” boundary can support behavior change, but it is not a guarantee of broader wellbeing outcomes.
Grayscale as visual frictionA preregistered field experiment with 112 participants found an immediate, significant reduction in objectively measured screen time versus control.Even small changes to an interaction can alter use; the intervention does not need to be punitive.

Sources: one sec (PNAS), Wellspent randomized controlled trial (JMIR), Digital Strategies for Screen Time Reduction (Cyberpsychology, Behavior, and Social Networking).

Why an explicit choice matters

With autoplay, doing nothing means consuming the next video. With manual play or a confirmation boundary, doing nothing means stopping. That is a small change in interaction design, but it changes the default.

NoLoop uses that principle in practical ways:

  • Keep a video paused until you choose to play it.
  • Add an intentional pause after a run of posts or Reels.
  • Block or interrupt surfaces you have decided are not useful right now.
  • Keep useful routes—Messages, profiles, notifications, and search—available.

The goal is not to make Instagram unpleasant. It is to create an episode boundary: a natural moment where a session can end before it becomes another open-ended stream.

Why we avoid nagging

“You have been scrolling for 20 minutes. Stop!” is easy to dismiss. A boundary that changes the next interaction is different. It does not tell you what you should do; it asks whether you want to recommit.

That is why NoLoop aims for calm, small, well-timed friction rather than repeated warnings. The best setting is personal: some people may prefer a pause every few Reels, while others may prefer no autoplay or a stronger block.

What NoLoop is designed to do

NoLoop is a Safari extension and companion app that changes selected Instagram interactions locally on your device. It can create a stopping cue before continuation becomes automatic. It cannot prove that every session will be shorter, diagnose an addiction, or replace professional support.

For the careful version of the evidence—including study limitations and the difference between related interventions and NoLoop itself—read What the evidence supports—and what it does not.