Before 2006, a web page ended. You reached the bottom, and there was a decision to make: click for the next page, or stop.
Aza Raskin, then a young interface designer, invented infinite scroll — the pattern in which reaching the bottom loads more content automatically, so that the page never ends. It was a genuine usability improvement, and it solved a real problem. It is now in essentially every feed on every platform.
Raskin has spent the last decade publicly regretting it, and has estimated that the pattern consumes hundreds of thousands of human lifetimes' worth of attention per year. He testified in litigation brought by the state of New Mexico against Meta.
The mechanism is the removal of a stopping cue, and it is worth being precise about why that matters.
Ongoing behavior is not continuously re-evaluated. Re-evaluation happens at boundaries — at the end of a chapter, at the bottom of a page, when the episode finishes, when the glass is empty. A boundary is what allows the prefrontal question to be asked: do I want to keep doing this. Remove the boundary and the behavior is governed by whatever moment-to-moment reward is arriving, and momentary reward is not where intention lives.
B. J. Fogg's Persuasive Technology Lab at Stanford supplied the design framework that the industry adopted. Fogg's behavior model holds that a behavior occurs when motivation, ability and a trigger converge, and that the cheapest lever is usually ability: rather than increasing motivation, reduce the effort required until the behavior happens on a trigger alone. Many of the people who built the attention economy passed through that lab, and Fogg has since written about the ethical consequences.
Combine the removed stopping cue with the variable-ratio reward of chapter 19 — each swipe delivering an unpredictable payoff, most of them worthless — and you have the slot machine's schedule with no coin slot and no closing time.
The distinguishing feature of this reward structure, and the reason it produces the specific experience users report, is that dopaminergic activity is highest in anticipation rather than at receipt. The seeking outlasts the satisfaction. That is why people describe scrolling as unpleasant and continue.
None of this required anyone to intend harm. It required a metric.
Cross-ref: Attack — Reward Unpredictably