Every medium before this one had an edge.
A newspaper ended. A magazine ended. A television schedule ended, eventually, with a national anthem and a test card. You could finish them. Finishing was the normal outcome, and the feeling of having finished (mild, unremarkable, slightly satisfied) was so ordinary that nobody thought to name it.
We have now spent roughly fifteen years without it, and I think we have underestimated what that cost.
Completion was a feature
The argument for the infinite feed was always framed as removing a limitation. Why should you stop reading because the paper ran out of pages? Why should the arbitrary boundary of an issue dictate your attention? Put that way it sounds like liberation, and for about two years it was.
But the boundary was doing work. An ending gave you a moment of evaluation: a natural pause at which you assessed whether the last forty minutes had been worth it, and decided, deliberately, whether to start something else. That decision point is where judgement lives. Remove it and you do not get more reading. You get less deciding.
The feed's genuine innovation was not infinite content. Libraries have had infinite content for centuries and nobody scrolls a library. The innovation was removing every place where a person might reasonably stop.
The tell is what gets measured
You can tell what a system is for by looking at what its builders count. Nobody at a platform company has ever been promoted for reducing session length. The metric is time, or sessions, or some laundered proxy for both, and every design decision that survives contact with that metric is a decision that makes stopping harder.
This is not a conspiracy and I want to be careful not to describe it as one. It is something duller and more durable: an optimisation process with a well-specified objective, run for a decade by intelligent people who were each doing their jobs correctly. Nobody had to intend the outcome. The objective function intended it on everyone's behalf.
What “just use it less” misses
The standard response to all of this is individual: use it less, set a timer, delete the app on weekends. I have no objection to any of that and I have done all three. But locating the problem entirely in the user's willpower is a category error, and a convenient one for the people who built the thing. We do not tell people to be more careful about buildings without fire exits.
The relevant question is not whether an individual can resist a system built by several thousand engineers to be difficult to leave. Some can, most cannot, and the distribution of who can is not flattering to anyone's theory of personal virtue.
Small evidence that they might
The email newsletter (an old, unglamorous format) has had a decade-long revival for reasons I think are mostly about endings. A newsletter arrives, you read it, it is over. The format cannot scroll forever because there is no forever in it. Its constraint is not nostalgia; it is the whole value proposition, arrived at accidentally.
The same is true of the print-run zine, the finite podcast series, the personal website with eleven pages on it. None of these are going to displace anything. They are not supposed to. They are evidence that a non-trivial number of people will actively choose a medium that permits them to be done.
What I actually want
Not less content. Not a return to scarcity, which was never as good as it is being remembered. An edge. A place the thing stops on its own, without requiring me to supply the discipline that the design deliberately removed.
Newspapers had one for three hundred years and it was not their weakness. It was the part that let you put them down and think about something else - which is, if you take reading seriously at all, the entire point of having read.