Think about what you read last week — news, feeds, posts, whatever your version is. Now try to say what you learned.
Most people can produce almost nothing, from several hours of reading. Not because the material was worthless, but because reading without writing produces a sensation of being informed and very little retention, and the sensation is convincing enough that nobody investigates it.
The problem is not volume, and reading less does not by itself fix it. Reading four hours and retaining nothing and reading one hour and retaining nothing differ only in the hours.
The three failure modes
The scroll. Continuous partial reading, mostly headlines and openings, producing a general awareness that things are happening and no specific knowledge of any of it. The most common by a wide margin.
The saved pile. Read-later services and a folder of tabs. Saving feels like reading and it is a substitute for it — the pile grows, occasionally gets declared bankrupt, and produces roughly nothing over its lifetime. The same failure as an unpruned someday list.
Following without accumulating. You genuinely read about a topic, regularly, over years, and cannot say what your view is — because each piece was consumed in isolation and nothing ever built.
All three are the same missing step. Nothing is written down, so nothing accumulates.
One note per topic, not per article
The change that fixes it.
Do not save articles. Keep a note per thing you care about — three or four of them, not thirty — and add to it when you read something worth adding.
Two or three lines: what this piece claimed, whether it changed your view, and the source. Over a year that note is a genuinely informed position, assembled from a hundred small readings, in your own words. It is the thing you wanted when you started reading about the subject, and no reading list ever produces it.
The reason it works is that adding to a topic note forces the step reading skips: deciding what this piece actually said and whether you believe it. That decision is the whole of the value, and it takes about forty seconds.
What is worth writing down
Not summaries of the article — the article summarises itself and you can find it again.
Anything that changed your mind. The most valuable and the rarest. Note what you thought before, which is the part that evaporates.
A number worth remembering, with its source and date. Numbers without sources are how people end up confidently wrong.
A claim you want to check. Rather than accepting it or dismissing it, park it. This is the single best defence against absorbing something because it was stated confidently.
A connection to something else you know. Where understanding actually comes from — the same material an idea bank collects.
What you now think. One line, occasionally. Your view, dated. Rereading your own view from two years ago is unusually informative, and unavailable any other way.
The three-day rule
Practical filter for anything that is news rather than knowledge.
If it is genuinely important, it will still be there in three days, better reported, with the corrections already made. The first version of a breaking story is routinely wrong in details and sometimes in substance, and reading it costs attention and then costs it again when the accurate version appears.
Deliberately waiting produces a strictly better version of the same information at lower cost. It also removes most of the emotional weight, which is largely a function of immediacy rather than importance.
This is the single change with the biggest effect on how much reading feels like drowning.
Sources, briefly
Two observations, both unfashionable.
A small number of good sources beats a large number of any. Three or four things you actually read beats forty subscriptions you skim. The marginal source adds volume, not information, because sources overlap enormously.
Pull beats push, here as elsewhere. A feed reader you open is better than notifications that arrive, for the same reason it is true of reminders: pushed information arrives when you cannot think about it, and gets processed with the reflex rather than the attention.
The quarterly read-back
Fifteen minutes, and it is what turns notes into knowledge.
Read your topic notes. Three things happen: you find you know more than you thought, which is genuinely encouraging; you find predictions you made that were wrong, which is instructive and slightly uncomfortable; and you find the topics you have stopped adding to, which is a signal you have stopped caring about them and can stop reading about them.
That last one is how the list of things you follow stays small. Topics you no longer add to are topics you can drop, and dropping them is what makes room.
The honest version
Some reading is for pleasure and should not be processed at all. A novel, a good essay, something interesting for its own sake — writing notes about that is a way of turning enjoyment into homework, and it is why some people ruin reading for themselves.
This applies to the reading you do because you want to know something, and the test is simple: if you would be annoyed to have retained nothing, write something down. If you would not, do not.
Cyanote holds the topic notes: one per thing you follow, growing over months, searchable by any word in them years later with ⇧⌘F, and [[ links tying a topic to whatever else it turns out to touch. It has no reader, no feed and no clipper — the article stays where it is, and what you write about it is the part worth keeping. One local database on your own Mac, so a decade of your own views on things you cared about stays yours.