The Noise Problem

Pick any day from the past decade and count what arrived in your inbox, your feed, your notifications. Now count how much of it mattered by the following week. The ratio is not improving. More arrives every year. The amount that turns out to matter stays roughly flat.

The standard complaint about the information age is that there is too much of it. That is the wrong diagnosis. The problem is not volume. Libraries have always had more books than any person could read. The problem is that we lost the filtering layer between the volume and the reader. We lost the people and institutions whose job was to say: this one is worth your time, that one is not.

What curation actually was

Editors decided what got into newspapers. Publishers decided what got into books. Programme directors decided what got on radio. These were not perfect systems. They had blind spots and biases and they kept out voices that deserved to be heard. But they also kept out an enormous quantity of drivel, and the drivel stayed out because getting past those gates required some minimum standard of coherence and evidence.

The internet did not just lower the cost of publishing. It removed the gate entirely. Every opinion now travels at the same speed and with roughly the same visual authority as every fact. A peer-reviewed study and a man shouting in his car look, in a feed, almost identical. The shouting man often gets more views.

What filled the gap left by editors and publishers was the algorithm. Algorithms do not curate. They optimise for engagement, which turns out to select strongly for outrage, novelty, and emotional intensity rather than accuracy, depth, or long-term usefulness. This is not a flaw in the algorithm's design. It is working exactly as built. The problem is that engagement and signal are not the same thing, and we handed curation to a system that cannot tell the difference.

How noise masquerades as signal

Noise does not usually announce itself. It wears signal's clothes. It has confident headlines. It arrives from sources that once were reliable. It has the look of urgency.

Take the daily news cycle. On any given day, a story will dominate coverage across every major outlet. By Friday it is gone, replaced by a different story that demands the same attention. The month of wall-to-wall coverage rarely produces any useful action or understanding in the reader. It produces the feeling of being informed, which is not the same thing.

Or take expert opinion on social media. A person with genuine expertise in one domain discovers that confident, short, strong-sounding statements get more reach than careful, qualified ones. The incentive is to compress and simplify. Nuance drops off first because nuance does not spread. What remains is the signal-shaped husk of the actual argument, minus the qualifications that made it true.

The reader on the receiving end often cannot tell. The stripped-down version sounds just as authoritative as the careful one. More, sometimes.

What judgement actually requires

Judgement is the capacity to assign weight to things without being told how heavy they are. It is calibration: knowing that some sources are worth trusting and some are not, that some questions admit confident answers and some do not, that urgency and importance are different properties.

Judgement does not develop from consumption. It develops from evaluation. There is a difference between reading a hundred articles on a topic and reading twenty of them slowly, asking which claims are supported, what is being left out, and what the author's incentives might be. The first produces familiarity. The second produces understanding.

This is why people who read fewer, better sources tend to have better-calibrated views than people who read everything. The disciplined reader who finishes three serious books on a topic and interrogates all three is usually better informed than the person who has followed the same topic daily across forty feeds for a year. Volume without evaluation is just exposure.

The other thing judgement requires is a willingness to be wrong, and to update. This sounds obvious but it runs against the logic of public discourse, where changing your position is treated as weakness. A person who reads carefully, finds new evidence, and revises what they think is doing exactly what good reasoning looks like. The alternative is opinions calcifying around whatever arrived first.

What you can do

None of this requires a dramatic retreat from information. It requires changing the ratio of consumption to evaluation.

Pick two or three sources you have found consistently reliable on questions you can actually check, and read those well rather than scanning twenty. When something arrives that provokes a strong feeling, wait before sharing it. Ask what evidence is behind the claim and what the strongest objection to it might be. Notice which topics you feel strongly about despite knowing little, and treat that as a flag worth examining.

Develop a short list of questions you apply to anything that arrives: Who made this claim? What do they stand to gain? Is this evidence or assertion? Is the urgency real or performed? None of these take long. They take habit, which is slower to build than a subscription but more durable than any tool.

The noise problem is not going away. The volume will increase. What can change is your relationship to it: from a reader who absorbs what arrives to a reader who decides what deserves absorption. That is the difference between being informed and being managed.