# Track Records for Attention

Canonical: https://mosiah.org/articles/track-records-for-attention/
Interactive: https://mosiah.org/#Articles%2Ftrack-records-for-attention

//Related:// [[sources|Article Sources/track-records-for-attention]] · [[notes|Article Notes/track-records-for-attention]] · [[metadata|Article Metadata/track-records-for-attention]] · [[Published Pieces]]

! Track Records for Attention

//Attention is the first bet, and public cognition needs records of who looked in the right direction early.//

Prediction is not the only public virtue. Attention matters too.

A person can predict correctly because they were lucky. A person can predict incorrectly while still attending to the right causal structure. A person can be early to the real issue before it becomes a clean prediction market. A person can preserve a distinction that saves future discourse from confusion. A person can notice the hinge that everyone else ignored.

Public cognition needs track records for attention.

Who noticed what mattered early?

Who ignored obvious evidence because it was low-status?

Who was late but loud?

Who saw a distinction before it had a name?

Who consistently chased fashionable noise?

Who attended to the boring mechanism that later became decisive?

These questions matter because the world is not presented to us as a list of well-formed bets. It arrives as a flood of fragments: leaks, speeches, datasets, rumors, market moves, court filings, technical papers, social media posts, private incentives, institutional behavior, and cultural signals. Before anyone can forecast an outcome, someone has to decide what deserves attention.

Attention is the first bet.

But our media systems barely record it. They record virality, publication date, follower count, prestige, outrage, and sometimes predictions. They do not reliably record salience judgment. They do not tell us who watched the right object when it still looked marginal.

This is a huge blind spot.

In politics, the person who notices the procedural change may matter more than the pundit who predicts the election. In finance, the analyst who attends to a balance-sheet footnote may matter more than the trader who guesses the quarterly print. In AI, the engineer who notices a harness bottleneck or eval failure may matter more than the public forecaster debating timelines. In media, the local reporter who preserves a fact may matter more than the national columnist who later synthesizes it.

A citation economy can make attention trackable.

When later work depends on an earlier artifact, the system can ask what that artifact did. Did it make a prediction? Did it preserve a source? Did it introduce a distinction? Did it map a field? Did it identify an incentive? Did it refute an error? Did it attend to a weak signal that later became central?

This creates a richer reputation than accuracy alone.

Accuracy matters. Calibration matters. But salience matters before accuracy can even be tested. A perfectly calibrated person who only prices already-legible questions is less valuable than someone who repeatedly notices the questions that will matter tomorrow.

This is why public speakers should eventually have attention track records.

Journalists, CEOs, investors, politicians, consultants, analysts, professors, and institutions all allocate attention. They decide what to cover, what to ignore, what to call serious, what to dismiss, what to amplify, what to bury. These choices shape reality. Yet they rarely face the kind of longitudinal accountability that athletes face after every game.

AI will change that.

Agents can maintain the record. They can compare what someone attended to with what later mattered. They can surface ignored priors, late pivots, failed frames, unacknowledged corrections, and early accurate salience judgments. They can turn “who saw this coming?” from a vibes question into a graph query.

That does not mean the system should create one crude score. A single attention score would become another Goodhart target. The record should be multidimensional: novelty, timeliness, source quality, correction speed, downstream citation, falsification survival, adversarial engagement, and relation to later events.

The point is not humiliation. The point is public memory.

A society without attention track records rewards confidence after the fact. A society with attention track records can reward the people who looked in the right direction before the crowd turned its head.