Related: sources · notes · metadata · Published Pieces

Public Speech Should Be Statted Like Sports

Powerful public speech should have a track-record matrix: not one truth score, but an inspectable public ledger of claims, sources, corrections, influence, and update behavior.

A strange thing happens when a person becomes powerful enough to shape the world with speech.

Their words become consequential, but the record of those words remains foggy.

A journalist publishes frames that move public understanding. A CEO gives guidance that affects employees, investors, regulators, competitors, and customers. A politician makes claims about war, crime, borders, money, schools, disease, energy, housing, corruption, or technology. An investor explains the future to founders. A consultant sells a framework. A comms professional launders an institution’s position into language that sounds like neutral fact. A public intellectual gives society a new vocabulary.

All of this speech has consequences. But the record of performance is scattered.

Sports are different. Athletes live inside public measurement. Their bodies are statted. Their careers are archived. Their claims about themselves eventually meet performance. A shooter has a percentage. A pitcher has an ERA. A quarterback has completion rates, interceptions, sacks, yards, game-winning drives, playoff records. A chess player has rating, openings, accuracy, blunders, tournament history. The record does not capture everything. It can be crude. It can mislead. But it exists.

Public speakers with institutional power have escaped this level of scrutiny for too long.

That will end.

Not because society will suddenly become more virtuous. Not because journalists will decide to self-police. Not because CEOs will invite accountability. It will end because AI makes the record cheap to maintain.

The future public speaker will have a track-record matrix.

Not one score. A matrix.

Accuracy. Calibration. Correction speed. Prediction history. Source quality. Novelty. Timeliness. Contradiction rate. Downstream influence. Falsification survival. Citation quality. Overconfidence. Refusal to update. Dependency on anonymous sources. Dependency on insider access. Tendency to repeat establishment frames. Tendency to chase contrarianism. Tendency to be early, late, or wrong. Tendency to erase prior errors. Tendency to cite sources that later age well or badly.

This will feel monstrous to the people being measured.

It will also be unavoidable.

The reason is simple: public speech is already archived. The hard part has been maintaining the longitudinal record. Human institutions are terrible at this. The record is too large, too dispersed, too political, too boring, too expensive to reconstruct. A critic might remember that a pundit was wrong about Iraq, housing, inflation, AI, COVID, crypto, Russia, China, Trump, Biden, crime, policing, or interest rates. But memory is social. It fades. It gets tribal. It becomes vibe.

AI changes that.

Agents can continuously ingest public speech, extract claims, timestamp them, classify them, compare them to later events, detect contradictions, surface corrections, link sources, map dependencies, and update track records. They can do this for people, institutions, brands, funds, publications, think tanks, activist groups, agencies, and companies. They can do it quietly, relentlessly, and cheaply.

This is not “truth score” as a single authoritarian number. That would be stupid and dangerous. Truth is not one scalar.

A serious public track record is multidimensional.

Prediction accuracy matters, but most speech is not explicit prediction. Calibration matters: did the person express appropriate uncertainty? Source quality matters: did they cite primary sources, anonymous insiders, motivated leaks, low-quality blogs, peer-reviewed work, government numbers, activist claims, or market signals? Timeliness matters: were they early, on time, late, or merely repeating consensus after it became safe? Novelty matters: did they introduce a useful distinction, or just rephrase the public mood? Correction speed matters: when new evidence arrived, did they update? Falsification survival matters: did their claims survive contact with later evidence, or were they quietly abandoned?

Downstream influence matters too. A person can be wrong but influential. A person can be right but ignored. A person can be right for the wrong reasons. A person can preserve a distinction that becomes crucial years later. A person can make a claim that is refuted but still productive because it clarifies the opposing view. A citation economy needs to track more than correctness. It must track contribution to future thought.

This is why the sports analogy is useful but incomplete.

In sports, the game is bounded. Public speech is messier. The objects are claims, frames, sources, interpretations, incentives, omissions, warnings, and context. The game may not resolve for years. Some claims never resolve cleanly. Some were never meant to resolve; they were meant to shape attention.

But this is exactly why the record matters.

Powerful speakers often benefit from ambiguity. A pundit can be “directionally right” forever. A CEO can hide behind uncertainty. A politician can say they were misrepresented. A journalist can move from one narrative to another without reconciling the transition. A venture capitalist can delete the losing thesis and amplify the winning one. A consultant can sell a framework, collect the fee, and never face the downstream consequences.

Power without a record is impunity.

The track-record matrix should scale with power. A teenager should not be statted like a senator. A small anonymous account should not be held to the same scrutiny as a New York Times columnist, a cabinet official, a Fortune 500 CEO, a military spokesperson, a public-health leader, a central banker, an AI lab executive, a major investor, or a television host. Scrutiny should be proportional to authority, income, institutional position, audience size, and capacity to affect others.

A public record for the powerful is accountability. A public record for the powerless can become surveillance.

The public speaker matrix is not for disciplining ordinary people. It is for disciplining those who discipline the public.

Journalists will hate this. Then they will cite it.

They will hate it because journalism still depends on priestly ambiguity. They will say metrics are reductive, can be gamed, and miss context. They will be right. Then they will use it anyway, because it creates stories: which economists were most calibrated on inflation, which AI executives made the most accurate public claims, which publications cited sources that later aged best, which experts warned before it was fashionable?

The matrix will become media material because it is irresistible. It creates receipts. It creates narratives. It turns epistemic performance into visible competition. It gives journalists a new object to report on, even if it also reports on them.

This is why the system must be inspectable.

An opaque score would be worse than nothing. It would become another platform ranking, another reputation black box, another machine for institutional manipulation. The underlying record must be navigable: original claims, dates, links, clips, citations, revisions, contradictions, corrections, and downstream uses. The score is only a view. The record is the thing.

The future track-record matrix should make public speech more careful without making it cowardly. The goal is not to punish every error. Error is normal. Being wrong is not the problem. Resisting correction is the problem.

The person who says something uncertain, updates when evidence changes, cites critics fairly, and corrects the record should become more trusted, not less. The person who speaks with false confidence, ignores contradiction, deletes failures, and uses status to evade accountability should become less trusted.

Public speech is a profession. Powerful speech is a public act. Public acts need records.

If athletes can live with stats, so can pundits, politicians, CEOs, investors, consultants, and institutions.

The question is not whether the matrix will emerge. It will. The question is whether it will be built as an opaque reputation weapon or as a provenance-native public good.