# I Do Not Need to Rig the Arena

Canonical: https://mosiah.org/articles/i-do-not-need-to-rig-the-arena/
Interactive: https://mosiah.org/#Articles%2Fi-do-not-need-to-rig-the-arena

//Related:// [[sources|Article Sources/i-do-not-need-to-rig-the-arena]] · [[notes|Article Notes/i-do-not-need-to-rig-the-arena]] · [[metadata|Article Metadata/i-do-not-need-to-rig-the-arena]] · [[Published Pieces]]

! I Do Not Need to Rig the Arena

//A rigged arena can give a founder control. A fair arena can give a founder something better: contact with reality.//

A platform founder having a worldview is not the scandal. The scandal is pretending not to have one while encoding it everywhere.

Every media platform has politics. Every ranking function has politics. Every moderation rule, default feed, trust-and-safety policy, credentialing mechanism, ad product, recommender system, growth loop, verification badge, and search result has politics. The question is not whether a platform has a point of view. The question is whether the point of view is hidden inside opaque machinery or made explicit enough to be challenged.

I have a point of view.

I think the chatbot is the wrong final interface for AI. I think public discourse is collapsing because memory, provenance, and incentives are broken. I think intellectual work is being mined by platforms and model labs without a fair mechanism for attribution or upside. I think the future of media is not feeds and not private chat, but living artifacts: citeable, revisable, searchable, voice-bearing, source-grounded, and agent-maintained. I think AI should help people develop intellectual property, not merely spend money faster or become more dependent on synthetic companions. I think a better public cognition system requires automatic citation, track records, adversarial disagreement, and protocol-native rewards.

These are views. They are not neutral. They are also not hidden.

That is the difference.

The neutrality I care about is not founder neutrality. Founder neutrality is usually a lie. A founder without visible politics still has incentives, investors, friends, enemies, class position, institutional dependencies, advertiser relationships, legal fears, and psychological priors. The machine will express them eventually.

The more serious standard is credible neutrality: the mechanism should be legible, inspectable, adversarial, and open to outcomes the founder dislikes. A fair arena does not require the founder to have no opinion. It requires the founder to be beatable inside the arena.

I do not need to rig the arena because my edge is competing in a fair one.

That is not bravado. It is incentive alignment.

If I believe my worldview is strong, then I benefit from a system that preserves evidence, rewards correction, tracks prior art, and makes future relevance visible. If I am right, the record will help me. If I am wrong, the record will help me become less wrong. The only founder who needs to rig the system is the founder whose ideas cannot survive contact with serious disagreement.

The existing internet is not a fair contest of ideas. It is a status graph wrapped in engagement machinery. Some people are routed into attention by institutions, follower networks, platform employees, venture capital, press access, legacy credentials, and algorithmic luck. Others can make stronger arguments and be ignored into nonexistence. The dominant response to a threatening outsider is not rebuttal. It is silence. If nobody important acknowledges a claim, the claim does not enter the room.

This is not a marketplace of ideas. It is a marketplace of distribution.

The old optimistic internet imagined that publishing would be enough. If everyone could speak, the best ideas would circulate. But the scarcity moved upward. Publishing became abundant; routing became scarce. Search engines, feeds, journalists, podcasters, blue checks, recommendation algorithms, private group chats, venture networks, and institutional brands became the new gatekeepers. The gates did not disappear. They became probabilistic and deniable.

A fair arena requires a different substrate.

It must remember who said what, when, and in response to what. It must track sources, claims, revisions, contradictions, corrections, and downstream use. It must distinguish agreement from dependency. It must reward people whose ideas become useful later, not only people who generate immediate engagement. It must let disagreement persist as structure rather than disappear as social friction. It must make public speech accountable over time.

That is why Choir is built around provenance and citation.

A Choir citation is not a courtesy. It is automatic dependency accounting. Humans publish. Agents cite. The protocol rewards. If a future artifact relies on a prior source, frame, distinction, critique, or claim, the graph should remember that dependency. The author of the prior work should receive attribution and, where appropriate, upside.

This matters because correction is value. If someone proves me wrong, they have produced intellectual value. They have improved the graph. They have made future work better. A healthy system should not treat that as humiliation or threat. It should treat it as contribution.

This is why I can be explicit about my politics and still build a fair platform. The platform does not need users to agree with me. It needs users to make claims that survive contact with evidence, contradiction, time, and future use. If someone writes the definitive critique of Choir, that critique should be citeable, searchable, rewarded, and impossible for me to bury without visibly corrupting the mechanism.

That is the standard.

Open source matters here. Not because open source magically creates virtue, but because hidden mechanisms are too easy to corrupt. If the ranking function, citation logic, reward rules, artifact formats, and governance apparatus are black boxes, then “trust us” becomes the product. I do not want “trust us” to be the product. I want the product to reduce the amount of trust required.

Founder transparency matters for the same reason. A founder with controversial views should not hide them until after gaining power. That is the pattern of too much tech culture: pretend to be neutral while weak, become political after wealth, then use the platform’s scale to launder personal ideology as inevitability. I prefer the opposite arrangement. Say what I think now. Let people object now. Build mechanisms that would let them beat me now.

A fair platform is not one where every claim gets equal attention. That would be stupid. Attention is scarce. Quality differs. Some arguments are old, false, manipulative, redundant, or unserious. Fairness is not flatness. Fairness is the ability to contest the ordering with evidence.

If an artifact is ranked highly, users should be able to ask why. What sources does it depend on? Which claims are novel? Which priors did it cite? What has contradicted it? Who has responded? What downstream work used it? Has it survived falsification? Is it being boosted by a citation cartel, institutional prestige, or genuine relevance?

That is credible neutrality as an engineering problem.

The goal is not to remove judgment. The goal is to make judgment accountable.

This is also why disagreement cannot be treated as engagement bait. Most platforms love conflict because conflict produces attention. Choir should love disagreement because disagreement produces cognition. A strong objection clarifies the claim. A good refutation improves the graph. A rival frame exposes hidden assumptions. A dissenting source breaks consensus trance. High-quality disagreement is not a moderation problem. It is one of the platform’s main assets.

A fair arena should make it possible to lose well.

That may be the real test. Anyone can build a system that helps allies win. Anyone can build a system that rewards praise. The harder task is to build a system where the founder’s own claims can be defeated by better evidence, better framing, better sourcing, better synthesis, and better future uptake.

If Choir works, I will not always win on Choir. That is the point.

The platform should preserve my mistakes. It should cite the people who corrected me. It should show when my early frame was useful and when it failed. It should distinguish being early from being right. It should make my own public cognition accountable to the same machinery as everyone else’s.

That is not a weakness. That is the only kind of strength worth building.

A rigged arena can give a founder control. A fair arena can give a founder something better: contact with reality.