# TruthExchange: Decentralized Wikipedia for Politics

**$TX — Infrastructure for Reasoning**

_Draft v0.4 — February 2026_

_Quis custodiet ipsos custodes?_

---

## I. The Problem

The political system isn't broken. It's working exactly as designed — for the people who maintain it.

There is a gap between a system's stated rules and its execution. Representative democracy promises that citizens surrender political power to elected officials who represent their interests. The implicit contract assumes accountability: if representatives fail, voters replace them. But when claims are costless, predictions carry no consequences, and actors engineer the information environment for confusion — accountability becomes theater. The system in motion remains in motion.

The system capitalizes on this implicit absolution of responsibility to function. Some systems rely on the ignorance of their users in order to flourish. The current political information ecosystem is both.

### Epistemological Fragmentation: The Core Weapon

The deepest problem isn't necessarily polarization. It's **epistemological fragmentation** — the deliberate fracturing of shared frameworks for evaluating truth.

This isn't an accident. Coordinated minorities win by fragmenting the majority. When a distributed public can't agree on _how to evaluate evidence_ — let alone what the evidence says — they cannot coordinate effectively. Institutions that benefit from the status quo have every incentive to maintain this fragmentation. Not through conspiracy, but through structural incentives: engagement algorithms that reward outrage over analysis, media business models that monetize tribal identity, and political incentive structures that punish compromise.

The same companies and algorithms that have captured enough mindshare to become technocratic nation states are being leveraged to minimize cooperative, productive discussion. Platforms with more daily active users than most countries are optimized for engagement — and engagement maximization is structurally opposed to cooperative discourse.

The result: modern political conversation has devolved into caricatures of productive policy discussion. Debates devolve into character attacks and capitalize on the reader's lack of context to engender ideologically driven hatred.

We are being made dumber. On purpose. Because dumber is more profitable.

### Word Theater: Costless Claims Without Accountability

Within this fragmented landscape, the information ecosystem runs on what we call **word theater** — performative speech disconnected from consequences.

Expert predictions carry no stake in accuracy. Journalists face no penalty for wrong forecasts. Politicians make promises without binding mechanisms. Academics publish unreplicable studies without reputational cost. The entire discourse operates like writing checks on an account with no balance — all commitment, no consequence.

Accountability mechanisms exist in theory: consumers choose their media outlets, politicians can be voted out, peer review exists. But the feedback loops have become divorced from execution reality. The mechanisms persist in form while the substance has been hollowed out.

Prediction markets like Polymarket partially address this by forcing capital behind claims. But they only work for binary empirical questions. They tell you _what_ the crowd believes but nothing about _why_. They can't handle the deeper layer — the reasoning frameworks, axioms, and evidence chains that lead people to different conclusions about the same facts.

### Why Now

Three technological primitives have converged to make this addressable for the first time.

**AI capabilities.** Large language models can extract structured arguments from natural language, conduct multi-turn Socratic dialogue, and surface evidence to help users articulate positions they've never formally examined. Combined with verification layers, they can scaffold human reasoning at scale.

**Blockchain infrastructure.** Immutable records create commitment devices. Smart contracts enable automated accountability. Token mechanics align economic incentives with epistemic quality. Fast chains — Solana for speed and cost, L2s like Base for Ethereum ecosystem access — enable real-time interaction at negligible cost.

**Cultural demand.** Trust in institutions is at historic lows. But it's worth noting: identity-driven politics is a recent phenomenon, not human nature. The Lincoln-Douglas debates, the Federalist Papers, even the televised policy debates of the 1980s were substantive engagements with reasoning and evidence. The current state is an aberration driven by algorithmic incentive structures, not some inevitable feature of democracy. People can engage with structured reasoning — the technological infrastructure just doesn't exist. The demand for substance over spectacle is not just real, but necessary for a democratic social order.

The convergence of AI reasoning, blockchain accountability, and cultural hunger for substance makes TruthExchange possible now.

---

## II. The Insight

You cannot fix political discourse from the top down. You cannot hold institutions accountable if citizens can't first articulate their own reasoning.

### The citizenry needs introspection

**The first step is self-knowledge.** Having individuals understand their own reasoning is necessary — though not sufficient — for rebuilding productive discourse. Before we can bridge ideological divides, we need to test a more fundamental question: can we find enough humans who are interested in mapping their ideologies?

If citizens can't articulate their own reasoning, they become puppets of emotionally charged rhetoric. Reason and logic brought us every major technological and political innovation — from constitutional democracy to the scientific method. The language of the American Constitution — despite its initial exclusion of populations — was to include everyone who could reason about the world. Every subsequent expansion of rights was an argument that said: "your own axioms require you to include us." TruthExchange builds infrastructure for that same tribe: people who believe reasoning is the path to cooperation, regardless of where the reasoning leads them.

But technology has outpaced politics. Our political infrastructure is rooted in a 250-year-old document that, brilliant as it is, could never foresee the scale and effect of information warfare on the populace. The accountability mechanisms designed for a world where information traveled at horseback speed collapse when algorithmic warfare operates at the speed of light.

Every person carries an implicit belief tree — a structure of axioms, evidence, warrants, and conclusions that produces their positions on policy questions. But this tree is invisible. It lives in intuition, tribal affiliation, and inherited assumptions. Most people have never seen it laid out.

TruthExchange makes the invisible visible.

### Why belief trees are effective

When you can see your own belief tree — traced from foundational axioms through evidence chains to policy conclusions — several things become possible. You discover internal contradictions: "I believe in individual liberty AND I support this policy that restricts it." You identify what would change your mind — each conclusion depends on specific evidence, and if that evidence is falsified, the conclusion can/should update. You find unexpected common ground with people whose reasoning diverges from yours at a specific, identifiable point rather than across an unbridgeable ideological chasm.

Think of it as a political horoscope with substance. Everyone wants to understand themselves better. But unlike a horoscope, a belief tree is built from your actual reasoning, responds to evidence, and connects you to others through shared logical structure rather than shared conclusions.

Politics, stripped of theater, is simply the collective decision of what to do with pooled resources. TruthExchange gives individuals the tools to reason about those decisions clearly — and to see where their reasoning genuinely converges with or diverges from others.

---

## III. What $TX Is

$TX is the native token of the TruthExchange protocol — a political reasoning marketplace where beliefs are living structures that update based on new information. Think of it as a decentralized, reimagined Wikipedia for political reasoning — collaboratively maintained, transparently sourced, with built-in dispute resolution and revision history.

### The Foundational Premise

TruthExchange operates from a premise shared by every blockchain ever deployed: **truth exists, and immutable systems are worth building because of it.**

Every block ever mined, every transaction ever validated, every smart contract ever executed rests on the same axiom — that there is a ground truth worth preserving and that systems which make truth tamper-proof are more valuable than systems that don't. The entire crypto ecosystem is a $2 trillion bet that immutability matters because truth matters. Blockchains don't work if "truth" is just a social construct — they work because the ledger either reflects reality or it doesn't.

STEM fields have advanced at extraordinary pace because claims are easy to falsify. Does the bridge hold weight or not? Does the equation predict the observation or not? Does the code compile or not? The scientific method imposes falsifiability on empirical claims, and the result is cumulative progress. Blockchains impose immutability on financial claims, and the result is trustless coordination.

The information economy has no equivalent. Political claims float in a vacuum of accountability, unfalsifiable and consequence-free. $TX extends the same logic that makes blockchains valuable — immutable records, verifiable claims, economic consequences for dishonesty — into the domain of political reasoning. Not by forcing agreement, but by making the structure of disagreement visible, traceable, and anchored to evidence that can be verified.

If you believe an immutable ledger is worth building for financial transactions, the question isn't whether truth exists. The question is why we haven't built one for political reasoning yet.

### The Core Product: Tree Builder

The foundation of TruthExchange is a conversational AI experience that maps your political reasoning into a structured, shareable belief tree.

**How it works:**

You engage in a quest-based conversation with an AI that surfaces politically salient questions derived from current events — not abstract philosophy, but the issues people are actually arguing about right now. The news cycle becomes the content engine: as stories break, the AI presents them and asks what you think.

The user takes a position. The AI drills down: _What evidence supports this? What assumption does that rest on? What would change your mind?_ Through multiple sessions, the AI constructs your belief tree — a structured graph from foundational axioms through evidence and warrants to policy conclusions.

The AI isn't just recording answers — it's actively surfacing evidence from its training data to help users substantiate their worldviews. It presents relevant studies, historical precedents, and philosophical frameworks that relate to the user's stated positions. The question of whether that evidence is authentic and accurately represented — that's what the verification layer addresses.

This tree is yours. It's a mirror of your reasoning. It's shareable, comparable, and — most importantly — it updates as new evidence enters the world.

### The Engagement Loop: Why Users Come Back

A belief tree isn't a one-time quiz. It's a living document that stays relevant because the world keeps moving.

**News-driven re-engagement.** TruthExchange maintains feeds of major news stories as they break. When a story intersects with a node in your belief tree — a Supreme Court ruling that challenges your stance on executive power, an economic report that tests your assumptions about trade policy — the system surfaces it. Not as a notification to doom-scroll, but as a prompt: _does this change your reasoning?_ The news cycle that currently fragments attention becomes the engine that deepens it.

**User-submitted claims.** Users don't just respond to AI-generated questions — they submit their own claims in response to events. "The Fed rate decision proves monetary policy is politically captured." "This trade deal validates industrial policy." These user-generated positions enter the ecosystem and are surfaced to other users whose belief trees intersect with the same evidence or axioms. You don't just see _that_ someone disagrees with you — you see _exactly where_ their reasoning diverges from yours and _why_.

**Social discovery.** When you update a node based on new evidence, users who share that node see the update. When someone challenges a claim you've staked, you're drawn back to defend or concede. The daily news cycle creates a perpetual stream of reasons to revisit, refine, and engage — turning a static belief map into a living, evolving representation of how you think.

The result: every major news event becomes a reason to open TruthExchange, not Twitter.

### Building Blocks: Evidence, CEWI, and Belief Trees

TruthExchange is built on three composable primitives, from atomic to compound:

**The Evidence Node (Atom).** The irreducible unit. A piece of evidence exists independently of any argument that references it: a study, a dataset, a court ruling, a legislative text. Evidence nodes are globally shared across the entire platform — the same piece of evidence can support claims in thousands of different belief trees. Each evidence node is linked to its source entity (person or organization) and carries verification status and confidence scores.

For v0, evidence verification is bounded and automated: a scraper fetches the source URL, validates the document exists and is accessible, archives a snapshot, and an LLM checks whether the excerpt is relevant to the claim it's attached to. More nuanced evidence evaluation — community review, salience scoring, methodological assessment — is built in later phases as evidence nodes accumulate usage across claims.

Some evidence sources present unique challenges. Social media posts (X, Truth Social) are subject to deletion and require archival at the moment of submission. Paywalled articles require special handling. These constraints are acknowledged as active design challenges for v0.

**The CEWI Node (Molecule).** A structured argument composed of evidence atoms bonded by reasoning. The CEWI framework — adapted from competitive debate methodology — gives every argument a consistent, traceable structure:

- **Claim**: A position on a political question
- **Evidence**: Pointers to evidence atoms that support the claim
- **Warrant**: The logical bridge connecting evidence to claim
- **Impact**: Why it matters and what follows
- **Falsification**: A pointer to the inverse CEWI node — the specific argument that, if true, would break this reasoning chain. Every claim contains its structural opposite, creating a graph where falsification is a traversable relationship, not a string someone typed.

This framework handles all forms of political reasoning — including emotional and aesthetic arguments. An appeal to emotion ("This terrible thing happened, therefore we must prevent it through policy X") is still a reasoning chain: the axiom is "preventing suffering justifies broad policy," the evidence is the specific event, and the warrant connecting them can be examined and challenged. The CEWI framework doesn't reject emotional reasoning — it makes its structure visible so it can be evaluated on its actual merits.

The LLM generates CEWI structures from the user's natural language responses and reflects them back for approval and editing. Users always have final say over their own reasoning.

**The Belief Tree (Compound).** Multiple CEWI molecules connected by logical dependencies into a coherent worldview. The tree is organized like a filesystem — topic domains (economics, social policy, civil liberties) serve as navigable folders, with individual CEWI chains as the contents. Users can drill into any branch, explore the axioms underlying their policy positions, and share specific slices without exposing the whole tree.

For sharing, the tree renders as a visual graph — the "poster" that gets shared on social media. For daily use, the filesystem navigation provides scannable, collapsible depth. Users are free to build trees as deep as they'd like, with daily session limits to encourage return engagement.

---

## IV. Token Utility

$TX unlocks the social, economic, and verification layers of TruthExchange:

**Free tier.** Build your belief tree. Full conversational AI experience, shareable output. Always free. The core product has to stand on its own — if the tree builder isn't valuable without a token, the token is worthless too.

**Social tier ($TX required).** Compare your tree with others. Reveal specific nodes in conversations and debates. See aggregate data on how your reasoning clusters with other users. Discuss and challenge specific nodes with other tree holders. See how other users responded to the same news events and where their reasoning diverges from yours.

**Staking.** Users can stake $TX on specific CEWI nodes they're confident in, inviting others to challenge. If a challenge surfaces evidence that falsifies a node in the chain, the challenger earns. If the defense holds, the staker earns. This creates skin-in-the-game at the node level — more granular than any existing prediction market. Polymarket tells you what the crowd believes. $TX staking tells you _why_ — and puts money behind the reasoning, not just the conclusion.

**Verification tier ($TX required).** Submit evidence nodes, verify source authenticity, contribute to the global evidence layer.

**Governance.** $TX holders vote on AI development priorities, verification standards, and protocol parameters.

The free product is genuinely valuable alone. The social layer makes it more engaging. Staking and verification make it economically productive.

---

## V. Evidence Verification

The belief tree builder is the foundation. Evidence verification is where accountability enters the system.

### V0: Automated Verification

For launch, evidence verification is bounded and automated:

1. **Source existence**: Does the document at the submitted URL actually exist?
2. **Archival**: Snapshot the source at the moment of submission (Wayback Machine API, on-chain storage). The archived version becomes the permanent record regardless of what happens to the original — tweets get deleted, articles get edited, pages go behind paywalls.
3. **Excerpt validation**: An LLM checks whether the cited excerpt exists in the source and whether it's relevant to the claim it's attached to.

That's it for v0. No community evaluation, no staking on verification outcomes, no salience scoring. Just: does this source exist, and does it say what you say it says.

### Future: Community Verification

As evidence nodes accumulate usage across thousands of claims, community verification becomes viable. $TX stakers evaluate citation accuracy, methodological quality, and contextual integrity. This is a Phase 2+ feature that emerges naturally once the evidence base has critical mass.

### Source Reputation: The Long Game

Over time, source-level reliability data accumulates naturally across the evidence graph. Domains and entities that consistently publish accurately cited, well-sourced material build visible track records. Those whose citations frequently fail verification — excerpts that don't match claims, sources misrepresented, context stripped — carry that record equally visibly.

This data is anchored on-chain using domains and entity IDs as identifiers. This isn't an editorial judgment about a source's ideology — it's a factual record of whether their citations check out.

The implications of source-level reputation data are significant. We'll let them speak for themselves.

---

## VI. Technical Architecture

### V0: What We're Actually Building

The first version of TruthExchange is straightforward: a conversational AI that helps users build belief trees, with basic evidence archival and a shareable output.

**Conversational AI layer.** A frontier LLM (Claude) conducts Socratic dialogue with users, surfaces relevant evidence, and structures responses into CEWI nodes for user approval. The AI derives questions from current news sources, keeping engagement tied to what's actually happening in the world. No custom RAG pipeline in v0 — the LLM's training data provides sufficient evidence for seeding initial belief trees.

**Evidence archival.** When users submit evidence URLs, the system archives them via the Wayback Machine API, validates accessibility, and stores snapshots. An LLM validates whether the cited excerpt supports the attached claim. Evidence nodes are submitted on-chain (L2) to establish an immutable, timestamped record of cited sources. This process also upserts domain nodes and entity nodes (person or organization) into the global evidence graph.

**Error handling.** V0 acknowledges and handles cases where archival fails — Twitter/X posts, Truth Social content, paywalled articles — with clear user feedback about what can and can't be verified at this stage.

**Tree output.** A shareable belief tree visualization (graph rendering for social sharing, filesystem navigation for daily use) that users can post, link, and compare.

**AI verification layer.** A working prototype (codenamed Sentinel) for cross-checking LLM reasoning against source material has already been built and tested against graduate-level academic texts. In v0, this ensures the conversational AI faithfully represents evidence when surfacing it to users. In later phases, Sentinel becomes critical infrastructure for verifying that emergent AI personas reason faithfully from real user cluster data.

### On-Chain Architecture

Not everything belongs on-chain. The architecture separates concerns:

- **On-chain (L2):** Evidence nodes (global, shared, verifiable), entity reputation data, staked claims, tree hashes for verification
- **Off-chain:** Belief trees (personal, evolving, frequently edited), conversation transcripts, user preferences, tree navigation state

Encoding sociopolitical knowledge on-chain requires a recursive data structure that current L1s and L2s aren't fully optimized for. Phase 1 operates off-chain with on-chain anchoring for evidence nodes and staked claims, giving us room to validate the data structures before committing to chain architecture. A dedicated L3 optimized for evidence graphs and claim verification is on the architectural horizon for Phase 3+.

### Privacy

Political beliefs are sensitive data. The LLM processes conversations server-side to build and maintain the tree at runtime, but:

- **PII stripping**: All API calls are stripped of personally identifiable information before transmission.
- **Selective publication**: Users choose which nodes to make public. Share your economic reasoning while keeping other views private.
- **User control**: Your tree is yours. Nothing is published or shared without explicit action.

---

## VII. Roadmap

### Phase 1: Belief Tree Builder + $TX Launch (Now)

**What we're building:**

- Conversational AI belief tree builder (web app) with news-driven questions
- Shareable belief tree visualizations (graph for social, filesystem for navigation)
- Basic evidence archival via Wayback Machine API + on-chain (L2) storage
- LLM-based excerpt validation (does the source say what you claim?)
- Domain and entity node creation from submitted evidence
- News feed integration for re-engagement prompts
- $TX token on Solana via Pump.fun (with Base/Clanker as future optionality)

**Validation milestones:**

- 100 completed belief trees
- 50% completion rate on tree-building quests
- 30% 7-day return rate
- Organic social sharing of belief trees

**Budget:** Minimal — API credits, hosting, token launch liquidity

### Phase 2: Social Layer + Evidence Marketplace (Month 3-6)

**What we're building:**

- $TX-gated social features: tree comparison, node-level discussion, selective reveal
- User-submitted claims surfaced to users with intersecting belief trees
- $TX staking on CEWI nodes (challenge and defend mechanics)
- Community evidence verification (source authentication, citation accuracy)
- Source and entity reputation accumulation

**Validation milestones:**

- 1,000+ active tree holders
- Active staking on CEWI nodes demonstrating skin-in-the-game engagement
- Measurable tree update rate (users engaging with new evidence)
- $TX utility demonstrated through social, staking, and verification usage

### Phase 3: Emergent Personas + Convergence (Month 6-12)

As users build belief trees, natural clusters emerge — standard ML territory (k-nearest neighbors, DBSCAN, hierarchical clustering) applied to graph-structured data. These clusters won't map to existing political labels. They'll reveal the actual structure of political reasoning across axiom similarity, evidence weighting patterns, and topological structure.

Once clusters reach critical mass, TruthExchange trains AI personas that represent them — not from canonical texts curated by the platform, but from the actual reasoning patterns of real user clusters. The Sentinel verification layer ensures these personas reason faithfully from cluster data. These personas articulate the cluster's strongest arguments, engage in cross-cluster dialogue, and evolve as users update their own trees.

The most valuable output: discovering where different clusters agree despite different axioms. Two clusters that disagree on healthcare policy might share the premise that "the current system has unacceptable inefficiencies." Clusters with opposing economic philosophies might converge on specific regulatory reforms. These convergence points are where real political coalition-building becomes possible.

**What we're building:**

- User clustering algorithms on belief tree topology
- AI personas trained on real user cluster reasoning (Sentinel-verified)
- Cross-persona dialogue system
- Convergence discovery engine
- Institutional API for policy analysis and research
- L3 architecture exploration for optimized evidence graph storage

**Validation milestones:**

- 5+ organic clusters identified with distinct reasoning patterns
- AI personas validated by cluster members as representative
- Convergence points discovered and surfaced to users
- First institutional partnerships (think tanks, research universities)

### Phase 4: Protocol Maturity (Year 2+)

- Open protocol for third-party belief tree applications
- Dedicated L3 for sociopolitical knowledge graphs (if validated)
- Cross-chain deployment (Base, additional L2s)
- Governance transition to token holder community
- Mobile applications
- Advanced analytics and research tools

---

## VIII. Why Me

This project sits at the intersection of three domains I've spent years in.

**Engineering and crypto execution.** My first engineering role was at a startup acquired by Time Magazine, where I built complex frontend systems including undo/redo/time-travel functionality on a multipage website editor — state management challenges directly analogous to belief tree architecture. I've been in crypto since 2017, leaving Uber within a year to build an NFT marketplace. I spent 3.5 years at Coinbase and won the 2024 Coinbase internal hackathon. I know how to ship production systems and I know the crypto landscape from the inside.

**Argumentation and epistemology.** Six years of competitive debate gave me the CEWI framework that structures belief trees — it comes from how formal argumentation actually works, not from a textbook. I'm currently pursuing graduate studies in theology and philosophy, with focus on epistemology and democratic theory. The question of how humans reason about truth, evaluate evidence, and reach political conclusions isn't a side interest — it's my academic discipline.

**AI verification (Sentinel prototype).** The core technical risk of this project — "can you trust the AI's reasoning?" — already has a working mitigation. I built a prototype that cross-checks LLM outputs against source texts, tested against graduate-level academic material. This becomes the verification backbone for evidence validation in Phase 1 and persona fidelity in Phase 3.

**Skin in the game.** I'm building TruthExchange because I want to use it. As someone who moves between secular tech culture, graduate philosophy, and crypto communities, I experience epistemological fragmentation daily. I'm building the tool I need to map my own reasoning and find common ground across communities that don't normally talk to each other.

---

## IX. Token Economics

### $TX Token

- **Chain:** Solana (Pump.fun launch), with Base deployment as future optionality
- **Total supply:** 1,000,000,000 $TX
- **Team allocation:** 20-30% (held throughout program)
- **Bonding curve:** Available via Pump.fun AMM
- **Protocol treasury:** Development, operations, evidence marketplace seeding

### Revenue Model

- **Free tier:** Belief tree building — core experience, always free
- **Social features:** $TX required for tree comparison, node discussion, selective sharing
- **Staking:** $TX staked on CEWI nodes creates economic activity at the reasoning level
- **Verification marketplace:** Protocol fee on evidence verification transactions (Phase 2+)
- **Institutional API:** B2B access for policy organizations and research institutions (Phase 3+)

### Value Accrual

$TX value is tied to protocol usage:

- More users building trees → more demand for social features → more $TX demand
- Active node staking → economic activity at the reasoning layer
- Verification marketplace → transaction fees → revenue to $TX holders
- Institutional adoption → API revenue → protocol treasury growth
- Emergent persona quality → attracts more users → flywheel

---

## X. Risks & Mitigations

### Technical Risks

**AI reasoning quality.** The conversational AI may not consistently extract well-structured CEWI nodes from natural language, especially on nuanced political questions. _Mitigation:_ Users approve and edit all generated CEWI structures. The AI scaffolds; the human decides. Sentinel verification layer (Phase 3) ensures emergent persona fidelity.

**On-chain architecture complexity.** Encoding recursive sociopolitical knowledge structures on-chain is a genuinely hard problem. _Mitigation:_ Phase 1 operates off-chain with on-chain anchoring for evidence and stakes only. L3 architecture exploration begins in Phase 3, after the data structures have been validated through usage.

**Evidence archival limitations.** Some primary sources (social media posts, paywalled content) resist automated archival. _Mitigation:_ V0 handles these cases with clear error messaging. Archival at moment of submission prevents evidence disappearance for accessible sources.

### Market Risks

**Insufficient initial demand.** The target audience — people who want to systematically map their political reasoning — may be smaller than hoped. _Mitigation:_ The "political horoscope" positioning targets self-knowledge seekers broadly. News-driven questions tie engagement to the daily cycle. Shareable trees create social proof loops. Phase 1 milestones are set conservatively (100 trees) to test demand honestly.

**Identity-driven politics as headwind.** Current political engagement is heavily identity-driven rather than reasoning-driven. _Mitigation:_ Identity-driven politics is a recent phenomenon, not human nature — a product of algorithmic incentive structures, not an inevitable feature of democracy. Substantive policy debate was the norm within living memory. TruthExchange provides infrastructure for the mode of engagement people are hungry to return to.

### Social Risks

**Structured reasoning could make divisive positions more articulable.** A well-structured belief tree is more persuasive than an incoherent rant, regardless of the position it articulates. _Mitigation:_ The verification layer exists precisely to ensure that well-structured arguments face scrutiny. Structure without accountability is rhetoric. Structure with verification is reasoning. The system provides both.

### Validation Milestones

We'll know the thesis is working when: users complete belief trees and return to update them, organic social sharing drives new users without token incentives, and staking activity demonstrates genuine engagement with reasoning accountability.

We'll know the thesis needs adjustment when: completion rates are low despite onboarding optimization, users build trees but never revisit them, or social sharing doesn't produce meaningful conversion.

---

## XI. Conclusion

TruthExchange doesn't claim to solve polarization. It doesn't promise consensus. It doesn't pretend technology alone can fix democratic dysfunction.

What it does: give individuals the tools to understand their own reasoning. Make that reasoning structured, transparent, and falsifiable. Create economic incentives for engaging with evidence. And surface pre-existing common ground that partisan framing — and the institutions that profit from it — have obscured.

The evidence node is the atom — a shared, verifiable, immutable piece of truth. The CEWI structure is the molecule — evidence bonded to reasoning. The belief tree is the compound — a coherent worldview, visible and accountable. $TX aligns economic incentives with epistemic rigor.

We're building infrastructure for a world where political claims have consequences, reasoning is visible, and changing your mind based on evidence is rewarded — not punished.

_Quis custodiet ipsos custodes?_

We do. Together. With receipts.

---

## Appendix A: Data Structures

EvidenceNode {
  id: hash
  content: string
  source_url: string
  archived_url: string          // Wayback Machine / IPFS snapshot
  excerpt: string               // exact quoted text
  source_type: PRIMARY | SECONDARY
  entity_id: hash               // who published this
  verified: boolean             // does source exist and say this?
  confidence_score: float
  referenced_by: CEWINode[]     // every argument using this evidence
  created_at: timestamp
}

Entity {
  id: hash
  type: PERSON | ORGANIZATION
  name: string
  domain: string                // for web sources
  reputation_score: float       // accumulated verification history
}

CEWINode {
  id: hash
  claim: string
  evidence_ids: hash[]          // pointers to EvidenceNode atoms
  warrant: string
  impact: string
  falsification_id: hash        // pointer to inverse CEWINode
  confidence: float
  parent_ids: hash[]
  child_ids: hash[]
  staked_amount: float          // $TX staked on this node
}

BeliefTree {
  id: hash
  owner: wallet_address
  root_nodes: hash[]            // top-level axioms
  nodes: CEWINode[]
  created_at: timestamp
  updated_at: timestamp
  tree_hash: hash               // on-chain verification
}

Evidence atoms are a globally shared pool. When someone verifies a piece of evidence, that verification propagates everywhere it's referenced. When source reputation accumulates on an Entity, it flows upward through every argument that depends on evidence from that entity.

---

**Website:** honestfarming.tech/truthexchange **Twitter/X:** @chungusfarmer **Token:** $TX on Solana (Pump.fun) **Contact:** [e-mail] **GitHub:** github.com/thehonestfarmer

---

_This document is a living draft. It will be updated as the product evolves and assumptions are tested._