Methodology

How Where the Truth Lies Works

Where the Truth Lies is not a fact checker, debate machine, or partisan scorekeeper. ClaimLab helps users think through claims by separating facts, assumptions, interpretation, values, uncertainty, and institutional structure.

Strip the Noise Excavation Reconciliation

1. The Noise Is Stripped Before Analysis Begins

ClaimLab starts by isolating what is actually being claimed before deeper analysis occurs. The original quote remains visible as context, but emotionally loaded rhetoric, implied conclusions, surrounding political narratives, and assumption stacking do not automatically become accepted premises inside the reasoning process.

This matters because hidden assumptions can distort every downstream layer of analysis. The purpose is not to rewrite claims. The purpose is to separate observation, interpretation, assumptions, emotional framing, and uncertainty so the system stays anchored to the actual claim being examined.

Direct Facts

The facts most directly tied to whether the claim holds, fails, or remains unresolved.

Adjacent Facts

Important context that informs the dispute without replacing the core claim.

Root Concern

The underlying civic, institutional, or practical issue giving the claim weight.

Values Divergence

The deeper priority conflict beneath the disagreement, such as liberty versus order, equality versus autonomy, or certainty versus caution.

2. Different Claims Require Different Reasoning Structures

Core Logic 2.0 recognizes that not all claims should be analyzed the same way. Constitutional claims are not treated like scientific claims. Economic claims are not treated like moral claims. Historical comparisons are not treated like predictive forecasts.

The system attempts to identify what kind of disagreement is occurring, what standards of evidence apply, what assumptions are being made, and where uncertainty genuinely exists.

For constitutional claims, ClaimLab separates categories such as citizen, person, the people, and the accused. It also distinguishes procedural protections from substantive rights and government limits from political participation.

3. Multiple Models Perform Different Roles

Different models are used for different reasoning tasks. Claude performs the primary excavation. Grok helps ground live or current context where needed. OpenAI challenges the reasoning by searching for drift, unsupported leaps, hidden assumptions, missing objections, and verdict overreach.

The reconciliation layer does not function like a vote. Its purpose is to surface meaningful disagreement, expose reasoning divergence, identify uncertainty, and reduce single-path reasoning failure.

The system is designed around the idea that imperfect systems become more useful when their assumptions, disagreements, and limitations are visible instead of hidden.

4. Human Review Matters

Where the Truth Lies is designed around structured human review, not blind automation. Claims can be disputed, challenged, revised, corrected, and improved over time.

AI may recommend changes, but important updates require human judgment, approval, and traceable revision history. The platform is designed to preserve reasoning history instead of silently rewriting conclusions.

5. What This Platform Is Trying To Solve

Public argument often collapses facts, assumptions, emotional reactions, historical analogies, legal categories, values, and institutional trust into immediate tribal conclusions. ClaimLab slows that process down.

The purpose is not to provide perfect certainty or replace human thought. The purpose is to help users think more structurally, identify hidden assumptions, understand where disagreement actually lives, explore uncertainty honestly, and navigate difficult questions more transparently.

Truth is often hidden beneath noise, rhetoric, incomplete information, and competing interpretations. Where the Truth Lies exists to help users explore where that truth may actually lie.

Closed Beta Access

Where the Truth Lies

Reasoning Made Visible

Log in with your username or email address if you already have access.

Forgot password?
Need access?

Join the waitlist for future access drops, release notes, and platform updates.

Close
Access Code

Enter your code

If you received a founding or closed beta access code, enter it below to continue signup.

Cancel
Original Quote — Exactly As Submitted
ClaimLab™ Claim Framing
This claim may be broad or interpreted in different ways. Add context if it helps clarify what is actually being claimed.
We’re reading your claim as
Add context if it helps
Similar Claims Found
This claim may already exist
These claims cover similar ground. Open one or continue creating yours.
Trending on X
Full Post
Claim Options
Full View

Unlock Full View

Challenge part of this excavation
Pick the section or sections you believe are wrong, incomplete, or unfairly framed. Be specific. If you have a source, include it.
Standard users get one dispute round before escalation to a human editor. Paid users get up to three pushback rounds before escalation. Human editor review may take 24 to 72 hours.
AI Recommended Update
Review the proposed changes before applying them to the live claim.
Generating preview...
Resolve this dispute?
Choose fields to update
Choose reanalysis type
Restore / Revert
Breakout Claim
Add context for this excavation (optional)
Excavating this claim will use 1 of your monthly claim allowances.
Help us improve Where the Truth Lies
Report bugs, broken claim behavior, confusing results, or suggest a feature or change.
SOURCE
Focusing on the claim. Stripping the noise.
The claim
Additional context (optional)