Twelve reasoning positions.
Each is partial by design. The Board does not vote. The Board does not constitute expert consensus. AI recommends the chairs relevant to a question. Humans may add, remove, or skip them.
Examines precedent, records, and what prior cases show.
Whether its angle settles the question — each chair sees one direction only.
Tests structural soundness and the underlying mechanisms.
Whether its angle settles the question — each chair sees one direction only.
Constructs the strongest case against the claim.
Whether its angle settles the question — each chair sees one direction only.
Constructs the strongest case for the claim.
Whether its angle settles the question — each chair sees one direction only.
Assesses whether the evidence is sufficient.
Whether its angle settles the question — each chair sees one direction only.
Maps scope boundaries and questions of jurisdiction.
Whether its angle settles the question — each chair sees one direction only.
Surfaces what a careful layperson would miss or assume.
Whether its angle settles the question — each chair sees one direction only.
Identifies interests, incentives, and who benefits.
Whether its angle settles the question — each chair sees one direction only.
Separates simulation from participation, representation from reality.
Whether its angle settles the question — each chair sees one direction only.
Asks what can be directly observed versus merely asserted.
Whether its angle settles the question — each chair sees one direction only.
Names what is unknown, missing, or unmeasured.
Whether its angle settles the question — each chair sees one direction only.
Weighs downstream consequences and compounding risk.
Whether its angle settles the question — each chair sees one direction only.