Testing Cohesion, Truthfulness, Quantification
Champions run real editorial and civic material through the three analysis modes and report where a score reads true, where it reads wrong, and why.
Early Testers
Pilot Champions — Early Testers of Dimensional Intelligence
Pilot Champions are the individuals and organisations who put HEUREKA to work on real material. They test the human-facing cognition tools and the machine-facing validation logic in the contexts those tools are meant to serve — newsrooms, classrooms, research groups, civic bodies, and safety teams — and their findings shape what the system becomes.
Human cognition ↔ Machine alignment
Role in the Misinformation Challenge
The human-facing side. Champions work with the cognition tools directly and report on whether they help a person see structure that was previously invisible.
Using the Hexagonal Cognition Engine
Reading and writing direction, analytical ↔ conceptual balance, and language settings are exercised against native working habits rather than lab defaults.
Exploring Categorised 3D Vocabulary
Champions walk the category tree, challenge placements, and surface the terms their own domain uses that the vocabulary has not yet absorbed.
Feedback on Alignment Gameplay
Spins, reels, colour tiers, and ranking behaviour are judged on whether they make a judgement legible — not on whether they look impressive.
Role in the AI Credibility Challenge
The machine-facing side. Champions test whether a result holds when someone else tries to reproduce it.
Testing machine-facing validation modules
Validation runs are executed against live systems and adversarial inputs, with failures logged as first-class results rather than exceptions.
Evaluating reproducibility and traceability
Every score must be re-runnable and every claim followable back to a source. Champions verify that both hold outside HEUREKA's own environment.
Feedback on agentic alignment
Where an agent acts on a judgement, champions assess whether the action matches the stated reasoning and the declared scope.
Who Becomes a Pilot Champion?
No single profile. What they share is exposure to material where getting it wrong matters.
Journalists
Newsroom verification under deadline — the hardest test of any credibility tool.
Educators
Teaching structure, evidence, and framing to students who will meet synthetic media daily.
Researchers
Methodological scrutiny of scoring, sampling, and reproducibility claims.
Civic organisations
Public-interest material where scope collapse and framing carry real consequence.
Fact-checking networks
IFCN-aligned practice, ratings discipline, and evidence-trail rigour.
AI safety groups
Adversarial evaluation, red-teaming, and alignment review of the machine-facing layer.
How Pilot Champions Contribute
Real-world testing
Live material, real deadlines, real stakes. Nothing about HEUREKA is validated on synthetic examples alone.
Feedback loops
Short, structured cycles between a champion's observation and a change in the engines or the vocabulary.
Reporting insights
Written findings that become part of the public record of how the system behaves under pressure.
Shaping validation standards
The reproducibility and traceability rules HEUREKA holds itself to are written with champions, not handed to them.
Become a Pilot Champion
Bring your own material, run it through the machine, and tell us what breaks. The pilot programme is open to individuals and organisations working on evidence, credibility, and public understanding.