UBT
Public Lab Door

The Trigon Challenge

Bring the idea. Choose its lane. Choose the depth. Run the test honestly. Preserve what worked, what failed, and what remained uncertain.

Participant firstAI assistsEvidence decides
◉Lifemind • people • meaningChoose lane ⌬Nature / Sciencebiology • ecology • physicsChoose lane ◇TechAI • systems • codeChoose lane ↑Submitshare the completed testReview first
Choose a Lane Choose Test Depth Open the Prompt Scan the Public Record
Step 1 · Choose the lane

Where does the question live?

Pick the lane that best matches the subject being tested. This organizes the run. It does not decide the result, raise the claim level, or prevent cross-lane evidence.

Your Trigon setupSelect one lane and one test depth. You can change either before copying the prompt.
Lane: not selectedDepth: not selected
01
Blue lane

Life

For people, behaviour, psychology, consciousness, relationships, experience, values, culture, and meaning.

Explore this lane
  • Human behaviour and decision-making
  • Mental and emotional systems
  • Consciousness and lived experience
  • Relationships, culture, and social systems
  • Ethical or interpretive frameworks
02
Green lane

Nature / Science

For living systems, ecosystems, physics, astronomy, biology, measurement, evidence, geometry, and natural processes.

Explore this lane
  • Biology, ecology, and environmental systems
  • Physics, astronomy, and cosmology
  • Geometry and mathematical models
  • Measurement, evidence, and uncertainty
  • Natural cycles, transitions, and structures
03
Red lane

Tech

For artificial intelligence, software, automation, digital infrastructure, engineering, and computational behaviour.

Explore this lane
  • AI models, agents, and evaluations
  • Software architecture and code
  • Automation, robotics, and digital tools
  • Technical safety and failure testing
  • Engineered networks and complex systems
Step 2 · Choose test depth

How far should this run go?

Light, Medium, and Heavy describe process depth. They do not rank the importance of an idea, and a Heavy test never guarantees a stronger conclusion.

△

Light Test

A clean first pass for clarity, overlap, obvious failure points, and readiness for deeper work.

What it includes
  • Public-record scan
  • Question and framework definition
  • Small scenario set
  • Ambiguity, null, and AI-addition notes
△△

Medium Test

A structured test with sources, controls, cross-checks, locked variables, and reusable result data.

What it includes
  • Archive scan and geometry preflight
  • Locked variables and conditions
  • Baseline, null, and negative controls
  • Downloadable data when rights allow
△△△

Heavy Test

A full research lane for preregistration, data lock, failure gates, robustness checks, and reproducibility.

What it includes
  • Preregistered hypothesis and failure criteria
  • Frozen data and analysis choices
  • Robustness, sensitivity, and provenance checks
  • Code, data, checksums, and Record-P when cleared
Step 3 · Run the challenge

Simple on the surface. Expandable when needed.

The participant runs the first pass. AI may help organize, question, and record the test, but it does not become the judge and it may not silently repair the framework.

1Bring the idea

Framework, model, diagram, theory, system, or focused question.

2Start clean

Use a fresh or temporary AI chat when separation from prior context matters.

3Scan the record

Check the public record before treating the work as new.

4Lock the setup

Define the lane, depth, variables, controls, and failure conditions.

5Run scenarios

Test the framework against cases, prompts, files, or situations.

6Record honestly

Keep success, ambiguity, nulls, failures, and AI-added meaning visible.

What the challenge is

The Trigon Challenge is a participant-led pressure test. Your framework remains yours. The record separates original material, participant decisions, source material, and AI-added scaffolding.

AI can assist, but it does not judge. The participant runs the first pass.
What counts as a useful result?

A useful result may reveal consistency, ambiguity, a weak spot, a better next version, a reason to rebuild, or a need for clearer definitions. Failure can be more informative than vague success.

Privacy and separation

Use a fresh chat, Temporary Chat, or Memory Off when the platform provides it. Do not upload private, sensitive, or unpublished material unless you understand the platform settings and release implications.

Public guide paper

Open public testing uses UBT for the World (v2), DOI 10.5281/zenodo.19421797.

Open the public guide
Step 4 · Use the testing prompt

Copy the setup into a fresh AI chat.

Your selected lane and depth are added to the top. The original testing rules remain intact.

TRIGON TEST SETUP
Lane: [SELECT A LANE]
Depth: [SELECT A TEST DEPTH]

You are helping me run a Trigon Challenge test.

You are helping me organize the test, record results, flag ambiguity, and check consistency.

You are not the judge.
You are not allowed to invent missing rules.
You are not allowed to silently improve my framework.
You are not allowed to hide failures.

My framework is:
[PASTE FRAMEWORK NAME]

My framework description is:
[PASTE DESCRIPTION]

My framework parts are:
[PASTE PARTS, DOMAINS, REGIONS, VARIABLES, OR LAYERS]

My test question is:
[PASTE TEST QUESTION]

The scenarios I want to test are:
[PASTE SCENARIOS]

Important rules:
1. Do not rename my framework parts unless I ask.
2. Do not invent meanings for unnamed parts.
3. Do not treat ambiguity as success.
4. Do not hide failures.
5. Do not claim this proves my framework.
6. Do not claim this proves UBT.
7. Record uncertainty honestly.
8. Ask only necessary clarifying questions.

For each scenario, create a result row with:
- scenario ID
- scenario summary
- activated framework parts
- overlap or interaction
- central or integrating region activation, if relevant
- confidence from 0 to 100
- ambiguity notes
- failure or null notes
- whether extra meaning had to be added

Before we begin, restate the task neutrally and confirm that you will preserve ambiguity, failure, and uncertainty.
Step 5 · Review and submit

Package the result without polishing away the signal.

Include enough information for someone else to understand what was tested and where the limits are.

  • Framework name and public name preference
  • Selected lane, depth, question, scenarios, and controls
  • Raw results, uncertainty, ambiguity, nulls, and failures
  • Sources, data files, permissions, and provenance
  • Permission level for public summary or archive release
Submission options Public Record Read the guide
↑Submitreview first
Public Record Archive

Scan what already exists.

Search released Trigon records before beginning new work. The archive preserves summaries, source trails, methods, evidence, uncertainty, failures, and approved downloads without repeating the testing instructions above.

Released public records

Only material that passes source, privacy, provenance, claim-boundary, and permission review appears here.

Loading public records…
Archive resultsNo released records
Truth boundary: exploratory, failed, null, ambiguous, blocked, and superseded records remain visibly labelled. A clean interface never upgrades the evidence.