Don’t defend the assumption.
Test the assumption.
Our own beliefs, published as research objects rather than positions: a status that will advance, a version history once there is one, and the condition that would defeat each one — written before any evidence exists.
What state the research is actually in.
Every number here is counted from the published tests. There is no status a test has not earned, and no completed test yet.
Three, and all of them unresolved.
A test that has skipped its falsification condition is an essay, and is not published here as a test.
Marriage
Marriage is primarily a matching problem.
We think this is probably false — and we have no evidence, and the counterarguments are unanswered.
Settlement and risk
When a natural event causes damage, the damage is primarily determined by the event.
Hazard is natural; exposure is architectural. Separating the two is the whole analytical content of the test.
Capability & responsibility
Greater technological capability creates a corresponding increase in human responsibility.
Probably true, and unproven. The central hypothesis of The Return to Humanity, stated as a test rather than a conviction.
No tests in that state yet. That is the honest answer rather than an empty promise.
The same fourteen stages, used flexibly.
Not every test needs every stage. Two rules are not negotiable: the falsification condition is written before the experiment, and the counterarguments are made properly before the tests are designed.
- 01
Assumption
Stated so it can be attacked, not so it can be agreed with.
- 02
Why do we believe it
The honest reason, including "because everyone around us does".
- 03
Historical origin
Where the belief came from, and what constraint it was a reasonable answer to.
- 04
Current system
How the thing is actually arranged today.
- 05
Evidence
What is known, with sources. Empty when it is empty.
- 06
Counterarguments
The strongest cases against, made properly — including our own bias.
- 07
The 1850 test
Would we design it this way from scratch, with today’s technology?
- 08
The today test
What does present technology actually change about it?
- 09
The future technology test
What changes at the technological boundary, and what stubbornly does not?
- 10
Alternative architectures
Options compared on real criteria, including reversibility.
- 11
Counterfactual
If this assumption had never existed, what would the system look like?
- 12
Experiment
Specific and falsifiable. Method public; location and operational design are not.
- 13
Result
Published either way. A result against us is the most valuable kind.
- 14
Updated assumption
The next version, and what changed. The history is preserved.
A status has to be able to go backwards.
- Under investigation
- Evidence collected
- Challenged
- Experiment
- Resolved
If a challenge cannot move a supported test back to challenged, the status is decoration. So it can, and the first genuine rejection will be the most valuable thing this section ever publishes.
The counts, the filters and the cards above are generated at build time from the test definitions themselves. There is no separate dashboard that could drift, and no way to publish a count that no test supports. Right now that means: three tests, all under investigation, no evidence collected.