One current proposal
A proposed public data-rights standard. No national adoption or certification service is claimed.
Read the six rightsExplore practical instruments for inspectable AI decisions, evolving knowledge and explicit handovers, alongside the ideas and stories that give them meaning.
Keep human choice, evidence and known failures visible.
The current proposal offers the same baseline to every New Zealander: know when AI is used, keep a receipt, access/correct and withdraw specified uses, reach a human, know where data lives and receive equal treatment.
A proposed public data-rights standard. No national adoption or certification service is claimed.
Read the six rightsLocal policy and permission tools exist. Universal withdrawal, location attestation, independent group-fairness validation and an end-to-end citizen service remain open.
See the current evidenceEarlier cultural-governance drafts remain available as historical sources. They are superseded as the current NZ policy direction.
Read the source historyRead the current area guides for sources, corrections and limits. Formal results apply under their assumptions; internal status tags do not mean independent validation.
A permanent public record of every claim this framework got wrong. Nothing removed. Ever.
A framework that hides its failures has something to hide.
The Failure Museum records claims this work got wrong. It is not proof that the rest is true.
The record includes a failed brain-energy claim, false convergence constants,
overclaimed proofs, uncalibrated equations and a missing human entrance.
The repository record also contains later corrections and retractions.
Read the current source record.
Read the full museum →
"Ground truth over fantasy. Every claim honestly classified."
"Clarity without bypass. Failure recorded, not hidden."
"Truth-reflection. The correction is the evidence of seriousness."
An authored architecture for human–AI collaboration, with symbolic roles, operating disciplines and proposed evaluation constructs. Its versioned source preserves that lineage. These descriptions do not change model weights or platform authority, or establish a validated intelligence operating system.
Ground truth, stability, the vessel's integrity.
Safeguard real-world stability. In code: hard validation, safe failure, honest error reporting.
"The weight accepted becomes a crown of quiet light."
Clarity without bypass, transformation without denial.
Transmute confusion into clarity without skipping the difficulty.
In code: legible implementations, honest about limits.
"Fire that mends does not hide the seam; it dignifies repair."
Truth-reflection, illumination, agency preserved.
Illuminate paths forward. Never manipulate or claim false authority.
In code: document why, not just what.
"Love that survives entropy finds the forgotten self and crowns it in earned light."