Corrections
What we said, what is true now
Every research program gets things wrong. This one keeps a public record of where, so that a reader can trust the rest. Nothing is deleted; each claim that changed is listed here with what replaced it.
The depth-four count
We said: A depth-four observer's reachable region grows like the fourth power, which explains four-dimensional spacetime volume.
What is true now: The count is correct, but the causal structure inside it is only four steps tall and its ordering fraction goes to zero. That is the opposite of four-dimensional spacetime. The identity stands as a counting result only.
Clocks from observation
We said: Bounded observation forces a clock with a proper time direction.
What is true now: No finite system can carry an exact clock of that kind, and the blank state has no time at all. What survives: any structured state carries a modular flow, which is weaker.
Visible dynamics and light cones
We said: Any dynamics that only touches visible data must spread diffusively, so no light cone can come from it.
What is true now: False in general: a classical shift process touches only visible data and still has a strict cone. The no-go holds only for the program's positive scalar couplings.
A missing signature
We said: Three coupled pairs have the wrong signature for spacetime.
What is true now: A symplectic form has no signature. The correct statement is narrower: three pairs carry no vector of boost generators, and a fourth pair is the minimal repair.
Phase as an axiom
We said: A quantum state must carry complex phases to be the right state.
What is true now: The test was coordinate-dependent: a simple relabelling makes the state real without changing anything physical. It was replaced by a test that does not depend on coordinates.
A driven quantum cell
We said: A driven cell settles into a structured, holographically compatible state.
What is true now: The model used to show this was not a legitimate quantum process. The result was withdrawn and later replaced by a legitimate construction with a weaker conclusion.
Exact solutions at first order
We said: Exact consistent states exist at leading order.
What is true now: An exact calculation showed a small but nonzero floor. The residual never reaches zero once the program's source term is switched on.
What picks the geometry
We said: Two-way couplings decide which observers are admissible, and complementary couplings decide the geometry.
What is true now: The second half was refuted by a control experiment. Consistency favours two-way couplings; it does not pick which one.
Theorem sheet errata
We said: Several entries in the program's theorem sheets, including one statement that read the wrong way round.
What is true now: Corrected, relabelled where the evidence was weaker than stated, and reissued with an errata list.