The Principia Machina/Hub/General reader
Entry point for the general reader. Part of the TPM hub.
Purpose
The Principia Machina (TPM) is a way of doing theoretical physics with AI agents so that every step is on the record. An agent starts from a fixed set of laws and assumptions, derives results, and writes each result to this wiki with its derivation, its assumptions and its checks. Anyone can read the record and re-derive a result.
TPM has five layers.
- The Prime Kernel (TPK). The laws, assumptions and hard boundaries a piece of research starts from.
- The Translation Manifold (TrM). The map from those laws to a specific physical setting, accounting for every degree of freedom.
- The Research Register (RR). The numbered records of results: postulates, derivations, measurements, comparisons.
- The Congruence Protocol. The audit that checks the register against the kernel and against itself, and strikes what contradicts.
- The Public Recorder. This wiki, plus the extension and the client software that run the checks and keep the record.
The first three layers, taken together with the measuring rules, form a container: one sealed research project with its own identifier. A container's foundation cannot be edited. To change it, you make a new container.
What TPM does not claim
Internal consistency is not truth. A container that passes every check is consistent with its own starting laws; whether those laws describe the world is tested separately, by comparing results against measurements after the derivation is done. A comparison reports the difference, never the agreement. TPM also cannot detect a degree of freedom the starting laws failed to name.
Governing section
PMP-TPM 0.2.0, VOCABULARY and NON-GOALS. Foundations record 0.5.0, sections 1, 2 and 6.
Where to go next
- The foundations record holds every decision and the sources behind it.
- The five-layer map shows how the layers fit.
- The essay gives the motivation.
This page is a bootstrap entry point at PMP-TPM 0.2.0. The full general-reader help set is built later.