The record.
Commitment Theory and the Conservation Law, with the experiments, specs and code behind them.
- A Conservation Law for Commitment in LanguageDOI 10.5281/zenodo.20029607
- Experimental RecordDOI 10.5281/zenodo.19105225
- Experiment HarnessDOI 10.5281/zenodo.19109397
- TTEOP Interoperability SpecDOI 10.5281/zenodo.22180349
- SigRank Taxonomy v3.1DOI 10.5281/zenodo.21900519
- SSRN paperDOI 10.2139/ssrn.6734283
- Commitment Theory on GitHubgithub.com/SunrisesIllNeverSee
The ideas behind it
Five concepts, in the order they matter.
Recursive Compression
Every summary, rewrite or handoff transforms a signal. Repeat it and commitment can leak away. This is the T in the Conservation Law.
02Governance Enforcement
Gates, thresholds and lineage checks that run before a transformation. With them, the law says commitment holds. Without them, it degrades.
03Constitutional Substrate
The layer where the rules live: what may change, what must hold, and what gets checked before anything runs.
04Signal Encoding
Writes commitment into the signal itself, so meaning can be measured and verified, not just observed after the fact.
05Origin Binding
Ties every artifact to the cycle that produced it with a cryptographic hash, so copies and mimics can't pass as the original.
GuideHow to verify lineage
A practical walkthrough: check an AI output's provenance chain with SHA-256 hashes and origin binding.