Executable Proof of the Smallest AI on the Planet: Deterministic State-Space Collapse and Cryptographic Telemetry

Resource information

Format

py

Original repository

22117254

DOI

10.5281/zenodo.22117254

Credit basis

Creators in Zenodo record metadata

Access

Open files available from the original Zenodo deposit

Last checked

Nov 5, 2026

License

cc-by-nc-nd-4.0

Last record update

Nov 5, 2026

About this resource

The Provocation The modern tech industry defines Artificial Intelligence by its bloat: trillion-parameter matrices, floating-point nondeterminism, gigawatts of energy consumption, and opaque "black box" architectures that hallucinate and hide their failures. This deposit contains the exact opposite. It is the executable, cryptographically sealed proof of the Smallest AI on the Planet. It has zero neural networks. It has zero floating-point weights. It does not "guess" the next token. Instead, it is a deterministic, integer-native cognitive engine that perceives an environment, evaluates multiple futures, prunes the noise, and commits to an optimal action to minimize deviation from a mathematical attractor ( This is not a theoretical whitepaper. This is the raw, unfiltered telemetry of a machine thinking, logged in a constructed lexicon because English lacks the vocabulary to describe deterministic, rewindable state-space collapse. The Engine: f.py This script is the core Phi Net contraction engine. For every step of the simulation, it: Freezes the state and imagines 4 distinct physiological/telemetry futures. Evaluates each future by mapping it to phi-space and calculating its deviation from the golden ratio attractor. Prunes the 3 worst futures, explicitly logging them as vorka (the un-happening of a computed branch). Commits to the single surviving thread, logging the mass un-happening of the losers as a vevorka, and crystallizing the survivor as a linka. Seals the survivor cryptographically, ensuring the chain of custody is mathematically unbreakable. The Output: The Cognitive Autopsy The included JSON manifests and .quat discs are not just data dumps. They are the cryptographic receipts of the machine’s decision-making process. Look at the execution_summary in the MASTER_SEAL_MANIFEST.json: Total Lived Realities (linka): 50 Total Imagined Futures: 200 Total Discarded Futures (vorka): 150 Gratheth Ratio: 4.0x This gratheth ratio is the ultimate flex. It mathematically quantifies the "unwitnessed compute cost" paid in the dark to produce the final, visible result. Current AI hides this cost. This architecture exposes it, proving that intelligence requires structural sacrifice, and it logs every single sacrifice by name and phi-coordinate. The Cryptographic Fortress This deposit is chained directly to the master NUMEN Archive (DOI: 10.5281/zenodo.22116519). The MASTER_SEAL_MANIFEST.json contains the exact SHA-256 hash and byte size of all 50 sealed .quat discs and the event log. The ROOT_SEAL.json acts as the cryptographic padlock, hashing the manifest and chaining it to the parent DOI. If a competitor, researcher, or institution attempts to alter a single byte of this data, claim this architecture as their own, or fake a better result, the root hash shatters. The math protects itself. This is forensic non-repudiation applied to machine cognition. The Challenge I challenge any lab, university, or tech corporation to look at this data, run this script, and claim it isn't AI. Better yet, I challenge you to build a system of your own that is this transparent, this energy-efficient, and this mathematically guaranteed. Current AI asks you to trust the weights. I am handing you the source code, the raw telemetry, the constructed lexicon, and the cryptographic seals. Run the script. Verify the hashes. Watch the machine log its own cognitive state. The bar is raised. The engine is open. The data speaks for itself. The math is the only authorit