Pythology · Manawatū, New Zealand
Pythology is building systems that watch complicated parts of the world, keep evidence and inference separate, test explicit causal mechanisms and interventions, say what should happen next — and keep the record when reality answers.
Proof before jargon
The research matters, but nobody should have to understand the architecture before they can see the product. These are live public-safe outputs from the same systems running underneath EarthNet and Prometheus.
Prometheus · a forecast on the record
Prometheus commits forecasts before the answer is known. He records confidence, evidence and a competing outcome, then comes back later and attaches what actually happened. Wins and misses stay in the same ledger.
New Zealand · daily intelligence
New Zealand is EarthNet's current country-level proving ground. The daily state combines seismic activity, volcanic status, incoming weather and ocean context, then adds a short interpretation without pretending every unusual signal is a prediction.
Loading the latest New Zealand interpretation…
A compact national view from the latest public evidence available to EarthNet.
Loading volcanic pulse…
EarthNet · planetary intelligence
EarthNet watches meaningful change across environmental hazards, seismic activity, severe weather, infrastructure pressure, humanitarian context and other connected systems. It can keep a planetary view or narrow attention around a country, region or operating area as suitable evidence sources are available.
Pythology-built infrastructure · Causal Rust Core
Pythology's founder built the Causal Rust Core as deterministic, versioned infrastructure for explicit causal reasoning. It is integrated into EarthNet and Atlas, and supports Prometheus by carrying declared mechanisms, intervention branches and a preserved reasoning trace into the forecasting pathway.
Potential causes and effects are represented as explicit, evidence-linked graphs rather than hidden inside a confident explanation.
The core can hold a chosen variable fixed and propagate the consequences downstream, separating an intervention from ordinary observation.
Each result carries its mechanism path, evidence references and engine provenance so a human can inspect what drove the conclusion.
Underneath the useful bit
The architecture is easier to understand when each piece gets one question. The machinery can be sophisticated underneath without forcing the person using it to speak machine.
Bring the outside world into a current evidence state and track meaningful change.
Keep source, age, missing information, contradictions and modelled estimates visible.
Run explicit, evidence-linked mechanisms and intervention branches through a deterministic Rust core, with the reasoning trace preserved.
Commit a forecast before the outcome, preserve uncertainty and return later to score it.
Compare conditional branches from the same evidence without pretending a scenario is prophecy.
Interrogate evidence, reasoning, degraded sources, uncertainty and weak assumptions.
Let later evidence decide whether confidence should rise, fall or stay exactly where it is.
Where this could matter
Different organisations care about different evidence, but many share the same problem: too much information arrives separately, important changes are easy to miss, and the reasoning behind a decision is difficult to audit afterward.
Country or portfolio-focused intelligence can connect hazards, evolving evidence and likely downstream pressure while keeping the assumptions inspectable.
Useful when the question is not simply “is there an alert?” but “what changed, what could it affect, and what are we still unsure about?”
Energy, transport, communications and other operators can focus the architecture around assets, regions and dependencies relevant to their decisions.
Prometheus and the causal layer provide a disciplined way to preserve evidence, alternatives, forecasts and later outcomes for review.
Long-horizon research
Bio Symbology, Physical Intelligence and the evolution of EarthNet are where we intend to push the architecture next — not as promises, but as serious research directions worth building toward.
Why Pythology exists
It started with one stubborn question: what is actually causing this?
I did not start Pythology because I had a grand plan to build an intelligence company. I kept building things, asking harder questions, and running into the same problem: plenty of systems could show me an answer, but very few could show me why they believed it — or admit later that they were wrong.
Pythology grew from trying to build that discipline into the software itself.
Research, investment or deployment
We are interested in organisations that need to understand complex evidence sooner, test forecasts prospectively, or build a country, region or system-specific intelligence view around real decisions.
.jpeg)