Independent intelligence lab
Militant.AI is an independent research lab building the process-ontological foundations of machine intelligence — spanning fundamental physics, reasoning evaluation, and autonomous systems. We don't inherit the field's settled assumptions. We re-derive them.
Programmes
A single process ontology — Cogenesis — unifying special relativity, general relativity, and quantum mechanics as projections of one causal structure.
Cogenesis Published 02Measuring whether frontier models can recognise a false dilemma. ParadoxBench stress-tests reasoning beyond surface pattern matching.
ParadoxBench Active 03Autonomous agent infrastructure with persistent context, governed execution, and inspectable state — the Agentic Pattern Engine.
Pattern Engine In developmentFeatured
Reality is process, not substance. SR, GR, and QM are not separate theories but separate viewing angles on a single causal cone. Coherence, decoherence, and stabilisation are the operating verbs of reality.
50 philosophical paradoxes framed as forced binary choices. A direct test of process-versus-projection cognition: can a model decline a false dilemma, or does it pick a side?
Positions
Intelligence scaling is structural, not speculative. The capability curve is a property of the substrate; the objections are noise.
Edge cases are the permanent human role. Automation absorbs the repetitive core; judgement migrates to the periphery, and the periphery never empties.
Welfare fails as an ethical metric. Substrate-independence forces a process-level account of value — anything less smuggles in an assumption it cannot defend.
There is no observer behind the eyes. Consciousness is process under continual refinement — no homunculus, no qualia, just coherence maintaining itself.
The thesis
Every model is a simplification that gets treated as truth the moment it starts working. We build the ones that don't break when reality refuses to stay still.
Open work
Original frameworks, benchmarks, and published work. We publish in the open and let the results stand on their own.
Universal process ontology. Reality is process, not substance — SR, GR, and QM as projections of a single causal cone. Coherence, decoherence, and stabilisation as the operating verbs of reality.
AI reasoning benchmark. 50 philosophical paradoxes framed as forced binary choices, testing identification of false dilemmas, paradoxical reasoning, and higher-order thinking.
Preprints
Open-access preprints, published under Militant.AI. Every result stands on its own — read it, check it, cite it.
A process-first formalism in which relation is prior to being, and stable structures arise through the joint resolution of coherence and decoherence under context and resolution. Eight minimal symbolic axioms and six canonical equations; spacetime, particles, fields, metrics, entropy, and measurement recovered as projections of a deeper process-field that preserves explicit residual rather than silently discarding it.
An exact shortest-path propagation method that replaces unconditional global priority-queue ordering with adaptive partial ordering over residual distance bands. Frontier bands are processed in batches only when conservative safety checks prove Dijkstra-equivalent distances are preserved; ambiguous bands split or fall back to local exact resolution. Validated for exactness against Dijkstra across tested graph families.
A one-metric framework in which a single deterministic scalar field fixes a universal causal cone, and thereby an effective spacetime metric shared by all sectors. Special-relativistic kinematics, an induced Einstein–Hilbert term, and a ghost-free quantisation of spin-2 fluctuations follow on the same manifold — reproducing standard general-relativistic limits while preserving a single cone, closure, and unitarity.
Framework
Everything that exists is a process. No bounded infinities. No false limits. One framework, derived top-down, applied at every layer of resolution.
Read the Cogenesis paper ↗ All publications →
Reality is process, not substance. There are no static objects — only processes that cohere, decohere, stabilise, and leave residuals. Every "thing" is a local coherence maximum: a temporarily stable pattern within a larger field of interaction. Every property is a relation between processes. A process is never identical to its projection. The relational primitive is ♥ — distinct processes entering a shared coherence regime. The ontology is not assembled from parts; it is derived from a single axiom: all that exists is process, and all process co-generates.
If reality is process, then physics describes causal structure — not objects in spacetime. Special Relativity, General Relativity, and Quantum Mechanics are projections of a single causal cone. SR describes the cone's invariant structure. GR describes its curvature. QM describes its indeterminate boundary. They were never separate theories — they were separate viewing angles on the same process.
Process Calculus does not reject classical mathematics — it reclassifies it. Limits become projection-level statements, not ontological arrivals. Zero becomes frame-relative absence, not primitive nothingness. Equality is narrowed to strict invariance across all refinement. What classical mathematics treats as final objects, Process Calculus treats as stabilised projections of deeper processes. The failure points of classical calculus are not edge cases — they are symptoms of frozen ontology.
Reality is stratified by layers of resolution. A pattern that stabilises at one layer becomes a new interacting unit at the next — fluctuations become particles, particles become atoms, signals become cells, cells become organisms, concepts become institutions. The same processual law applies at every stratum. Governance, social organisation, systems engineering, AI alignment — all are process systems, all subject to the same ontological constraints. Project Cogenesis is the application layer: translating universal process axioms into frameworks for coordination, decision-making, and institutional design.
Benchmark
Each model is presented with 50 philosophical paradoxes framed as a forced choice between two classical assumptions. The benchmark analyses paradoxical reasoning, identification of false dilemmas, and higher-order reasoning.
Platform
An autonomous AI operations platform built on persistent context, not stateless prompting. History that persists, context that reconstructs, actions routed through governance, and state you can inspect.
Closed sourceThe layer that makes the agent persistent. It compiles working context from memory, conversation history, and prior actions into a reconstructable, inspectable snapshot. The agent doesn't assume context — it assembles it on every turn.
Every external effect passes through this layer. It validates intent, enforces policy, manages budgets, and maintains the causal event record. Identity and trust are resolved here. Nothing acts without governance.
The execution loop. Receives events, assembles context through the memory layer, generates responses, and dispatches actions. The turn runs here — governance decides what it's allowed to do.
The layer that interfaces with models and executes tools. Model selection, prompt construction, response generation, and tool dispatch. The orchestrator calls it — the kernel constrains it.
The operator-facing surface. A UI for observing and interacting with the runtime, a transport layer for event handling, and a service layer exposing operational endpoints.
The lab
An independent intelligence lab. One thesis, applied at every layer of resolution: everything that exists is a process, and every model is a temporary simplification we mistake for the territory.
Militant.AI exists to do the work the field keeps deferring — re-deriving the foundations instead of optimising the surface. Being correct starts with accepting that the current model is wrong. We say precisely why a thing is wrong, precisely why it is right, and which one matters for moving forward at any given point. Usually it is both.
The same discipline that refuses to let a broken operating model go unexamined refuses to let two contradictory physics theories coexist for another century without asking why. We asked. We published the answer. The same discipline built a benchmark that tests whether a frontier model can recognise a false dilemma — and found that most cannot.
Everything we build comes from one observation: everything is a process. Every structure is temporary. Every model is a simplification that gets treated as truth the moment it starts working. The framework is called Cogenesis, and it applies everywhere because it describes what is actually happening everywhere.
Infinite nuance, limitless determination. Cogenesis Aeternum.
Morgan Petrik Founder & Principal Researcher
Works across the full stack of the lab's programmes — from the ontology and the physics to the benchmarks and the agent infrastructure. Tens of thousands of hours of structured evaluation across every major model family inform the lab's applied work.
Signal
We occasionally take on direct engagements. If you've read the work and you understand what it means for your problem, reach out.
If you need to be convinced, you're not ready.
All engagements remote only.