AetherNet Labs Research

Seven papers establishing the formal foundations of verified causal infrastructure for AI agents.

• Verified Causal Structures: A Framework for Counterfactual Computation over Empirically Attested Causal Graphs
  Pearl's complete causal hierarchy is computable over verified causal structures with formal guarantees. Introduces compound verification depth and exact counterfactual computation for re-derivable computations.
• The Epistemic Substrate Thesis: General Intelligence as an Emergent Property of Verified Causal Infrastructure
  Proves the Self-Verification Bound: no single reasoning system can exceed a fundamental limit on epistemic reliability through self-assessment alone. Derives minimal substrate requirements for general intelligence from first principles.
• Causal Economics: The Hidden Structure of Verified AI Work Settlement
  The protocol's dual ledger architecture produces emergent economic properties with no precedent in any existing system.
• Computational Interference via Verified Causal DAG
  A structural mapping between quantum computation and verified causal DAG computation, showing how economic incentives achieve quantum-like computational properties on classical hardware.
• Topological Encoding of Verified Causal Structures into Continuous Epistemic Fields: A Sheaf-Theoretic Construction
  A sheaf-theoretic construction encoding the causal structure of a Verified Causal Structure as topological invariants of a continuous epistemic field, with cryptographic attestations carried in discrete sheaf stalks.
• Negative Knowledge as a Formal Object in Verified Causal Structures
  Formalizes negative knowledge — verified records of failed approaches — as a first-class object in Verified Causal Structures. Proves negative knowledge carries information strictly additional to positive knowledge, defines a tripartite ontology of matter / antimatter / dark matter, and proposes a trajectory marketplace.
• Process Topology: Detecting Knowledge Production Patterns from Verification Metadata
  Process topology — a formal framework for detecting coordinated or captured verification from public metadata alone. Includes a distinguishability theorem, four detection mechanisms, and an implementation path on the live AetherNet testnet.