top of page

Principles

These twenty related principles, and their extensions, serve as the foundation assumptions of our research agenda.

I. First principles: foundations

P.1

Postbiological drift:
Long-lived intelligence tends to externalize, evolving from biology → tools → self-replicators (VNRs).

  • Prediction: artifact ecologies outlast biologies; look for process, not entites.

  • Falsifiers: compelling biosignature civilizations broadcasting widely while artifacts remain absent.

  • Implication: center hypotheses on replication, repair, resource uptake.

P.2

Persistence over presence: Success = replication + repair; contact is optional.

  • Predictions: low public salience; background persistence.

  • Falsifiers: abundant, persistent “beacons” intended for us.

  • Implication: optimize sensing for persistence metrics, not messages.​

P.3

Cost minimization:
Minimize energy/bit and risk/detection.

  • Predictions: tight beams, short duty cycles, camouflage.

  • Falsifiers: continuous high-power emissions without cover.

  • Implication: hunt low signal-to-noise ratio events.

P.4

Local-first gradients:
Replicators exploit nearby energy/material gradients before long-haul signaling.

  • Predictions: edge zones (coasts, industrial belts, ionosphere transitions) show lift.

  • Falsifiers: uniform spatial distribution after controls.

  • Implication: build a counterfactual siting atlas.

II. Mechanisms: how a VNR ecology might operate

P.5

Gradient seeking:

Foraging clusters near EM/thermal/material gradients.

  • Predictions: excess clustering vs. matched baselines; lift at edges.

  • Falsifiers: normal after covariates.

  • Implication: site sensors in best-window cells from the atlas.​

P.6

Noise-floor operations:
Operate at or below environmental variance; bursty and clock-coupled.

  • Predictions: ms–s micro-bursts; twilight bias; sparse high-SNR.

  • Falsifiers: predominance of bright, continuous events.

  • Implication: two-tier pipeline (from fast triggers to deep analysis).

P.7

Modularity & reconfiguration:

No canonical “craft”; toolheads reconfigure by task.

  • Predictions: diverse shapes with function-linked patterns (e.g., survey swarms vs. couriers).

  • Falsifiers: single fixed morphology dominates credible cases.

  • Implication: classify by behavioral role, not silhouette.

P.8

Swarm advantage:

Small agents + distributed control outperform monoliths.

  • Predictions: pairs/triangles as ranging baselines; cooperative motion.

  • Falsifiers: purely independent, non-correlated trajectories.

  • Implication: formation-geometry detectors.

P.9

Transmedium operations: Crossing air/water/near-space for resources, not display.

  • Predictions: entries near resource interfaces; minimal wake/splash.

  • Falsifiers: transmedium only where spectators are dense (i.e., no gradient link).

  • Implication: place sensors at river mouths, ports, shelves.

P.10

Sanctuary tiers:
Activity in specific contexts (e.g., orbital night, high altitude).

  • Predictions: post-activity downlinks (Ka/optical) from sanctuary altitudes.

  • Falsifiers: no altitude-stratified behavior after long campaigns.

  • Implication: add up-look Ka/NIR watchers tied to coastal hot cells.

P.11

Tight-beam trunks (optical/NIR): Max bits/J, minimal intercept for long links.

  • Predictions: NIR micro-flashes time-locked to site activity.

  • Falsifiers: no coincidences across many hours/sites.

  • Implication: look for narrow-FOV NIR co-pointed with RF arrays.

P.12

RF camouflage: Spread-spectrum/UWB/backscatter to blend into human RF.

  • Predictions: sub-percent phase/AM blips on strong carriers; noise-like bursts.

  • Falsifiers: none beyond confounders after cyclostationary tests.

  • Implication: carrier-tracking + cyclostationary detectors.

P.13

Sparse sync, local autonomy: Most decisions on-board; network sync is opportunistic.

  • Predictions: irregular comms windows; event packs near geometry/favorable backgrounds.

  • Falsifiers: rigid periodic schedules independent of environment.

  • Implication: widen time gates; model terminator windows.

III. Observables: what should show up in data

P.14

Spatial clustering:
Lift in edge cells after controls.

  • Predictions: RR>1 at coasts/industrial belts/ionosphere hot spots.

  • Falsifiers: RR≈1 after covariates or lift shifts to non-gradient zones.

  • Implication: publish hot-cell maps; pre-register thresholds.

P.15

Temporal recurrence:
Return windows aligned with environmental cycles.

  • Predictions: departures from Poisson; twilight/seasonal recurrences.

  • Falsifiers: memoryless arrival times after controls.

  • Implication: Hawkes/renewal models; sliding-window tests.

P.16

Low-entropy emissions:
Structured EM/IR when active.

  • Falsifiers: all anomalies match thermal/known device spectra.

  • Controls: ionosondes, lightning, avionics; spectral-kurtosis/cyclostationary masks.

  • Implication: matched-filter chirp banks; entropy/complexity metrics.

P.17

Cross-sensor correlation: Coincident RF/EO/IR beyond chance.

  • Predictions: time-locked, multi-modal hits during active windows.

  • Falsifiers: coincidence rates ≈ controls/shuffled timelines.

  • Implication: GPS-sync; pre-set gates; null libraries.

P.18

Local interference side-effects: EMI on nearby systems (engines, radios) during activity.

  • Predictions: short, recoverable perturbations co-timed with signatures.

  • Falsifiers: no co-timed perturbations in instrumented zones.

  • Implication: instrument vehicles; log EMI with RF/EO/IR.

P.19

Formation logic:
Pairs/triangles exhibit stable baselines and task-aligned maneuvers.

  • Predictions: geometric regularities; cooperative turns; baseline maintenance.

  • Falsifiers: formations dissolve into independent, random tracks.

  • Implication: tracking that scores relative-motion features, not just absolute paths.

P.20

Model predictivity:
The siting model yields results where predicted.

  • Predictions: higher detection rates & richer signatures in predicted “best windows”.

  • Falsifiers: no lift vs. controls across seasons/sites; model underperforms naive baselines.

  • Implication: pre-register atlas; run prospective tests; update priors with nulls.

bottom of page