Maximum receptivity
AI extracts literal, symbolic, narrative, and contextual elements from the original dream text while marking speculation as speculation.
Project Brief
Dream Journal / Precognition
AI-assisted inquiry into possible anomalous information transfer across time.
Future of Inquiry is working with Andrew and one of the largest personally maintained dream corpora we are aware of to build a defensible AI research system for studying possible dream-to-public-event correspondences.
Protocol
Separate discovery, calibration, and proof.
The project is not trying to turn striking stories into instant evidence. It separates known-case calibration from blind discovery and requires provenance, contamination flags, ordinary-exposure alternatives, denominator analysis, and adversarial decoys before a case is treated as serious.
Blinded recovery
The target event is hidden. The system receives cleaned dream text and attempts to recover plausible later public events from the dream alone.
Adversarial scoring
Candidate matches are tested against decoys, timing, provenance, failed predictions, exposure pathways, and corpus-frequency alternatives.
Why It Matters
Precognition becomes testable only when the boring parts are handled well.
Claims about future information are easy to exaggerate and easy to dismiss. The FOI approach is to make the evidential chain visible: what was written, when it was written, what later event is proposed, what else could explain the match, and whether a blinded or adversarial process still finds the correspondence interesting.