How ReincarnatedAI separates experience, evidence, interpretation, and speculation — and what it would take to change its mind.
Not all data is equal. ReincarnatedAI applies a layered evaluation — starting with what was directly observed, then what can be checked, then what ordinary explanations remain, and finally what larger theory (if any) is justified.
The core principle: A claim about survival or nonlocal mind gets stronger only when it survives the strongest available alternatives — not when it accumulates supporting anecdotes. The question is always: what is the minimum interpretation consistent with the evidence? We work upward from that floor.
Give greatest weight to statements, drawings, or reports recorded before any connection to a previous personality is established, before media exposure, and before family contact. The timing of documentation is often more important than the content of a claim. A verified detail recorded in writing two weeks before investigation is worth more than a hundred details recorded two years later.
Each evidential domain involves stacked claims that require separate evaluation: (a) the experience happened, (b) the experience was anomalous, (c) the anomaly requires a non-ordinary explanation, (d) that explanation is survival or nonlocal consciousness. These are not the same claim. A veridical NDE perception is evidence for anomalous perception — it is not automatically evidence for a full afterlife.
Studies that define targets, samples, and methods before data collection are far stronger than retrospective collection. Case investigation is not reproducible in a laboratory sense, but investigators can still commit to methods before beginning — and should document the investigation process as carefully as the report.
For each piece of apparently anomalous knowledge, systematically investigate all conventional routes: overhearing, media access, family stories, investigator leakage, cryptomnesia, prior contact, childhood exposure. The question is not whether fraud is conceivable, but whether any specific normal channel could have supplied the specific knowledge in the specific case.
A single anomaly is interesting. The same anomaly appearing under multiple methodologies, across independent investigators, across cultures and eras, with consistent phenomenology, carries more weight. Convergence across NDE studies, reincarnation cases, and deathbed vision research does not prove survival — but it narrows the space of viable explanations.
The absence of an accepted mechanism does not disprove an anomaly. Quantum entanglement was accepted before its mechanism was understood; the hard problem of consciousness itself lacks an accepted mechanism. But anomalies without mechanism cannot be integrated into broader science, so mechanism-building is a legitimate research priority — not a burden of proof that must be cleared before taking evidence seriously.
Anecdotes are not worthless — they are evidence of the weakest kind. The question is what they can and cannot support.
Strength is a compound score — no single factor is sufficient, and multiple factors pointing in the same direction is more meaningful than any one of them alone.
Claims recorded in writing before any investigation or family contact. Ideally by someone with no stake in the outcome, using a standard protocol. The van Lommel study collected NDE reports before patients left the hospital for this reason.
Not general impressions but specific, low-probability details: names, addresses, physical descriptions, unusual events, wound locations. A child saying "I died near water" is less specific than "I drowned at a particular well in a named village."
Evidence that the child or experiencer could not have accessed the information through any conventional route — geographic isolation, age, language barrier, family history, media access. The rarer and more obscure the previous personality, the stronger the case.
In reincarnation cases: birthmarks corresponding to documented wounds, verified by medical attestation from both ends of the case. In NDE cases: medical records confirming the duration and completeness of brain shutdown during the claimed perception.
The investigator should have no prior relationship to either family, no financial stake in the outcome, and should ideally approach the case from a skeptical or neutral starting position. Self-reports by believers carry less weight than skeptical investigation with positive findings.
In reincarnation cases: unusual behaviors, phobias, skills, or food preferences consistent with the claimed past life, present before the verbal reports began and not explainable by the current family environment. Behavioral evidence is harder to fabricate retroactively than verbal claims.
A serious research program must name what would strengthen or weaken its preferred interpretations. Here is ours.