Medical Reviewer for AI-Assisted NDE Case Profile Calibration

The Future of Inquiry Institute is looking for a medically trained reviewer to help with a small AI-assisted research project involving near-death experience accounts from NDERF. This is not an advocacy, religious, or metaphysical assignment. The task is to help evaluate AI-generated case profiles as medical self-report narratives. We need someone who can review the AI’s medical reasoning and help us improve the prompt/rubric used to classify and score cases. The immediate calibration issue is this: The AI may be over-penalizing cases for lack of formal external documentation, such as hospital records, contemporaneous clinical notes, or written witness statements. In this field, that kind of documentation is often unavailable, so missing documentation should be noted clearly but should not automatically count as a major evidential weakness. At the same time, the AI still needs to evaluate ordinary explanations, internal inconsistencies, medically unrealistic details, and the distinction between: - self-report, - reported corroboration, - independently documented corroboration. We are looking for someone who can review one AI-generated case profile and provide a short calibration memo. Ideal background: - MD, DO, MBBS, medical resident, or clinically trained reviewer - Emergency medicine, critical care, anesthesiology, cardiology, neurology, internal medicine, hospital medicine, or oncology background helpful - Experience with clinical narratives, case reports, chart review, medical documentation, or medical research writing helpful - Comfortable working with uncertainty and incomplete documentation - Able to give specific feedback, not just a general opinion - Comfortable reading Markdown or Google Docs Initial paid trial: We would like to start with one fixed-price paid sample task. The reviewer will receive: - one AI-generated case profile, - a short project summary, - the current calibration issue, - specific questions to answer. Deliverable: A short written memo answering questions such as: - Does the AI classify the medical event correctly? - Does it separate medical-event verification from content verification? - Does it over-penalize missing external documentation? - Does it appropriately identify limitations? - Does it overestimate ordinary guessability? - Are there medical details the AI misunderstands or underweights? - Are any details medically unrealistic, internally inconsistent, hallucinatory, or more plausibly ordinary? - Are the case rating and score medically reasonable? - What prompt/rubric changes would improve future reviews? This first task is expected to take about 2-3 hours. If the trial is a good fit, there may be additional work reviewing more case profiles and helping refine the scoring protocol. Please answer these questions when applying: 1. What is your medical training/background? 2. Have you reviewed clinical narratives, case reports, chart notes, or medical documentation before? 3. Are you comfortable reviewing patient self-report when formal records may not be available? 4. How would you distinguish “lack of documentation” from “evidence against the claim”? 5. Are you comfortable giving prompt/rubric feedback, not just a medical summary?