Future Mind Institute Research Governance Charter
Version: 1.0, draft for Board adoption
Institution: Future Mind Institute, a United States nonprofit institution
Effective date: Upon Board adoption; reviewed annually
Texts: Chinese and English provisions correspond. The Board shall publicly resolve interpretive discrepancies without reducing research protections.
Article 1. Mission, Positioning, and Authority
Future Mind Institute shall neutrally fund consciousness and life-sciences research without committing to a metaphysical position. Its publicly identified Seth Material Research Program shall maintain a separately disclosed budget and project portfolio. The Seth Material is strictly a hypothesis generator, never evidence, a diagnostic authority, or clinical guidance. Funding shall depend on adjudicability, methodological quality, and public value, regardless of whether findings support the material.
This Charter binds directors, employees, advisers, grantees, and subcontracted teams through grant agreements. The Board retains ultimate governance responsibility; the Scientific Advisory Committee exercises scientific judgment. An independent compliance officer reports directly to the Board and may suspend noncompliant activities. Applicable law takes precedence; institutional safeguards may exceed legal requirements.
Bial Foundation, Fetzer Institute, Templeton World Charity Foundation, BICS, UVA DOPS, and IONS are comparative reference organizations, implying neither endorsement nor equivalent policies. COGITATE provides a reference for preregistered adversarial collaboration and cross-laboratory testing. COGITATE description
Article 2. Scientific Advisory Committee
The Committee shall comprise nine voting members covering consciousness science, statistics and measurement, life sciences, physical metrology, clinical ethics, and historical-source methods. At least two-thirds shall be independent of institutional employment and donor relationships. At least one member shall hold a substantively different interpretation of relevant claims. Specialist reviewers shall supplement these capabilities; one generalist cannot substitute for multidisciplinary review.
Members serve staggered three-year terms, renewable once. The Board shall publish selection criteria, biographies, and appointment or removal reasons. Donations confer no entitlement to membership. Quorum requires two-thirds of eligible, unconflicted members and at least five members. Ordinary decisions require a majority; major agenda changes require two-thirds approval. Public decision summaries shall record dissent.
Members shall disclose financial interests, appointments, collaborations, family relationships, intellectual property, and relevant advocacy annually and for each case. Joint publication, direct supervision, or material financial relationships with an applicant during the preceding three years require recusal from discussion, case-material access, and voting. Theoretical disagreement alone does not require recusal. Independent alternates shall restore quorum when necessary.
Article 3. Claims Ledger and Source Review
Within ninety days of adoption, the data steward shall publish a machine-readable CSV and JSON ledger, data dictionary, and validation rules. The draft’s 383 concepts constitute a coverage baseline pending verification. Each shall be classified as mapped, duplicated, untestable, or awaiting review. Counts of concepts, entries, or publications shall not represent counts of independent evidence.
Each record shall contain a stable claim_id; concept links; original wording and language; title and edition; page, chapter, and session references; composition date; context; source permissions; version relationships; review status; direct versus adjacent evidence; independent-sample identifiers; effect estimates and intervals; scope; adjudication conditions; and responsible reviewer. Splits and mergers shall retain previous identifiers and relationships. Missing information shall be explicitly marked, never inferred into apparent fact.
Two reviewers shall check the original English corpus, recording editorial changes, editions, translation choices, and disagreements. Review shall distinguish gaps in consciousness from attention fluctuations, interoception from access to hidden external information, and physical temperature from temperature imagery. A public mapping shall connect original assertion → added bridging hypothesis → experimental endpoint. Weak magnetic inputs, light-mediated cellular communication, and random-target tasks shall be identified as added operationalizations, with reasons they provide fair tests and explanations they may omit.
Historical envelope, dark-room, and dream-precognition cases require inventories of every recoverable attempt, original report timestamps, target custody, feedback, omissions, and retrospective revisions. Inadequate denominators or custody chains restrict records to archival leads; they cannot support success-rate estimates or confirmatory conclusions.
Article 4. Funding Process and Review Criteria
Funding shall proceed through letter of intent → independent peer review → preregistration review → milestone disbursement. Letters shall identify claims, bridging mappings, competing explanations, capabilities, and budgets. Each application requires at least three external reviewers, collectively providing domain expertise, methodological expertise, and a different explanatory perspective. Internal administrators cannot replace external review.
| Review dimension | Weight | Required submission |
|---|---|---|
| Adjudicability and source correspondence | 20% | Mapping, scope, and three outcome categories |
| Design, statistics, and measurement | 25% | Smallest meaningful effect, reliability, power simulations |
| Alternative explanations and leakage controls | 20% | Blinding, randomization, competing models |
| Capability and independent replication | 15% | Expert responsibilities, facilities, replication plan |
| Ethics and open research | 15% | Ethics pathway, data and publication plans |
| Cost and public value | 5% | Itemized budget and expected information gain |
Reviewers score each dimension from zero to five, converted to a weighted total of 100. Recommendation requires at least 80. Failure on ethics, adjudicability, data integrity, or essential expertise requires return or rejection regardless of aggregate score. The Committee recommends awards; the Board approves budgets and publicly explains departures from scientific ranking.
Ordinarily, disbursements shall be 20%, 40%, 30%, and 10%, respectively tied to contractual and compliance preparation; approved preregistration and ethics clearance; compliant collection milestones; and complete reporting and archiving. Final payment cannot depend on positive findings or journal acceptance. Applicants may appeal procedural errors within thirty days to reviewers uninvolved in the original decision.
Article 5. Preregistration, Open Data, and Negative Results
All scientific projects—including pilots, replications, archival audits, and theoretical-methodological work—shall register on OSF or AsPredicted before undertaking the planned work. A full protocol shall supplement a short-form registration whenever necessary. Existing-data studies shall register before new analyses and disclose prior access and known information; retrospective registration shall never be represented as prospective. OSF guidance; AsPredicted description
Registration shall fix endpoints, independent sampling units, eligibility rules, randomization, blinding, stopping rules, missing-data treatment, multiplicity adjustments, the smallest meaningful effect, power simulations, and success, failure, and inconclusive criteria. Analyses shall distinguish event frequency from time proportion, trials from independent samples, equivalence from nonsignificance, absence of report from unconsciousness, and prediction from causal influence. Parameter-recovery studies and measurement-reliability validation are required where applicable.
Registrations shall normally become public before recruitment or work begins. Targets and keys susceptible to leakage shall be sealed with an independent custodian, accompanied by public cryptographic hashes. Amendments shall preserve timestamps and previous versions. Analyses formulated after outcome knowledge shall be labeled exploratory.
Completion status shall be public within thirty days of data lock, and a results preprint within 180 days. Deidentified data, dictionaries, materials, code, and execution environments shall be released by the first public results report or 270 days after lock, whichever occurs first. For terminated projects, deadlines run from termination. Restrictions justified by privacy, consent, or legal rights require independent approval, public reasons, metadata, a controlled-access pathway, and annual reassessment. Confidentiality does not waive reporting.
Every award shall contain protected funding for preparing, storing, and reasonably publishing negative, null, and inconclusive findings. The Institute guarantees funding and public archiving, not journal acceptance. Registered Reports are preferred; Stage 1 in-principle acceptance shall be distinguished from Stage 2 assessment of completed results.
Article 6. Adversarial Collaboration and Common Adjudication
Contested claims require proponents and researchers favoring different explanations to jointly lock the protocol under an independent facilitator. Neither side may block reporting because of its preferred conclusion. Without agreement on a fair test, funding is limited to public methodological negotiation; confirmatory experimentation shall not begin.
Both sides shall specify success, failure, inconclusive outcomes, exclusions, and a bounded set of amendment triggers. Conditions shall not expand indefinitely after results become known. Even successful hidden-information identification requires comparison with leakage, randomization defects, unknown ordinary channels, prediction, and influence models. It does not uniquely establish disembodied consciousness, temporal travel, or personality persistence. Reliable negative findings constrain only the registered version and tested effect range.
An independent statistical team shall lock code while blinded. Each side may submit separately identified analyses without unilateral authority to rewrite or suppress the other’s work. Before the initial experiment, the award shall designate an independent replication team, budget, and activation criteria. Replication shall not share critical personnel, targets, or the original collection chain.
Article 7. Specialist Capabilities and Evidence Gates
Evidence review requires legally accessible full texts and checks of Registered Report stage, retractions, corrections, sample overlap, underlying data, effect sizes, and high-quality negative replications. A valid DOI does not establish evidential reliability. Support for an adjacent mechanism shall not be promoted into support for the original proposition.
Immunology, microbiology, developmental biology, and genetics specialists shall respectively review immune function, cellular communication, bioelectricity, and genetic regulation, supplemented by additional expertise as the task requires. The Institute shall publish a collaboration capability map identifying personnel, facilities, quality controls, and unresolved gaps. A general cross-scale cognition laboratory cannot stand in for these distinct capabilities.
Mass-change, remote electromagnetic, matter-pulse, and cross-universe coupling proposals require independent metrology, thermodynamics, information-security, and instrumentation-engineering review before experimental funding. Submissions shall provide error budgets, carrier descriptions or coupling equations, sensitivity ranges, and parameter intervals that the design could exclude. Proposals lacking identifiable predictions cannot pass this gate.
Cross-life personality, post-resuscitation experience, and anomalous ancient-knowledge proposals require joint review spanning clinical resuscitation, reconstructive memory, developmental psychology, history, archaeology, and cultural anthropology. Cardiac arrest shall not be equated with irreversible death. Contemporary ability experiments cannot establish that an ancient event occurred.
Article 8. Donor Firewall
Donors have no result-veto rights and no authority to select recipient teams, specific experiments, endpoints, analyses, or publication timing. Restrictions may refer only to broad themes previously approved by the Board. All such restricted funding combined shall not exceed 20% of the annual research budget, establishing x = 20. Related donors shall be aggregated; multiyear commitments count according to expenditure in the relevant year. Midyear budget reductions require recalculation.
Gift agreements shall expressly exclude scientific control. The Institute shall reject conditions tied to outcomes, withdrawal of findings, or secrecy. Within ninety days, disclosures shall identify institutional donors, individual names or lawful anonymity categories, amount bands, restrictions, and relevant relationships. Public anonymity does not conceal actual identity from the compliance officer or independent auditor and cannot disguise directed influence. Project communications between investigators and donors shall be retained.
Article 9. Ethics and Clinical Boundaries
Before recruitment, human research requires review by a qualified Institutional Review Board, or a written exemption or non-human-research determination from an authorized institutional body. Investigators cannot exempt themselves. Applicable United States human-subject, privacy, and clinical-trial requirements shall be followed. Even when federal funding rules do not directly apply, equivalent participant protections remain a contractual minimum. 45 CFR 46
Standard treatment shall never be replaced, delayed, or discouraged. Prohibited practices include blaming patients, attributing illness to thoughts or beliefs, inducing false identity or past-life memories, and sleep deprivation, hypoxia, fasting, or other distress-based procedures intended to produce effects. Research shall neither induce cardiac arrest nor interfere with resuscitation.
Consent shall explain investigational status, uncertainty, withdrawal rights, and the absence of consequences for treatment. Additional protections shall address children, people with impaired decision-making capacity, and bereaved families. Serious or unexpected risks require immediate suspension of affected procedures and notification to the Institute within twenty-four hours, alongside applicable IRB reporting. Resumption requires written safety review.
Article 10. Conflicts, Publication, and Public Communication
Relevant interests shall be disclosed in applications, reviews, annual reports, and publications. Authorship shall reflect actual contribution and accountability, documented through contributor statements. Funding, donations, or supplying source material confer no automatic authorship. Ghostwriting, honorary authorship, and funder suppression are prohibited.
Preprints shall clearly state their unreviewed status and report registration deviations, all primary endpoints, and uncertainty. Journals shall be selected for disciplinary fit, transparent review, and openness, not ideological alignment. Factual and privacy review shall be limited to fifteen working days and creates no veto over conclusions.
Public communications shall not claim to “validate Seth,” prove an entire worldview, or establish clinical efficacy. Statements may describe only the specific registered hypothesis, conditions, effect, and limitations. Errors shall be corrected promptly, with corrections linked to the original communication.
Article 11. Licensing and Versioned Preservation
Ledgers, data, reports, and publications that the Institute is entitled to license shall use CC BY 4.0, identifying authors, sources, licensing, and modifications. Code shall use MIT or Apache-2.0 with dependency licensing documented. Personal information, third-party original texts, and material lacking relicensing rights do not acquire an open license merely through inclusion in a compilation. CC BY 4.0 terms
Outputs shall be deposited in Zenodo with version-specific DOIs, file hashes, change logs, and registration links. Citations shall identify the version actually used; historical records shall not be overwritten. Controlled original records shall be retained for at least ten years, or longer when required by law or consent terms. Zenodo record management
Dataset DOI 10.5281/zenodo.22484474, its stated CC BY 4.0 license, and the research website are information supplied for this draft. Their contents could not be retrieved for verification. The data steward shall verify accessibility, license scope, the 383-concept inventory, and version history before representing them as audited.
Article 12. Annual Audit, Agenda Renewal, and Enforcement
Within 120 days after each fiscal year, the Institute shall publish an independent governance audit covering finances and the 20% ceiling, recusals, ethics, registration deviations, failure denominators, release deadlines, replication, communications, and ledger completeness. This is distinct from any separately required statutory financial audit. The report shall disclose every project’s status, corrective-action owners, and deadlines, with follow-up after ninety days. Whistleblowers shall be protected against retaliation.
Annual agenda renewal shall begin with thirty days of public comment. The Committee shall recommend retention, revision, suspension, or retirement based on evidence and adjudicability; the Board shall approve changes and publish version differences. The draft’s stated eighteen directions receive no automatic authorization merely through inclusion. Unadjudicable cosmological or consciousness-unit assertions shall not be included to achieve nominal coverage. Normative recommendations are not themselves natural-science propositions; their measurable institutional consequences require separate operationalization.
Theory without differential predictions may receive methodological microgrants only, capped at 0.5% of the annual research budget per award and 5% in aggregate. Such work shall not be listed as an experimental direction. Reliably excluding a specific proposed benefit or clarifying a measurement confusion is publishable progress and need not be disruptive.
Responsibility for the ten identified gaps is assigned as follows: gaps 1–4 to Article 3; gap 5 to Article 7; gap 6 to Article 5; gaps 7–9 to Articles 7 and 9; and gap 10 to Articles 6 and 12. Violations may trigger withheld disbursements, corrective action, termination, and legally available recovery, but never deletion of negative findings. Charter amendments require thirty days of public consultation and approval by two-thirds of the Board; they cannot retrospectively change locked adjudication standards.