Survivors Watch Platform
Submit consequential actions for review, find active cases across Earth, the Moon, and Mars, and route them through AI-assisted evidence work and human comparator review.
Operational Governance Pipeline
The platform turns the OPT governance standards into a working civic pipeline. It is built for proposed actions, live deployments, institutional decisions, AI releases, infrastructure changes, policy moves, planetary risks, and rare public-actor decisions whose effects may scale beyond ordinary organizational boundaries.
Early in the workflow, the platform uses plain terms: cases, actions, subjects, reviewers, evidence, and decisions. Once a case is scoped, the formal OPT review object is a branch: an action-conditioned continuation proposed, initiated, or governed by an institution, AI deployment, policy process, infrastructure operator, planetary system, or exceptional public actor.
How the Pipeline Works
1. Discover or Add
Search finds known review records. The 3D heat map lets contributors select active situations spatially across Earth, the Moon, and Mars. Origin lists surface cases whose source is systemic, cosmic, or unforeseen. Cases that are not already recorded enter through structured intake.
2. Triage and Scope
Operators classify the subject, action, location, affected moral patients, origin class, evidence links, sensitivity, reversibility, expected decision horizon, and effective consequentiality class. Triage decides whether the case becomes a formal review assignment.
3. Assign Evidence Work
Voluntary contributors and domain reviewers receive bounded tasks: source verification, affected-party mapping, gate evidence, CPBI notes, translation, legal or safety review, and right-of-reply handling. The Branch Card records the six hard gates, CPBI dimensions, comparator requirements, monitoring triggers, rollback triggers, and draft decision.
4. Comparator Review
AI can assist triage, de-duplication, evidence-gap detection, and synthesis, but cannot approve final decisions. Human comparators review the evidence, publish redacted rationales where appropriate, and keep STAGE decisions under monitoring with correction and rollback paths.
Can Companies Submit Their Own Actions?
Yes. Pre-action submission is one of the intended uses. A company, public agency, AI lab, infrastructure operator, or other high-impact actor can submit a proposed action for review before it is launched. That registration is disclosure, not endorsement.
- Marked source: subject-submitted cases are visibly labeled so reviewers can account for conflicts and information asymmetry.
- No self-certification: the submitter can provide evidence, safeguards, and right-of-reply material, but cannot approve its own branch.
- Confidentiality with limits: sensitive pre-release details can be scoped or embargoed where disclosure would create risk, while the audit trail remains available to authorized reviewers.
- Same burden of proof: subject-submitted actions pass through the same triage, hard gates, CPBI scoring, comparator review, and monitoring logic as public or contributor-discovered cases.
This allows serious actors to ask for independent review before harm occurs, while preventing governance review from becoming reputation laundering.
Search, globe selection, origin lists, and subject submissions are different doors into the same review object.
The review asks whether the branch preserves or damages the future conditions under which observers can keep modelling reality.
Project Model
- Case discovery: ordinary search, a global heat map extended to the Moon and Mars, and lists by systemic, cosmic, or unforeseen origin.
- Contributor network: voluntary workers, domain reviewers, translators, safety reviewers, and comparators cooperate through bounded assignments.
- AI-assisted workflow: AI support helps with intake normalization, duplicate detection, evidence routing, draft summaries, and gap checks while final authority stays with human comparators.
- Governance firewall: donations and voluntary contributions support the commons, but donors, submitters, contributors, comparators, and publishers remain distinct roles.
This is not a reputation engine, public-shaming tool, or paid certification service. It is open-source civic infrastructure for reviewing high-consequence institutional actions, AI deployments, infrastructure changes, public policy proposals, planetary-scale risks, and rare public-actor actions with global, civilizational, or multi-planetary effects.