Who has authority to decide
The responsible real-world actors retain decision-making authority. Launch-stage entry is valid only within a clearly defined high-consequence context, responsibility, task and scope of authorization.
P05-HUMANCOS-005 · One-page launch-stage introduction
Human-COS at the Launch Stage: Judgment and Responsibility Before Irreversibility
A one-page guide to when deeper judgment is needed: facts and unknowns, Causal Map, Responsibility Map, Stakeholder Visibility, alternative paths, termination conditions, the remaining window, and No Silent Crossing.
Boundaries to clarify before real-world use
Why this mechanism is needed, and how far it may go
After understanding why Human-COS is needed, we must also limit when it enters, what it examines, who is responsible and when it stops, so that a high-consequence tool does not expand into routine management of people’s thoughts.
What it can do
Before consequences become irreversible, make causal relationships, others’ perspectives, responsibility, long-term consequences, crucial dissent and stopping conditions visible in support of controlled high-consequence judgment.
What it cannot do
It cannot generally evaluate personhood, thoughts or values; execute automatically while bypassing the responsible actors; or make temporary powers permanent under the name of the launch stage.
How to test whether it actually works
Use real cases, isolated shadow operation, independent review and records of failure to compare whether it reveals crucial omissions earlier than the original process.
When to stop, exit or roll back
Stop immediately if authority is exceeded, responsibility is unclear, explanations are unavailable, judgment is misled or risk increases. Temporary powers must end when the launch phase ends.
How real-world results revise upstream design
Use results to revise launch conditions, the topology of responsibility, allocation of roles, and the interfaces between Human-COS, Nebula and AGI-COS.
The problem this work addresses
A one-page guide to when deeper judgment is needed: facts and unknowns, Causal Map, Responsibility Map, Stakeholder Visibility, alternative paths, termination conditions, the remaining window, and No Silent Crossing.
Place in the framework
This work belongs to the Human-COS family in the AGI governance and civilizational safety field. Its current role is a one-page launch-stage introduction.
Who should read it
General readers, decision-makers and professionals
Reading guidance
Begin here when a high-consequence matter arises. For theoretical and engineering details, continue to the formal Human-COS baseline and Runtime.
Version and publication status
- Current version
- V1.0
- Publication status
- Current concise entry
- Role of this work
- One-page launch-stage introduction
- Stable English page
- https://www.civitas.top/en/documents/p05-humancos-005/
Continue reading
- Seeing Earlier What Would Otherwise Be Seen Only Later: A Public Introduction to Human-COS — Full public explanation and participation paths.
- AGI-COS Controlled Version Family: A One-Page Guide by Stage — Select machine-side risk and permission tracks.
- Public Window Protocol: An Open Future and Three Constraints — Minimum boundaries for an open future.
Publication boundaries
This page describes the current controlled public version only. Audience-specific editions, research tracks, historical versions and engineering details retain their respective authority. Public explanation does not create new theory, permissions or authority to deploy.
Translation scope
This page translates the public web description. The reviewed English manuscript is available in the English PDF section below; its source edition is identified there.
English PDF
Download the English PDF · V1.0