terminology
AI Graded Governance and Universal Access Terminology Standards
Document Positioning: This document unifies terminology usage standards for the Web showcase layer, ensuring consistent concepts and unified expression while preserving the tone of research draft, public discussion, and open questions.
Update Rule: Any new terminology, terminology change, or obsolescence must be synchronized with this table.
I. Obsolete Terminology (Prohibited in New Writing)
| Obsolete Term | Replacement Term | Reason for Obsolescence |
|---|---|---|
| Level 1-5 | Baseline Universal Layer/Restricted Operation Layer/Professional Execution Layer/Risk Decision Layer/System Definition Layer | Eliminate numerical numbering, weaken hierarchy metaphor |
| Level | Risk Threshold/Stair Position | Avoid technocratic numbering system |
| Hierarchy | Risk Threshold/Stair Position | Avoid hierarchy associations |
| Low-Authority Layer | Baseline Service User/Universal Layer | Eliminate "authority = rank" implication |
| High-Authority Layer | High-Risk Threshold Holder | Precise role description, avoid identity labels |
| Low-Authority Person | Baseline Service User | Unified |
| High-Authority Person | High-Risk Threshold Holder | Unified |
| Upgrade | Threshold Unlocking/Capability Development | Weaken progressivist ideology |
| Upgrade Path | Capability Development Path | Emphasize growth rather than climbing |
| Upgrade Channel | Threshold Unlocking Channel (mechanism documents) / Capability Development Channel (manifesto, social contract) | Contextual differentiation |
| Hierarchy Solidification | Threshold Access Solidification | Precise mechanism description |
| Cross-Layer | Cross-Threshold | Unified |
| 90% Quality Baseline | Quality Baseline | Eliminate specific numbers, avoid precision illusion |
| Stairway Universalism | AI Graded Governance and Universal Access | Avoid misreading as a new political ideology or faction |
| Manifesto | Research Draft / Problem Framework | Reduce declarative and finalizing tone |
| Political Manifesto | Technical Ethics and Public Policy Discussion | Clarify that the Web layer is not a mobilizing political text |
| Democratic Constraint | Multi-Stakeholder Institutional Constraint | Reduce institutional overclaiming and emphasize procedural participation |
| Technological Colonialism | Cross-National Capability-Access Inequality | Shift from accusatory language to mechanism-risk analysis |
| Anti-Technocratic Aristocracy | Preventing Excessive Concentration of Capability-Definition Power | Shift from camp critique to governance risk |
| Public Movement | Public Policy Discussion | Avoid mobilizing language |
II. Retained Terminology (Continue using, but must follow usage standards)
| Term | Usage Standard | Prohibited Variants |
|---|---|---|
| Risk-Graded Permission Model | Refers to risk threshold gradient, not capability level. Ensure context clearly defines this when using | Capability ladder, rank ladder, permission ladder in public titles |
| Capability Certification | Pre-authority unlocking assessment, containing three-dimensional assessment (technical, social coordination, ethical judgment) | Assessment, evaluation, qualification exam |
| Safety Encapsulation | Baseline service users' restrictive AI function access mechanism | Encapsulation (used alone) |
| Spillover Feedback | Mechanism for high-authority advantages flowing back to baseline layer | Feedback (used alone unless context is clear) |
| Quality Baseline | Baseline service minimum alarm line, not long-term justice line | 90% baseline, performance baseline |
| Risk Threshold | Operation authority level corresponding to different risk scenarios | Level, hierarchy |
| Baseline Service User | Citizens enjoying baseline universal layer services | Low-authority layer, low-authority person |
| High-Risk Threshold Holder | Operators obtaining Risk Decision Layer or System Definition Layer authority | High-authority layer, high-authority person |
| Human Rights Protection Channel | Emergency service mechanism bypassing normal certification in life-and-death scenarios | Emergency channel, privilege channel |
| Chronic Under-Provisioning | State where baseline services do not collapse but long-term stably lag behind | Quality gap, service downgrade |
| Threshold Access Solidification | Phenomenon where specific groups long-term monopolize specific risk thresholds | Class solidification, rank solidification |
III. Contextually Differentiated Terminology
The following terms adopt different expressions according to document type:
| Concept | Mechanism Design Documents | Manifesto/Social Contract/Theoretical Foundations |
|---|---|---|
| Authority Elevation | Threshold Unlocking | Capability Expansion |
| Ascension Channel | Threshold Unlocking Channel | Capability Development Channel |
| Authority Reduction | Threshold Downgrade | Responsibility Adjustment |
IV. New Terminology Approval Process
- Proposal: When any author needs to use new terminology in a document, first search this table for equivalent terminology
- Argumentation: If new terminology is indeed needed, must explain:
- Why existing terminology is insufficient
- What problem the new terminology solves
- The new terminology's relationship with the existing concept network
- Recording: After confirmation, write the new terminology into the appropriate section of this table
- Synchronization: Update all related documents referencing this concept
VI. Special Notes
On Publicly Downgrading the "Stairway" Metaphor
The Web showcase layer no longer uses "stairway" as a public-facing main title, and prioritizes "risk-graded permission model." This is because:
- Public expression needs to reduce hierarchy associations: "Stairway" can be misread as capability ranking, ascension narrative, or a new status order.
- Grading still needs to remain: Completely removing grading would obscure the fact that high-risk AI capabilities do require responsibility, audit, and access conditions.
- Critique documents continue supervision: Foucault's and Sandel's critiques of the stairway metaphor remain valid; these critiques are retained in theoretical foundation documents as continuous tension reminders.
On "Baseline Service User" vs. "Universal Layer"
Priority is given to "baseline service user" because it:
- Emphasizes "usage" relationship rather than "belonging" relationship
- Avoids the hierarchy implication of the word "layer"
- Forms a symmetrical verb-object structure with "high-risk threshold holder"
In contexts emphasizing emotional connection (such as social contract), "universal layer citizen" may be used at discretion.
Institutional Engineering Honesty: Any terminology is constructive. This standard does not pursue "the most correct terminology," but "the least bad terminology"—maintaining balance between operability, precision, and anti-hierarchization.