Skip to Content

Changes for page Automation

Last modified by Robert Schaub on 2026/02/08 21:22
From version 2.5
edited by Robert Schaub
on 2026/01/20 20:25
Change comment: Renamed back-links.
To version 1.1
edited by Robert Schaub
on 2025/12/16 21:42
Change comment: Imported from XAR

Summary

Details

Page properties
Parent
... ... @@ -1,1 +1,1 @@
1 -Archive.FactHarbor 0\.9\.40.Specification.WebHome
1 +FactHarbor.Specification.WebHome
Content
... ... @@ -3,7 +3,6 @@
3 3  Automation in FactHarbor amplifies human capability while implementing risk-based oversight.
4 4  
5 5  This chapter defines:
6 -
7 7  * Risk-based publication model
8 8  * Quality gates for AI-generated content
9 9  * What must remain human-only
... ... @@ -19,18 +19,14 @@
19 19  
20 20  1. **Input**: User pastes a block of raw text.
21 21  1. **Deep Analysis (Background)**: The system autonomously performs the full pipeline **before** displaying the text:
22 -
23 23  * Extraction & Normalisation
24 24  * Scenario & Sub-query generation
25 25  * Evidence retrieval with **contradiction search**
26 26  * Quality gate validation
27 27  * Verdict computation
28 -
29 29  1. **Visualisation (Extraction & Marking)**: The system displays the text with claims extracted and marked.
30 -
31 31  * **Verdict-Based Coloring**: The extraction highlights (e.g. Orange/Green) are chosen **according to the computed verdict** for each claim.
32 32  * **AI-Generated Label**: Clear indication that content is AI-produced
33 -
34 34  1. **Inspection**: User clicks a highlighted claim to see the **Reasoning Trail**, showing exactly which evidence and sub-queries led to that verdict.
35 35  
36 36  === 1.2 Technical Scope ===
... ... @@ -42,38 +42,33 @@
42 42  * **No Human Approval Gate**: Demonstrates scalable AI publication
43 43  * **Structured Sub-Queries**: Logic generated by decomposing claims into the FactHarbor data model
44 44  
40 +
45 45  == 2. Publication Model ==
46 46  
47 47  FactHarbor implements a risk-based publication model with three modes:
48 48  
49 49  === 2.1 Mode 1: Draft-Only ===
46 +* Failed quality gates
47 +* High-risk content pending expert review
48 +* Internal review queue only
50 50  
51 -**Mode 1 (Draft-Only)**: Failed quality gates or high-risk content pending expert review. Internal review queue only.
52 -
53 -See [[AKEL Publication Modes>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome#HPublicationModes]] for detailed mode specifications.
54 -
55 -
56 56  === 2.2 Mode 2: AI-Generated (Public) ===
51 +* Passed all quality gates
52 +* Risk tier B or C
53 +* Clear AI-generated labeling
54 +* Users can request human review
57 57  
58 -**Mode 2 (AI-Generated, Published)**: Passed all quality gates, risk tier B or C, clearly labeled as AI-generated. Users can request human review.
59 -
60 -See [[AKEL Publication Modes>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome#HPublicationModes]] for detailed requirements.
61 -
62 -
63 63  === 2.3 Mode 3: Human-Reviewed ===
57 +* Validated by human reviewers/experts
58 +* "Human-Reviewed" status badge
59 +* Required for Tier A content publication
64 64  
65 -**Mode 3 (Human-Reviewed, Published)**: Validated by human reviewers or experts, highest trust level. Required for Tier A content publication.
61 +See [[AKEL page>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome]] for detailed publication mode descriptions.
66 66  
67 -See [[AKEL Publication Modes>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome#HPublicationModes]] for detailed requirements.
68 68  
64 +== 3. Risk Tiers and Automation Levels ==
69 69  
70 -== 3. Risk tiers and Automation Levels ==
71 -
72 -
73 -Risk tiers determine review requirements and automation levels. See [[Governance>>FactHarbor.Organisation.Governance.WebHome#HRiskTierPolicyGovernance]] for tier policy governance.
74 -
75 75  === 3.1 Tier A (High Risk) ===
76 -
77 77  * **Domains**: Medical, legal, elections, safety, security
78 78  * **Automation**: AI can draft, human review required for "Human-Reviewed" status
79 79  * **AI publication**: Allowed with prominent disclaimers and warnings
... ... @@ -80,7 +80,6 @@
80 80  * **Audit rate**: Recommendation: 30-50%
81 81  
82 82  === 3.2 Tier B (Medium Risk) ===
83 -
84 84  * **Domains**: Complex policy, science, causality claims
85 85  * **Automation**: AI can draft and publish (Mode 2)
86 86  * **Human review**: Optional, audit-based
... ... @@ -87,12 +87,12 @@
87 87  * **Audit rate**: Recommendation: 10-20%
88 88  
89 89  === 3.3 Tier C (Low Risk) ===
90 -
91 91  * **Domains**: Definitions, established facts, historical data
92 92  * **Automation**: AI publication default
93 93  * **Human review**: On request or via sampling
94 94  * **Audit rate**: Recommendation: 5-10%
95 95  
84 +
96 96  == 4. Human-Only Tasks ==
97 97  
98 98  These require human judgment and cannot be automated:
... ... @@ -104,6 +104,7 @@
104 104  * **Audit system oversight** and quality standard definition
105 105  * **Risk tier policy** adjustments based on societal context
106 106  
96 +
107 107  == 5. AI-Draft with Audit (Semi-Automated) ==
108 108  
109 109  AKEL drafts these; humans validate via sampling audits:
... ... @@ -118,12 +118,12 @@
118 118  * **Public summaries** and accessibility text
119 119  
120 120  Most Tier B and C content remains in AI-draft status unless:
121 -
122 122  * Users request human review
123 123  * Audits identify errors
124 124  * High engagement triggers review
125 125  * Community flags issues
126 126  
116 +
127 127  == 6. Fully Automated Structural Tasks ==
128 128  
129 129  These require no human interpretation:
... ... @@ -137,25 +137,40 @@
137 137  * **Layout generation** (diagrams, summaries, UI presentation)
138 138  * **Federation integrity checks** (cross-node data validation)
139 139  
130 +
140 140  == 7. Quality Gates (Automated) ==
141 141  
133 +Before AI-draft publication (Mode 2), content must pass:
142 142  
143 -Before AI-generated publication (Mode 2), content must pass four automated quality gates:
135 +1. **Source Quality Gate**
136 + * Primary sources verified
137 + * Citations complete and accessible
138 + * Source reliability scored
144 144  
145 -1. **Source Quality** - Primary sources verified, citations complete
146 -2. **Contradiction Search** (MANDATORY) - Counter-evidence actively sought
147 -3. **Uncertainty Quantification** - Confidence scores calculated
148 -4. **Structural Validation** - Required fields present, format valid
140 +2. **Contradiction Search Gate** (MANDATORY)
141 + * Counter-evidence actively sought
142 + * Reservations and limitations identified
143 + * Bubble detection (echo chambers, conspiracy theories)
144 + * Diverse perspective verification
149 149  
150 -See [[AKEL Quality Gates>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome#HQualityGates]] for complete gate specifications.
146 +3. **Uncertainty Quantification Gate**
147 + * Confidence scores calculated
148 + * Limitations stated
149 + * Data gaps disclosed
151 151  
151 +4. **Structural Integrity Gate**
152 + * No hallucinations detected
153 + * Logic chain valid
154 + * References verifiable
152 152  
156 +See [[AKEL page>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome]] for detailed quality gate specifications.
157 +
158 +
153 153  == 8. Audit System ==
154 154  
155 155  Instead of reviewing all AI output, systematic sampling audits ensure quality:
156 156  
157 157  === 8.1 Stratified Sampling ===
158 -
159 159  * Risk tier (A > B > C sampling rates)
160 160  * Confidence scores (low confidence → more audits)
161 161  * Traffic/engagement (popular content audited more)
... ... @@ -163,9 +163,7 @@
163 163  * User flags and disagreement signals
164 164  
165 165  === 8.2 Continuous Improvement Loop ===
166 -
167 167  Audit findings improve:
168 -
169 169  * Query templates
170 170  * Source reliability weights
171 171  * Contradiction detection algorithms
... ... @@ -173,11 +173,11 @@
173 173  * Bubble detection heuristics
174 174  
175 175  === 8.3 Transparency ===
176 -
177 177  * Audit statistics published
178 178  * Accuracy rates by tier reported
179 179  * System improvements documented
180 180  
183 +
181 181  == 9. Automation Roadmap ==
182 182  
183 183  Automation capabilities increase with system maturity while maintaining quality oversight.
... ... @@ -185,7 +185,6 @@
185 185  === 9.1 POC (Current Focus) ===
186 186  
187 187  **Automated:**
188 -
189 189  * Claim normalization
190 190  * Scenario template generation
191 191  * Evidence metadata extraction
... ... @@ -195,7 +195,6 @@
195 195  * **Risk tier assignment**
196 196  
197 197  **Human:**
198 -
199 199  * High-risk content validation (Tier A)
200 200  * Sampling audits across all tiers
201 201  * Quality standard refinement
... ... @@ -204,7 +204,6 @@
204 204  === 9.2 Beta 0 (Enhanced Automation) ===
205 205  
206 206  **Automated:**
207 -
208 208  * Detailed scenario generation
209 209  * Advanced evidence reliability scoring
210 210  * Cross-scenario comparisons
... ... @@ -213,7 +213,6 @@
213 213  * **Increased AI-draft coverage** (more Tier B content)
214 214  
215 215  **Human:**
216 -
217 217  * Tier A final approval
218 218  * Audit sampling (continued)
219 219  * Expert validation of complex domains
... ... @@ -222,7 +222,6 @@
222 222  === 9.3 Release 1.0 (High Automation) ===
223 223  
224 224  **Automated:**
225 -
226 226  * Full scenario generation (comprehensive)
227 227  * Bayesian verdict scoring across scenarios
228 228  * Multi-scenario summary generation
... ... @@ -231,12 +231,12 @@
231 231  * **Most Tier B and all Tier C** auto-published
232 232  
233 233  **Human:**
234 -
235 235  * Tier A oversight (still required)
236 236  * Strategic audits (lower sampling rates, higher value)
237 237  * Ethical decisions and policy
238 238  * Conflict resolution
239 239  
237 +
240 240  == 10. Automation Levels Diagram ==
241 241  
242 242  {{include reference="FactHarbor.Specification.Diagrams.Automation Level.WebHome"/}}
... ... @@ -254,7 +254,8 @@
254 254  
255 255  == 13. Related Pages ==
256 256  
257 -* [[AKEL (AI Knowledge Extraction Layer)>>Archive.FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome]]
255 +* [[AKEL (AI Knowledge Extraction Layer)>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome]]
258 258  * [[Requirements (Roles)>>FactHarbor.Specification.Requirements.WebHome]]
259 259  * [[Workflows>>FactHarbor.Specification.Workflows.WebHome]]
260 -* [[Governance>>Archive.FactHarbor.Organisation.Governance.WebHome]]
258 +* [[Governance>>FactHarbor.Organisation.Governance]]
259 +