Changes for page Automation

Summary

• Page properties (1 modified, 0 added, 0 removed)

Details

Page properties

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 ===
...	...	@@ -42,12 +42,12 @@
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 ===
50		-
51	51	**Mode 1 (Draft-Only)**: Failed quality gates or high-risk content pending expert review. Internal review queue only.
52	52
53	53	See [[AKEL Publication Modes>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome#HPublicationModes]] for detailed mode specifications.
...	...	@@ -54,7 +54,6 @@
54	54
55	55
56	56	=== 2.2 Mode 2: AI-Generated (Public) ===
57		-
58	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	59
60	60	See [[AKEL Publication Modes>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome#HPublicationModes]] for detailed requirements.
...	...	@@ -61,7 +61,6 @@
61	61
62	62
63	63	=== 2.3 Mode 3: Human-Reviewed ===
64		-
65	65	**Mode 3 (Human-Reviewed, Published)**: Validated by human reviewers or experts, highest trust level. Required for Tier A content publication.
66	66
67	67	See [[AKEL Publication Modes>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome#HPublicationModes]] for detailed requirements.
...	...	@@ -71,9 +71,7 @@
71	71
72	72
73	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
	85	+
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
	97	+
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
	117	+
127	127	== 6. Fully Automated Structural Tasks ==
128	128
129	129	These require no human interpretation:
...	...	@@ -137,6 +137,7 @@
137	137	* **Layout generation** (diagrams, summaries, UI presentation)
138	138	* **Federation integrity checks** (cross-node data validation)
139	139
	131	+
140	140	== 7. Quality Gates (Automated) ==
141	141
142	142
...	...	@@ -155,7 +155,6 @@
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
	169	+
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
	223	+
240	240	== 10. Automation Levels Diagram ==
241	241
242	242	{{include reference="FactHarbor.Specification.Diagrams.Automation Level.WebHome"/}}
...	...	@@ -257,4 +257,5 @@
257	257	* [[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]]
	244	+* [[Governance>>FactHarbor.Organisation.Governance]]
	245	+