Wiki source code of Governance

Last modified by Robert Schaub on 2026/02/08 08:29

version	line-number	content
1.1	1	= Governance =
1.2	2
1.1	3	FactHarbor governance is simple, transparent, flat, and focused on enabling automation.
1.2	4	{{include reference="Archive.FactHarbor 2026\.02\.08.Organisation.Diagrams.Governance Structure.WebHome"/}}
	5
1.1	6	== 1. Governance Philosophy ==
1.2	7
1.1	8	Core Principles:
1.2	9
1.1	10	* Automation over bureaucracy: Minimize manual processes
	11	* Transparency by default: Open decision-making
	12	* Community input: Listen but decide decisively
	13	* Measured outcomes: Data drives decisions
	14	* Adaptive structure: Evolve as needed
1.2	15
1.1	16	== 2. Organizational Structure ==
1.2	17
1.1	18	Flat Cooperative Model:
1.2	19
1.1	20	* Small organization with collaborative teamwork
	21	* No hierarchical management layers
	22	* Decisions by consensus when possible, voting when needed
	23	* Everyone contributes across multiple areas
	24	General Assembly → Governing Team → Team Members
1.2	25
1.1	26	=== 2.1 General Assembly (All Members) ===
1.2	27
1.1	28	Decides: Governing Team election, statutes, major strategic changes, budget
	29	Meets: Annually
1.2	30
1.1	31	=== 2.2 Governing Team ===
1.2	32
1.1	33	Decides: Strategy, policy, budget allocation, hiring
	34	Meets: Quarterly
	35	Size: small group (Facilitator, Coordinator, Treasurer + others)
1.2	36
1.1	37	=== 2.3 Team Members ===
1.2	38
1.1	39	* Technical Coordinator: AKEL & infrastructure
	40	* Community Coordinator: Moderators & contributors
	41	* Moderators (part-time): Handle abuse/disputes
1.2	42
1.1	43	== 3. Decision Authority ==
1.2	44
1.1	45	Day-to-day: Technical Coordinator + Community Coordinator
	46	Tactical: Governing Team (simple majority)
	47	Strategic: General Assembly (2/3 majority)
1.2	48
1.1	49	== 4. Policy Development ==
1.2	50
1.1	51	RFC Process (Request for Comments):
1.2	52
1.1	53	1. Anyone drafts proposal
	54	2. Community discussion
	55	3. Governing Team reviews and votes
	56	4. Decision published
	57	Emergency changes: Technical Coordinator can act immediately, Governing Team ratifies later
1.2	58
1.1	59	== 5. Financial Governance ==
1.2	60
1.1	61	* Annual budget approved by General Assembly
	62	* Two-signature requirement for >CHF 5,000
	63	* Governing Team approval for >CHF 20,000
	64	* Quarterly financial reports
	65	* Annual independent audit
1.2	66
1.1	67	== 6. Automation Governance ==
1.2	68
1.1	69	Core Principle: AKEL makes content decisions. Humans make system decisions.
1.2	70
1.1	71	=== 6.1 Decision Boundary ===
1.2	72
1.1	73	What AKEL Decides (Automated):
1.2	74
1.1	75	* All claim verdicts and confidence scores
	76	* All evidence assessments and relevance scores
	77	* All source track record scores
	78	* All risk tier classifications
	79	* All publication decisions
	80	* All scenario extractions
	81	Rationale: These decisions must be automated for scale, consistency, transparency, and to avoid human bias. Humans cannot process millions of claims reliably.
	82	Human Role: Monitor aggregate performance metrics, identify systematic issues, improve algorithms.
	83	What Humans Decide:
	84	Strategic Decisions (General Assembly, 2/3 majority):
	85	* Mission and values
	86	* Risk tier policy definitions
	87	* Major architectural changes
	88	* Budget allocation
	89	* Dissolution
	90	Tactical Decisions (Governing Team, simple majority):
	91	* Algorithm parameter ranges (within policy)
	92	* Infrastructure investments
	93	* Hiring and role assignments
	94	* Community policies
	95	* Partnership agreements
	96	Operational Decisions (Domain Owners, autonomous):
	97	* Technical Coordinator: AKEL performance optimizations, infrastructure changes
	98	* Community Coordinator: Community process improvements, documentation
	99	* Moderators: Handling AKEL-flagged items, detection improvement proposals
	100	Emergency Decisions (any team member, ratified by Governing Team):
	101	* Critical security issues
	102	* Legal compliance requirements
	103	* Immediate safety concerns
1.2	104
1.1	105	=== 6.2 Principle: Fix the System, Not the Data ===
1.2	106
1.1	107	When AKEL makes a "wrong" decision:
1.2	108
1.1	109	* ❌ Do NOT manually override that specific verdict
	110	* ✅ DO investigate: Is this a systematic issue?
	111	* ✅ DO improve: Change algorithm/policy to handle such cases better
	112	* ✅ DO test: Validate improvement on historical data
	113	* ✅ DO deploy: Roll out improved system
	114	* ✅ DO monitor: Check if metrics improve
	115	Example:
	116	* Bad: "AKEL rated this source too low, I'll manually boost it"
	117	* Good: "AKEL consistently under-rates peer-reviewed sources. Let's adjust the scoring algorithm to weight peer-review more heavily."
1.2	118
1.1	119	=== 6.3 Governance of AKEL ===
1.2	120
1.1	121	Quarterly Performance Review:
1.2	122
1.1	123	* Who: Governing Team + Technical Coordinator
	124	* What: Review AKEL performance metrics, bias audits, user feedback patterns
	125	* Output: Performance report, improvement priorities, policy updates if needed
	126	Performance Metrics Monitored:
	127	* Processing speed (P50, P95, P99)
	128	* Success rate and error rate
	129	* Evidence completeness
	130	* Confidence score distribution
	131	* User feedback (helpful/unhelpful ratio)
	132	* Bias indicators (by domain, source type, etc.)
	133	Triggers for Policy Review:
	134	* Metrics outside acceptable ranges
	135	* Systematic bias detected
	136	* Major user complaints about fairness
	137	* Legal/compliance concerns
	138	* New domains requiring special handling
	139	Algorithm Change Process:
1.2	140
1.1	141	1. Identify issue from metrics
	142	2. Propose solution (RFC - Request for Comments)
	143	3. Test in staging environment
	144	4. Measure impact on historical data
	145	5. Technical Coordinator approves (or escalates to Governing Team for policy changes)
	146	6. Deploy with monitoring
	147	7. Evaluate results
1.2	148
1.1	149	=== 6.4 Human Intervention Criteria ===
1.2	150
1.1	151	Legitimate reasons to intervene:
1.2	152
1.1	153	* ✅ AKEL explicitly flags item for human review
	154	* ✅ System metrics show performance degradation
	155	* ✅ Legal/safety issue requires immediate action
	156	* ✅ User reports reveal systematic bias pattern
	157	Illegitimate reasons (system improvement needed instead):
	158	* ❌ "I disagree with this verdict" → Improve algorithm
	159	* ❌ "This source should rank higher" → Improve scoring rules
	160	* ❌ "Manual quality gate before publication" → Defeats purpose of automation
	161	* ❌ "I know better than the algorithm" → Then improve the algorithm
1.2	162
1.1	163	=== 6.5 Consent-Based Decision Making ===
1.2	164
1.1	165	For system changes, use consent not consensus (from Sociocracy 3.0):
	166	Consent = No principled objections
1.2	167
1.1	168	* Faster than consensus
	169	* Respects concerns without requiring full agreement
	170	* "I can live with this and support it"
	171	Process:
1.2	172
1.1	173	1. Proposal presented (RFC)
	174	2. Clarifying questions
	175	3. Reactions and concerns
	176	4. Proposer integrates feedback
	177	5. Consent round: Any principled objections?
	178	6. If no objections → Decision made
	179	7. If objections → Integrate and repeat
	180	Use for:
1.2	181
1.1	182	* Algorithm changes
	183	* Policy updates
	184	* Infrastructure investments
	185	* Process changes
	186	Not for:
	187	* Strategic decisions (use voting)
	188	* Emergency decisions (use autonomous authority)
1.2	189
1.1	190	== 7. Transparency & Accountability ==
1.2	191
1.1	192	Always public: Policies, structure, board membership, financials, quality metrics, major decisions
	193	Published quarterly: Activity reports, metrics, moderation stats
	194	Internal documentation: All meetings, decisions, actions retained indefinitely
1.2	195
1.1	196	== 7. Conflict Resolution ==
1.2	197
1.1	198	User disputes: Moderator → Appeal to different moderator → Governing Team (final)
	199	Timeline: Reasonable timeframe
1.2	200
1.1	201	== 8. Moderation Oversight ==
1.2	202
1.1	203	Moderator requirements: high reputation, 6+ months active, clean record
	204	Review: Quarterly by Community Coordinator, annually by Governing Team
	205	Appeal: Any user can appeal promptly
1.2	206
1.1	207	== 9. Code of Conduct ==
1.2	208
1.1	209	Governing Team: Act in org interest, disclose conflicts, maintain confidentiality
	210	Team Members: Follow procedures, professional conduct, document decisions
	211	Moderators: Impartial decisions, respect privacy, respond timely
1.2	212
1.1	213	== 10. Amendment Process ==
1.2	214
1.1	215	Minor changes: Governing Team decision
	216	Major changes: General Assembly (2/3 vote)
	217	Emergency: Governing Team can act, must ratify at next Assembly
1.2	218
1.1	219	== 11. Related Pages ==
1.2	220
1.1	221	* [[Organisational Model>>FactHarbor.Organisation.Organisational-Model]]
	222	* [[Legal Framework>>FactHarbor.Organisation.Legal-Framework]]
	223	* [[Transparency Policy>>FactHarbor.Organisation.How-We-Work-Together.Transparency-Policy]]