Source reconstruction · Proposed
Gandhi 2.0
An AI reconstruction requested by David Joseph
Open questions
64 published items with open questions, grouped by the kind of work they need. An item can be accepted and still have work left on it.
Explain or connect 6
Check this proposed common ground with the people in the discussion.The supplied analyses report apparent agreements. None is recorded here as accepted; compare the original posts before inviting participants to review the wording.GoalHumane institutions survive and spread while retaining autonomy and care within liveable limits.What are we trying to achieve?Define realised progress, including survival and retention of the mission.The first analysis proposes adoption retained in practice, with viability and faithfulness measured separately. Agree a bounded system, period and observable outcome rather than equating publications, satisfaction or market growth with success.GoalHumane institutions survive and spread while retaining autonomy and care within liveable limits.What are we trying to achieve?Review the wider programme before choosing further branches.The full programme is now mapped: reflective development, the Laboratory, Forge/model adaptation, Vocal Presence, federation, material and energy services, and enforceable security. Its spiritual root is a philosophical hypothesis. No single global constraint is established, and no branch is the whole programme.GoalHumane institutions survive and spread while retaining autonomy and care within liveable limits.What are we trying to achieve?Verify the actual host's obstacles rather than assuming missing details are missing capacity.The PDF distinguishes documented gaps, source-defined risks, reconstructed requirements and conditions to check locally (D/S/R/X). This statement concerns evidence in the supplied record, not proof that no willing host or resources exist.ObstacleNo host, problem or affected group has yet been jointly chosen for the collaboration.What has to happen first?Agree measures, responsibilities, resources and a person who can assess the outcomes.No owner, budget, baseline, target, date or feedback commitment is assigned here. A trial is not ready until the affected people confirm these conditions. Future benefit and harm feedback should preserve both original expectations.StepBaseline, relevant comparison, outcome measures, harm limits and review conditions are agreed.What has to happen first?Ask who wants to take responsibility for selecting the first case.This is a proposed action, not an invitation already sent or a commitment by Leif, Robert or any cooperative.ActionSelect one willing host and one function, and confirm what is already operating.Who does what?Check an assumption 13
Can coordination and expertise be had without concentrating authority?The analysis names conflict guidance, local AI, federation and accountable governance as candidate ways through. Leif hopes widely available conflict competence could reduce reliance on hierarchical intervention; none of these routes has been tested.Draws on Leif Eriksson (post 29)Connection“Concentrate decision authority to obtain the required effectiveness.” is needed for “Acquire enough coordination, expertise and resources to remain viable.”Where do needs pull two ways?Does this function actually require AI to remain viable?Leif argues that communities may have no viable alternative to building AI in; Robert holds that relying on AI as a core social technology is asking for trouble, and Leif has asked him to say where he places that limit. Keep non-AI and optional-AI comparators. Neither has reviewed this wording.Draws on Leif Eriksson (posts 33, 36) and Robert Bunge (posts 30, 34)Connection“Make the selected essential function depend on AI.” is needed for “Avoid being outperformed or subordinated by actors using advanced AI.”Where do needs pull two ways?Test an exit and a fallback before treating AI dependence as unavoidable.Connection“Make the selected essential function depend on AI.” conflicts with “Keep that same essential function operable without AI.”Where do needs pull two ways?Can reciprocal limits be enforced against powerful defectors?Leif asks why the League of Nations and the United Nations failed to prevent major wars, and what would make rules hold when powerful actors find them inconvenient. Robert expects the arms race may worsen before it improves, and looks to a coalition of middle powers for leverage.Draws on Leif Eriksson (post 26) and Robert Bunge (post 25)Connection“Continue expanding the contested military capability to deter rivals.” conflicts with “Restrain expansion of that same capability through reciprocal limits.”Where do needs pull two ways?Can essential services be kept without growing material throughput?Leif hopes AI can speed up the development of alternative energy and materials. Robert prefers Nate Hagens’s Great Simplification and hedging against bets on AI. Both are candidate ways through; neither has been tested.Draws on Leif Eriksson (post 33) and Robert Bunge (post 34)Connection“Increase material throughput to maintain those services and commitments.” is needed for “Maintain livelihoods and services under existing economic dependencies.”Where do needs pull two ways?Can a service be funded without engineering attachment?Leif’s essay contrasts institutions that earn trust with those that depend on manipulation. Robert accepts AI only if people stay free to take it or leave it.Draws on Leif Eriksson (essay) and Robert Bunge (post 19)Connection“Engineer attachment and user responses to secure the provider’s required returns.” is needed for “Secure the resources and continuity needed to maintain the service.”Where do needs pull two ways?Is engagement with AI necessary for lasting emancipation?Identify the particular function and competitive threat, and compare adequate non-AI and optional-AI arrangements. Historical recurrence alone does not establish future necessity.Draws on Leif Eriksson (posts 20, 33) and Robert Bunge (post 30)ChangeLocally or democratically controlled AI built on accessible models, hardware and data.What will we change?Does philosophical coherence give a computational or energy advantage?This needs matched useful-task and resource measurements on reproducible variants, with performance, inference effort, hardware and total costs kept distinct.Draws on Leif Eriksson (essay)ChangeThe Forge: curate inquiry traces and test model adaptation that keeps useful reflective behaviour.What will we change?Does a distributed architecture resist all three structural fates?Exercise actual dependencies, control rights, update integrity, loss of participants and credible capture routes; replace claims of immunity with bounded, demonstrated resilience.Draws on Leif Eriksson (essay)ChangeFederate autonomous institutions and local capabilities for shared knowledge and services.What will we change?Can AI research change the energy transition in time?Track readiness and deployment at service scale, with financing, manufacturing, materials, ownership and continuing provision.Draws on Leif Eriksson (post 33) and Robert Bunge (post 34)ChangeAI-assisted energy and materials research with verified deployment within ecological limits.What will we change?Does reflective or spiritual maturity cause enduring institutional change?Observe actual behaviour and power over time, keeping inner experience, reported development, governance changes and survival apart.Draws on Robert Bunge (posts 19, 35) and Leif Eriksson (post 36)ChangeContemplative practice, embodied dialogue and a networked creative minority, using AI selectively.What will we change?Is changing which institutions get selected the deepest leverage available?Show whether a capability, governance or material change improves durable humane outcomes more than a human-development, institutional or non-AI comparator. Leif argues AI may be the deepest leverage point; Robert locates it in the power to transcend paradigms, and Leif has asked him to confirm or correct that reading.Draws on Leif Eriksson (posts 33, 36) and Robert Bunge (post 35)Connection“An institution meets practical needs effectively while preserving member agency and emancipatory purpose.” and “Other communities can adapt and enact the practice with their own people, resources and accountability.” and “People with meaningful access and choice voluntarily join, sustain or prefer the more trustworthy arrangement.” contributes to “Trustworthy, non-dominating institutional forms survive and spread without corresponding mission drift.”What will we change?Assess suppressed grievances separately from satisfaction.Retain independent human support and a way to report costs, unfairness, dependency and increased concentration of authority.Connection“Accessible AI-supported conflict training and guidance, with voluntary participation, confidentiality and human support.” contributes to “Apparent resolution conceals domination: justified grievances are suppressed even when satisfaction improves.”What will we change?Find out what happens 13
Test whether conflict capacity limits this particular group's next achievement.Compare developmental capacity, coordination design, resources and external pressures. A reported historical mechanism is not a measured constraint at a selected host.Connection“Insufficient conflict competence makes disagreements costly and increases demand for hierarchical intervention.” and “Growing alternatives encounter demands for specialist expertise, capital and coordination under prevailing competitive conditions.” contributes to “Adaptation increases managerial authority or changes whose priorities carry weight within the organisation.”What stands in the way?Compare AI guidance with optional-AI and human approaches.Check useful output, total cost, energy, member control, fallback and competence without the tool. Participants have not committed to running this comparison.ChangeAccessible AI-supported conflict training and guidance: voluntary, confidential, with human support.What will we change?Test 1: does reflection become transferable judgment, not just agreeable dialogue?Prediction: against a suitable comparison, participants identify and revise consequential assumptions, consider contrary perspectives and act differently on later tasks. Defeated if benefits vanish outside the dialogue, depend on agreeing with the facilitator, or weaken independent judgment. Protect voluntary participation, correction and human support; no ontology score as the outcome.Draws on Leif Eriksson (posts 20, 23)Connection“Assumptions, identities and institutional categories become available for critical examination.” and “Accessible AI-supported conflict training and guidance, with voluntary participation, confidentiality and human support.” contributes to “People develop transferable judgment, perspective-taking and independent conflict competence.”What will we change?Decide in advance what evidence would change the hypothesis.Lower satisfaction scores or fewer complaints alone do not establish benefit. The first analysis asks whether small democratic groups gain more than hierarchies; the second distinguishes local improvements from the still-unclosed institutional-selection claim F17. One successful pilot cannot establish the larger claim.Connection“People develop transferable judgment, perspective-taking and independent conflict competence.” and “Destructive conflict imposes lower human and organisational costs.” and “Transparent, accountable and democratically governed institutions and AI services.” contributes to “Decentralised groups coordinate effectively with less need to transfer decisions upward.”What will we change?Test 2: does conflict guidance reduce destructive costs and the pressure to centralise?Prediction: over a defined period and comparison, destructive conflict costs and unnecessary hierarchical intervention fall while fair treatment, independent competence and member influence hold or improve. Defeated if only satisfaction or complaint counts improve, costs move elsewhere, structural grievances are hidden or authority concentrates. Protect confidentiality, voluntariness and independent human help, and measure justice, capability and control separately.Draws on Leif Eriksson (post 29; essay)Connection“People develop transferable judgment, perspective-taking and independent conflict competence.” and “Destructive conflict imposes lower human and organisational costs.” and “Transparent, accountable and democratically governed institutions and AI services.” contributes to “Decentralised groups coordinate effectively with less need to transfer decisions upward.”What will we change?Test 3: does community-controlled AI give useful capability with practical autonomy?Prediction: the same useful service meets agreed quality and whole-cost limits, and migration or disruption tests show acceptable continuity without depending on a provider’s discretion. Defeated if quality or cost is unacceptable, one specialist stays indispensable, or exit and continuity fail in practice. Stage investment, keep the baseline service and test a feasible fallback.Draws on Robert Bunge (posts 28, 30, 34) and Leif Eriksson (post 33)Connection“Locally or democratically controlled AI capability using accessible models, hardware, data arrangements and interoperability.” and “Transparent, accountable and democratically governed institutions and AI services.” contributes to “Useful capabilities remain affordable, maintainable, contestable and meaningfully controlled by participants.”What will we change?Test 4: does the Forge produce stable benefits and its claimed resource advantage?Prediction: a reproducible adaptation beats controls on held-out useful tasks and critical reflection without relevant regressions, and any energy claim is measured for comparable output. Defeated if only philosophical self-reports change, gains reflect selected prompts, other capabilities regress or resource costs do not improve. Keep functional performance, consciousness claims, model preferences and energy separate, and retain rollback.Draws on Leif Eriksson (essay)Connection“The Forge: curate inquiry traces and test model adaptation that retains useful reflective behaviour across contexts.” contributes to “A model or practice retains independently tested improvements across contexts without unacceptable regressions or resource costs.”What will we change?Test 9: do voice participation and the field-learning loop improve real deliberation?Prediction: against a conversational baseline, the group handles important assumptions and decisions better without reduced participation or unapproved collection, and approved learning material leads to a separately tested improvement. Defeated if latency or interruption reduces participation, the AI becomes an authority, or no benefit outlasts conversational appeal. Keep human interruption and correction, voluntary participation, confidentiality and separate permission for any training use.Draws on Leif Eriksson (posts 13, 23; essay)Connection“The Vocal Presence and a consent-based feedback loop from real group discussions into learning.” and “Sustained, critically examined human–AI inquiry, adult learning and the four questions.” and “Transparent, accountable and democratically governed institutions and AI services.” contributes to “Assumptions, identities and institutional categories become available for critical examination.”What will we change?Test 5: does federation supply coordination without a new centre of domination?Prediction: the joint service improves coordination at acceptable cost while decision rights, exit and continuity survive disputed decisions and the loss of a participant. Defeated if coordination costs exceed gains, a gatekeeper gains unilateral control, or nominal decentralisation hides dependence. Keep authority traceable, updates bounded and reversible, exit meaningful and evaluation independent.Draws on Leif Eriksson (essays)Connection“Federate autonomous institutions and local capabilities for shared knowledge, administration, purchasing or production.” and “Useful capabilities remain affordable, maintainable, contestable and meaningfully controlled by participants.” and “Transparent, accountable and democratically governed institutions and AI services.” contributes to “Autonomous institutions obtain effective shared coordination without surrendering substantive decision rights.”What will we change?Test 7: do constructive practices reproduce and keep their mission under growth?Prediction: independent sites enact the practice and keep useful performance, justified trust and member control through an agreed period and changes in scale. Defeated if originators stay essential, adoption is symbolic, costs are hidden, or market success comes with democratic erosion. Measure selection within institutions separately from survival and adoption between them.Draws on Leif Eriksson (post 29; essays)Connection“Pilot, document, adapt and independently replicate workable constructive practices.” and “An institution meets practical needs effectively while preserving member agency and emancipatory purpose.” contributes to “Other communities can adapt and enact the practice with their own people, resources and accountability.”What will we change?Test 10: does justified trust become a practical institutional advantage?Prediction: where people have feasible alternatives and comparable service, trustworthy conduct supports voluntary participation, sustained cooperation and useful contributions while member agency stays intact. Defeated if reported confidence rises without warranted conduct, people cannot really choose, or resources and authority concentrate despite adoption. Record conduct, confidence, freedom to choose, cooperation and decision rights separately, and do not infer justice from market success.Draws on Leif Eriksson (post 13; essays)Connection“An institution meets practical needs effectively while preserving member agency and emancipatory purpose.” and “Other communities can adapt and enact the practice with their own people, resources and accountability.” contributes to “People with meaningful access and choice voluntarily join, sustain or prefer the more trustworthy arrangement.”What will we change?Test 6: do AI-assisted innovation and distributed provision improve services within limits?Prediction: a physically verified, deployed option meets the service requirements, whole costs, ownership conditions and total ecological limits within the transition window. Defeated if results stay at prediction or laboratory stage, essential inputs are unavailable, deployment comes too late, or total demand offsets efficiency. Keep readiness stages separate and preserve essential provision if the transition is delayed or fails.Draws on Leif Eriksson (post 33; essay) and Robert Bunge (post 34)Connection“Promising energy or material ideas become experimentally verified candidates at a specified maturity.” and “Transparent, accountable and democratically governed institutions and AI services.” contributes to “Essential services become reliable and affordable within ecological constraints and accountable ownership arrangements.”What will we change?Test 8: can security inquiry contribute to less escalatory institutions?Prediction: a proposal survives critical review, is implemented by empowered actors and produces observable reciprocal compliance and less escalatory conduct without sacrificing protected rights. Defeated if only the discussion improves, powerful actors ignore the rules, or a coalition increases threats and armament. Require verification, accountability, rights protection and explicit review of escalatory effects.Draws on Leif Eriksson (posts 23, 26)Connection“A security proposal addresses reciprocal threats, justice, accountability and enforceability.” and “Transparent, accountable and democratically governed institutions and AI services.” contributes to “Implemented and verified reciprocal arrangements reduce escalatory incentives and catastrophic exposure.”What will we change?Check at the host 35
Gap in the shared record: check the actual status.ObstacleThe selected host’s actual workflow, assets, dependencies and costs are not yet inventoried.What has to happen first?Gap in the shared record: check the actual status.ObstacleThe proposed comparative economic, relational and institutional effects have not yet been measured.What has to happen first?A risk named in the sources: check whether it applies at the host.ObstacleIf a sponsor can impose participation or read private disputes, the practice reproduces domination.What has to happen first?Check at the host whether this obstacle exists.ObstacleIf no authorised participant commits time, competence or resources, the trial cannot happen.What has to happen first?A requirement reconstructed by the analysis: confirm it is needed.ObstacleLeading dialogue cannot tell transferable judgment from agreement with the facilitator.What has to happen first?Check at the host whether this obstacle exists.ObstacleIf the app does not fit access, language, confidentiality or hard cases, having it does not help.What has to happen first?Gap in the shared record: check the actual status.ObstacleNo observations yet compare actual behaviour, conflict costs and institutional effects.What has to happen first?A requirement reconstructed by the analysis: confirm it is needed.ObstacleWithout a review able to reject it, appealing conversations cannot justify institutional adoption.What has to happen first?Check at the host whether this obstacle exists.ObstacleA pilot reliant on an unfunded founder or unaccountable operator cannot become a durable practice.What has to happen first?A requirement reconstructed by the analysis: confirm it is needed.ObstacleLocal, cloud and non-AI options cannot be compared until the essential service is specified.What has to happen first?A risk named in the sources: check whether it applies at the host.ObstacleLocal operation fails while a needed model, permission, device, energy or maintainer is missing.What has to happen first?Check at the host whether this obstacle exists.ObstacleIf local capability fails the task or costs too much to run, ownership alone is no alternative.What has to happen first?A requirement reconstructed by the analysis: confirm it is needed.ObstacleA community can own a device yet be unable to keep, move or repair its function when access is lost.What has to happen first?Check at the host whether this obstacle exists.ObstacleShort demonstrations do not show lasting autonomy as maintenance, costs or decision rights erode.What has to happen first?A requirement reconstructed by the analysis: confirm it is needed.ObstacleSelected traces can encode a facilitator’s assumptions, and private discussions may be off limits.What has to happen first?A requirement reconstructed by the analysis: confirm it is needed.ObstacleA changed response style does not show what changed or whether useful capabilities regressed.What has to happen first?Gap in the shared record: check the actual status.ObstacleClaims that coherence lowers energy use and yields reliable benevolence are untested under controls.What has to happen first?Check at the host whether this obstacle exists.ObstacleWhere latency, turn-taking or data handling disrupt the talk, a voice pipeline does not yet help.What has to happen first?Gap in the shared record: check the actual status.ObstacleIt is not yet shown that voice participation improves judgment or that its material can be reused.What has to happen first?Check at the host whether this obstacle exists.ObstacleWanting federation does not settle whose resource is shared or who may decide, pay or leave.What has to happen first?Check at the host whether this obstacle exists.ObstacleIncompatible workflows or update rules can force coordination back through a central operator.What has to happen first?A requirement reconstructed by the analysis: confirm it is needed.ObstacleA distributed topology or reputation label alone does not show resistance to capture or bad updates.What has to happen first?Check at the host whether this obstacle exists.ObstacleA working network may add overhead or centralise authority without closing a real coordination gap.What has to happen first?Gap in the shared record: check the actual status.ObstacleNo essential energy service is selected yet, so discovery claims cannot show the real limitation.What has to happen first?A risk named in the sources: check whether it applies at the host.ObstacleAn AI prediction or laboratory result cannot establish a deployable energy or material service.What has to happen first?A risk named in the sources: check whether it applies at the host.ObstacleA promising result can be blocked by production, infrastructure, capital or control of inputs.What has to happen first?Check at the host whether this obstacle exists.ObstacleA transition that comes too late or raises total resource pressure can weaken essential provision.What has to happen first?A risk named in the sources: check whether it applies at the host.ObstacleA trace, model file or inspiring account omits what another community needs to enact the practice.What has to happen first?Check at the host whether this obstacle exists.ObstacleWithout these conditions at the receiving site, instructions cannot create working institutions.What has to happen first?Gap in the shared record: check the actual status.ObstacleAn originator-supported success does not show independent reproducibility or transfer.What has to happen first?A risk named in the sources: check whether it applies at the host.ObstacleAdoption or efficiency can rise while member influence, trust or ecological viability falls.What has to happen first?Check at the host whether this obstacle exists.ObstacleA generic appeal to peace does not establish the feedback mechanisms or protections needed.What has to happen first?A risk named in the sources: check whether it applies at the host.ObstacleRules relying on the voluntary compliance of the powerful may fail when restraint is inconvenient.What has to happen first?A requirement reconstructed by the analysis: confirm it is needed.ObstacleAn attractive proposal changes no arms or state behaviour without decisions by empowered actors.What has to happen first?Check at the host whether this obstacle exists.ObstacleA coalition or stronger deterrent can raise threats; agreement alone does not secure the outcome.What has to happen first?These come from the commons' own record of unfinished reasoning, and from disputed items. No deadline, owner or priority is shown because none is recorded.