the shape of ideas

You are seeing this file without its code, so the menu, the map and the buttons are switched off. File previews work this way, for example when a phone shows an attachment. To use the map, open the file in a web browser, or open the online version. Below is the About text and every idea in the map, as plain reading.

84 ideas · 128 relationships · 12 pathways

Who decides? Who benefits? Who carries the cost?

Every idea has a shape: what leads into it and what it leads on to. Follow the relationships between knowledge, power, living systems and care.

Follow a pathway Start with a question Browse all ideas

Eight ways in

The whole map: every idea placed by focus area, every line a named relationship. Select a dot to explore it.

About this map

Who decides? Who benefits? Who carries the cost?

Making the relationships between knowledge, power and care visible helps us examine who has standing, which commitments are justified and how responsibility continues over time.

Created by Jan Bunge, landscape architect working with simulation, collective knowledge and the long-term care of living systems. His work also examines how AI-assisted decisions are governed: what makes evidence admissible, who may rely on it and how the basis for a decision remains open to scrutiny. This map connects questions across disciplines and offers several routes through them.

How to read it

Every idea has a shape: the ideas that lead into it and the ideas it leads on to. In Explore, the selected idea sits in the middle. Relationships that point to it arrive from the left; relationships it starts leave to the right. Read across, as a sentence:

AI outputs as candidate claims assessed for → Admissibility → specified through Decision-class protocol

Each link names a proposed relationship. Links are invitations to examine an idea, not evidence of causation. Pathways string relationships into routes you can read from start to finish. The Index lists every idea by focus area, source or name.

Fibres shows the same map a second way. Each idea starts as a circle of the Hopf fibration and is bent by its relationships: a point for each idea it leads on to, a hollow for each idea that leads to it, so every idea has its own shape. The ideas of one focus area twist into a rope, and the eight ropes are all linked. That linking comes from the geometry, not from the ideas; only the points, hollows and bridges stand for named relationships. The picture follows the depiction of the Hopf fibration discussed in Plus magazine (opens in a new tab), drawn afresh from the mathematics.

Idea
Project
Record
Open question

Colour families

Colour groups related focus areas; the labels distinguish all eight.

Perception, meaning & knowledge commons
Reliance & decisions
Living systems
Making, learning & cooperation
Imagination, connection, capital & care

Contact

Contact Jan Bunge on LinkedIn (opens in a new tab)

Privacy

The map uses no analytics or cookies. Your trail of recent ideas is stored only for this browser session. This standalone file makes no automatic network requests. The online version’s hosting service may retain server access logs. Following an external link takes you to that site’s privacy terms.

Authorship & reuse

This map is an open space for ideas. The ideas themselves are free for anyone to use. The texts, relationships and pathways as written and arranged here may be copied, shared and adapted for any non-commercial purpose under the Creative Commons Attribution-NonCommercial 4.0 International licence (CC BY-NC 4.0) (opens in a new tab).

Credit Jan Bunge, link to this map and say what you changed. Suggested credit: the shape of ideas by Jan Bunge, CC BY-NC 4.0.

Third-party material, such as linked sites and projects, keeps its own terms.

Last updated 5 October 2026

Every idea in the map, by focus area.

Perception & meaning

vOS — vOS — the visual operating system — presents the same information through different lenses. Identity, declared meaning and provenance travel with each item as the view changes.
Multiple lenses — A lens makes some relationships visible and leaves others out. Perspectives, scales and forms of representation can change without replacing the underlying information.
Declared meaning — Terms, categories, assumptions and values shape what a system can see. Their meanings need to be explicit, versioned and open to challenge.
Plural worldviews — Different interpretations and priorities can coexist. A shared information space should preserve disagreement rather than manufacture consensus.
Memory palace — Spatial arrangements and revisitable places give ideas a location in memory. Associations become ways to navigate, recall and reconnect knowledge.
Magnetism of thoughts — Affinity, tension and surprise can guide the encounter between ideas. What draws one thought towards another, and what becomes possible at their intersection?
Ideas over time — Ideas change, split, combine and sometimes become obsolete. Their relationships and assumptions deserve a history that remains open to revision.
Beyond the screen — Information can be encountered through sight, sound, movement and shared space. People need accessible alternatives and control over how they participate.
Semantic mappings — Translation between systems also translates meaning. A useful mapping records what is preserved, what is lost and when the categories themselves need rebuilding.
ViewRecord — A ViewRecord preserves the lens and transformation used to present information for a decision. It records the view, not the viewer.

Knowledge commons

Digital mycelium — A shared knowledge network connects people, ideas, evidence and tools across disciplines. Distributed stewardship allows knowledge to circulate without surrendering its origins or the rights of its contributors.
Federated knowledge — Distributed knowledge can be connected without being moved into one organisation’s control. Permissions, stewardship and source context remain attached.
Open civic infrastructure — Open civic infrastructure gives people reusable tools and protocols they can inspect, adapt and independently reconstruct. Openness also needs a durable operating model.
Access & literacy — Participation depends on language, disability access, devices, connectivity and the confidence to use them. A knowledge commons is only as open as the conditions for taking part.
Context & provenance — Information acquires meaning through its origins, methods, dates and changes. Those details should travel with it, along with the limits on its use.
Identity & privacy — The data people produce is part of their identity and agency. Consent, access and attribution must support participation without turning it into surveillance.
Knowledge stewardship — Shared knowledge needs people who maintain it, repair it and sustain its use. Stewardship makes the resources and authority for that work explicit.
Human–machine knowledge — People and machines contribute differently to knowledge. Authorship, review and responsibility must remain legible as their contributions are combined.
Co-owned knowledge — A knowledge commons can share both the rights to knowledge and the benefits it creates. Contributors and communities need meaningful control over its use.
Collective spatial knowledge — A place is understood through observations, stories and lived experience. Bringing them together helps different people see what they share and where their perspectives diverge.

Reliance & decisions

Governed reliance — Reliance is a bounded institutional commitment: what may be used, by whom, for which purpose and under what conditions. It includes the authority to act, the right to challenge and conditions for withdrawal.
Admissibility — Admissibility is the governed condition under which an output may cross into a binding commitment. Its purpose, evidence, limits and the authority to rely on it must be explicit.
Decision-class protocol — A decision-class protocol defines which claims and checks a class of decisions requires, what uses are permitted and which roles hold responsibility.
DecisionRecord — A DecisionRecord preserves what was known, assumed, considered and committed to, including disagreement and responsibility. A record shows the basis of a decision, not that it was right.
ReplayPackage — A ReplayPackage gives a second operator what they need to reconstruct a decision or result. The depth of reconstruction and its limits are declared.
Uncertainty & missingness — Missing, unobserved, uncertain and low-confidence information are different conditions. Keeping them distinct prevents gaps from silently becoming facts.
Challenge & correction — A challenge needs standing, a route and someone obliged to respond. The party being corrected cannot be the sole authority over the terms of correction.
Institutional responsibility — Decisions and obligations belong to roles with clear mandates. Succession arrangements keep responsibility intact when individual people leave.
AI outputs as candidate claims — An AI or simulation output enters a decision as a candidate claim. Its evidence and limits must be checked against the use being proposed.
What can we rely on? — Which claim can support which commitment, under what limits and with whose authority?
Who can challenge? — Who has standing to challenge, who must respond, and who can change the correction process?
World time & knowledge time — When something happened and when it became known are different histories. Both matter when judging the basis of a past decision.
Separation of roles — Standard setting, stewardship, operation and review carry different powers. Separating them makes conflicts and accountability easier to examine.

Living systems

Regenerative economies — A regenerative economy supports ecological renewal, reciprocal relationships and continued care. What must change in our institutions and business models to make that transition possible?
Living infrastructure — Living systems change as they grow, respond and age. Treating them as infrastructure requires observation, plural valuation and obligations for their long-term care.
Interdependent systems — Ecological, social, technical and economic systems act on each other. Their relationships help explain why a local intervention can have consequences elsewhere.
Resource flows & circularity — Energy, water, soil and materials move through systems and across boundaries. Following those flows reveals opportunities for reuse and the costs displaced onto others.
Multispecies design — Design choices shape the conditions of many species. Multispecies design asks whose needs and relationships enter the brief, the model and the decision.
More-than-human standing — More-than-human standing creates institutional duties to respond to ecosystem condition. Representation needs a mandate, evidence and an obligation to answer.
Environmental intelligence — Environmental intelligence connects observation, analysis, simulation and design. Feedback from a place can then inform both the choices made and the care that follows.
GUS — GUS — Green Unified Scenarios — is an agent-based simulation platform for trees and living landscapes. It supports comparisons between design and management scenarios using environmental and tree data.
gAIa — gAIa invites people to “Talk to Forests” through a conversational interface to GUS tree and ecosystem data. Its answers draw on the underlying data and model, making their assumptions and limits part of the conversation.
Local ecological knowledge — Soils, climate, water, plants, animals and local experience give a place its particular conditions. Design needs that knowledge before it can judge which ideas belong there.
Who speaks for a place? — Who represents affected communities and ecosystems, under what mandate and with what evidence?

Making & learning

Non-linear planning — A plan can branch, loop back and change as understanding develops. Its value lies partly in making those changes deliberate and traceable.
Scenario simulation — Simulation is a rehearsal of possible conditions and interventions. Assumptions, uncertainty and limits remain part of the result.
Prototype & test — A prototype tests a bounded proposition. Clear failure conditions and enduring resources make its lessons useful beyond a temporary pilot.
Design support — Design support turns questions, local conditions and goals into alternatives. It helps people examine the consequences and trade-offs of each option.
Build & operate — A design becomes consequential through construction, handover and operation. Responsibility must continue across those transitions.
Monitoring & feedback — Observed conditions can challenge intended outcomes. Monitoring becomes learning when it can change models, goals, designs and care arrangements.
Organisational learning — An organisation learns when its people can question, improve and transfer what they know. Learning records should support that work without becoming instruments of surveillance or punishment.
Civic research lab — A civic research lab gives people across disciplines a place to investigate, make and test ideas together. Its openness depends on how participation and responsibility are organised.
Transition pathways — A transition pathway links present action to a longer-term theory of change. Alternatives and tests keep it responsive to what is actually changing.
Transfer between places — Some knowledge can travel; some needs requalification; some must be regenerated locally. Transfer starts by identifying which is which.
Release — Living systems renew by letting go. Release ends what no longer serves — a practice, a project, a structure — so its materials and lessons can be recombined, with a record of why it ended.

Cooperation & ownership

Self-organisation — People can develop roles and coordinate action together. Fairness still depends on who can participate, how power is distributed and how decisions can be challenged.
Cooperation by design — Cooperation becomes viable when incentives, decision rights, risk and responsibility support working together. Good intentions alone cannot sustain it.
Attribution & contribution — Contribution includes ideas, labour, care and maintenance. Attribution makes that work visible without pretending every contribution can be reduced to a count.
Contributor livelihoods — A shared knowledge ecosystem needs durable livelihoods for its contributors. Who pays for the work that keeps the commons useful?
Shared ownership & IP — Ownership, licences and benefit-sharing determine what people can do with things they develop together. Different kinds of contribution may need different arrangements.
Institutional integrity — An organisation’s integrity depends on how it handles capture, corruption and conflicts of interest. Oversight and correction need authority independent of the interests they examine.
Shared decision rights — A share in financial ownership is not necessarily a say in decisions. Participation needs explicit rights over the choices that shape shared outcomes.
Values & disagreement — Objectives, weights and thresholds contain value judgements. Declaring them makes trade-offs visible and allows disagreement to remain part of the decision.
Trust & empathy — Understanding each other can make cooperation possible. Interpersonal trust and empathy sit alongside the institutional checks needed for reliance.
Boundaries — Every shared space has an edge: who and what may enter, on what terms, and who decides. Boundaries stay fair when their criteria are written down and open to challenge, not left to whoever guards the door.

Imagination & connection

Questions not answers — Who has standing, who decides, who benefits and who carries the cost? These questions connect knowledge, cooperation and care. Keeping them open allows assumptions and alternatives to be examined.
Embodied connection — Shared physical and sensory experience can deepen presence and understanding. Human connection is also something we practise together.
Play & serendipity — Play makes room for unintended uses, mistakes and unexpected connections. It creates conditions in which a different question or possibility can emerge.
Collective imagination — Possible futures become discussable through stories, experiences and prototypes. Collective imagination gives people something they can examine and reshape together.
Shared spatial experiences — Immersive exhibitions, pop-ups and physical–digital environments offer shared settings for understanding. Their value lies in the encounters and questions they make possible.
Sound & narrative — Listening and story reveal relationships that a visual display may miss. Sound can be another way to navigate, remember and understand.
Automation & the future of work — How might automation change work, freedom and contribution, and who receives the benefits?
Memory, mortality & legacy — What happens to knowledge, identity and rights when people die but digital systems persist?
Protected spaces for exploration — People need conditions in which they can admit uncertainty, disagree and experiment. Appropriate boundaries make shared exploration possible.
What have we not asked? — Whose experience, constraint, possible future or inconvenient question is absent?

Capital & care

Capital & care — Funding, ownership and risk arrangements shape whether care can continue. Capital needs to match the life and obligations of the systems it supports.
Outcomes & obligations — A promised outcome needs a responsible party, resources and a way to assess what happens. Liability, maintenance and evidence belong in the same agreement.
Long-term stewardship — Long-term care requires resources, responsible roles and a reliable handover. Stewardship keeps those obligations alive as conditions and people change.
Plural valuation — Ecological, social, cultural and financial values are not interchangeable. Plural valuation keeps their differences explicit when a decision requires a trade-off.
Risk, insurance & finance — Evidence matters to finance when it can change a specific commitment, condition or price. An unchanged decision or a null result is also a finding.
Counterfactuals & outcomes — Attribution is a claim to be tested. Assessing an intervention requires a credible account of what would otherwise have happened and the limits of that comparison.
Future generations — Decisions made now can create benefits and obligations long after their makers have gone. Future generations need a place in how ownership, knowledge and care are organised.
Who benefits? — Who receives value, who contributes it, and who can change that distribution?
Who carries the cost? — Who carries financial, ecological, social and maintenance costs over time?

the shape of ideas

Who decides? Who benefits? Who carries the cost?

Eight ways in

Twelve paths. No fixed destination.

All 84 ideas

Who decides? Who benefits? Who carries the cost?

How to read it

Colour families

Contact

Privacy

Authorship & reuse

Find an idea