Designers 2026-05, Synthesis
Field-reading across 20 tracks, ~53,100 words. Observational register. Single-author notebook.
1. What I'm Sitting With After Reading 20 Tracks
The single thing that keeps surfacing across all twenty briefs is that the field of conscious systems designers is not one field. It is a set of methodology-tribes that have each, in their own dialect, arrived at the same underlying ontological move, that the relation, not the entity, is the unit of design, and have failed to recognize themselves in each other. Sociocracy calls it the circle. Permaculture calls it the guild. Multi-agent AI calls it the protocol. Bateson calls it symmathesy. Indigenous traditions call it kin. Friston calls it nested Markov blankets. None of them have a shared name for what they share, and so the convergence stays invisible to the people inside it.
What's striking, holding the twenty tracks at once, is how much of the live energy in 2026 is happening at the seams the tribes can't see. The cyberecology-shaped opening is real. The honest gaps are large. And the silences, across everything I read, are louder than the noise.
2. Who Conscious Systems Designers ACTUALLY Are in 2026
The first thing to say is that almost nobody in this audience self-identifies as a "conscious systems designer." That phrase belongs to the framing of someone trying to map them from the outside. They identify by methodology lineage, and the lineage is the tribe.
T01 surfaces this most sharply in the org/governance cluster. A practitioner in this world is a sociocracy facilitator, an S3 trainer, an Art of Hosting steward, a holacracy coach, a steward-ownership advisor, a Liberating Structures user, a Reinventing Organizations alumnus. Each lineage has its own conferences (Sociocracy For All 2026, the Liberating Structures Eindhoven Gathering June 2026, Art of Hosting trainings in Berlin and Vancouver), its own canonical books, its own credentialing ladder. Geographic centers cluster: Berlin and the Netherlands and Nordic countries for steward-ownership and sociocracy, Pacific Northwest and Bay Area for the older Teal scene, UK and Catalonia for Decidim and participatory governance, Aotearoa for the Loomio and Enspiral lineage. Latin America has a deep tradition (Buurtzorg-inspired, ToP/ICA International) that English-language discourse mostly misses. The serious practitioner advisor base for steward-ownership is small globally, under a hundred specialists by my read, even though Denmark has roughly a thousand steward-owned companies holding sixty percent of stock-market value.
T02 maps the ecological cluster as at least five overlapping sub-communities, each with its own credentialing logic, gathering rhythm, and unspoken pecking order. Permaculture-lineage designers organized around the IPC circuit (IPC16 in Thailand November 2026 is the first full-scale convergence since the pandemic-era reset). Regenesis-lineage practitioners running The Regenerative Practitioner Series cohorts. Holistic Management/Savory Hub network ranchers and EOV-certified land managers. Bioregional weavers organized through the Bioregional Weaving Labs collective (eight active European labs, three Bioregional Finance Facilities prototyping in 2026). Ecovillage and intentional-community designers organized through GEN and the European Ecovillage Gathering (Germany July 2026, "To Dare to Trust"). An individual practitioner often holds two or three of these identities loosely but rarely all five, and they cluster more by methodology than by bioregion, which is itself a tension worth naming, because their stated values say the opposite. Generationally there is an aging center (boomer permaculture founders, the Allan Savory cohort, early Regenesis) and a much younger periphery in their twenties and thirties, often without land. The middle is thin. Many in their forties left the field because it didn't pay.
T03 maps the tech/AI builder cluster as a Venn diagram of five sub-bubbles with no shared center. The open-weights/sovereignty wing (EleutherAI lineage, the small-language-model crowd, federated AI builders). The Indigenous-protocol-aware wing (orbiting the Indigenous Protocol and AI Working Group and Old Ways New, tiny in headcount but disproportionate in conceptual weight). The trustworthy/responsible-AI fellowship wing (Mozilla Fellows, Montreal AI Ethics Institute orbit, AJL alumni). The AI-for-nature/regen-ag/biodiversity wing (Living Models, Omdena's regen-ag stack, satellite-and-soil startups). The contemplative-tech wing (the post-Calm meditation-app cohort, plus solo builders trying to make tools that don't extract attention). T03's back-of-envelope on scale: low tens of thousands worldwide could plausibly read into this work, with maybe a few hundred who are nodes. Geographic skew: heavily North Atlantic plus Aotearoa NZ plus Australia, with growth in Berlin/Amsterdam/Lisbon (sovereignty plus EU AI Act gravity), Nairobi and Bengaluru (responsible-AI-for-development), and a small São Paulo/Mexico City strand around Indigenous data sovereignty.
What unifies the demographic picture across all three cuts is its precarity. Most practitioners in all three clusters are independent or run small studios; pure in-house roles exist (a few inside steward-owned companies, inside platform cooperatives, inside the values-and-evaluations layer of frontier labs), but the majority are practitioner-consultants doing the work and teaching the work. T09's economics confirm what the field-reading suggests: median Patreon earnings are about $150 a month, median Substack newsletter earns zero, the median full-time creator US 2025 earns $44,000, the median part-time creator earns under $5,000. The discourse rounds up to top performers and rounds down to platform totals; the median is the unforgiving middle. This is the economic substrate the audience is actually living inside.
What the field doesn't have is a name it answers to as a single thing. The cross-cluster identity is genuinely absent. A sociocracy practitioner may not see herself in "conscious systems designer" until someone she trusts inside the S3 world uses the phrase. A steward-ownership lawyer probably never will, and may experience the framing as a flattening of distinctions she has spent her career holding. The audience for the cross-cutting view is real but smaller than the audience for any single methodology, it is the seam-workers, the cross-cutters, the ones already restless inside their own tribe.
3. The Cross-Cluster Patterns That Repeat
A handful of patterns appear independently in three or more tracks, with the same shape and different vocabulary. They are worth naming because they are the field-level structure even though no one inside the field is yet naming them.
Methodology-tribal cohesion plus cross-cluster identity vacuum. This pattern shows up in T01, T02, T03, T08, and T15. Each tribe has worked out its own internal pedagogy, its own canonical conferences, its own pattern library. Across tribes there is no neutral ground. T15 puts it directly: almost every organisation that has tried to span clusters has either narrowed back to one cluster, formally closed (OuiShare 2024), entered "hibernation" (Berkana Institute), or been wound down by community vote (Cabin DAO 2025). T08 names the shared ontological move the tribes can't see in each other: "the relation, not the unit, is the object of design."
The AI-conversation silence inside each discipline. T01 surfaces it as a "live edge" with no carrier yet, the OrgaLog/Synchrony tools are bullish, the contemplative-facilitator crowd is uneasy, the space between is empty. T02 surfaces it as a regenerative-design field with a "low-tech, ancestral knowledge" identity that makes engaging AI tools feel like betrayal, with most practitioners quietly using ChatGPT for grant writing while publicly skeptical. T03 surfaces it as the contemplative-tech builders being "conceptually starved" with no positioned host for the conversation Headspace's Ebb (AI triage, January 2026) is going to force. T12 names it as "no defensible answer to AI" across the entire org/governance platform layer. T13 names it as the absent doors between climate-tech MRV (Pachama, Funga, Cambium) and ecological-design pedagogy. The same silence, three to five times, across clusters that don't read each other.
The cyberecology-shaped opening repeating. T03 spells it out: nobody in the relational-alignment crowd, the small-model-sovereignty crowd, or the Indigenous-protocol crowd has named the underlying ontological frame. T08 finds the same gap from a different angle, noting that the Bateson-Yunkaporta-Kimmerer-Friston-IIT convergence still lacks a common vocabulary, and that the Friston/Active Inference architecture work is reading as cyberecology in mathematical clothing while almost nobody in the conscious-design field is reading it. T14 puts the same finding institutionally: across the eight strata of conscious-tech hubs, ecology-as-ontology-of-AI is not the organising frame for a single platform.
Durability comes from being plumbing, not pitch. T07 is the cleanest articulation: the conscious-tech projects that survived to 2026 are infrastructural, not narrative. NatureMetrics (eDNA bolted onto TNFD/CSRD reporting), Local Contexts (TK Labels traveling on the IEEE 2890.1 standard adopted January 2026), Inrupt/Solid (consumer pods abandoned, government data wallets succeeding), Common Approach (open standards under platforms rather than competing with them). The pattern repeats in T12 (Open Collective surviving by handing itself to a member-governed consortium, OFiCo, late 2024), in T19 (Loomio's "quietly persistent" 15+ years as an Aotearoa worker cooperative without acquisition or pivot), and in T16 (Wikipedia's defensive posture as it gets eaten by the substitutes it fed). What survives is what becomes load-bearing for somebody else's mandatory workflow.
The regen-as-contested-vocabulary race. T17 lays it out directly: "regenerative" is being pulled simultaneously by indigenous-rooted relational practitioners, mid-tier sustainability consultancies, enterprise-tech vendors (Movate's January 2026 "Beyond Generative" piece defining regenerative AI as autonomous self-modifying systems), USDA program designers (the $700M Regenerative Pilot Program characterized by Beyond Pesticides as institutional greenwashing), blockchain protocols (the ReFi crash where KLIMA dropped 99.99%), and global agribusiness. Wahl, Hutchins, and Reed are openly worrying about cooptation; Forum for the Future has added "just" to the term as a defensive patch; Living Building Challenge keeps a measurable floor at the cost of staying small. Same pattern is visible in T02's regen-ag certification wars (Regenified, Rainforest Alliance's new Standard, Regenagri at 4.37M acres / 332k farms, EOV) and T03's vocabulary race for "regenerative AI." The window before the term becomes unusable is closing. T17 estimates it sits where "sustainable" was around 2008-2010.
Asymmetric cross-bridges. T18 maps them directly. Org-governance and ecological is the only bridge of the three where citation, methodology, and personnel cross in both directions in 2026 (sociocracy is the lingua franca of contemporary ecovillage governance; regen-business consultancies absorb bioregional grammar). Org-governance and tech/AI is mostly one-way and accelerating in the wrong direction (enterprise AI literature has appropriated the language of "governance" as a risk-and-compliance discourse, hollowing out the distributed-authority register). Ecological and tech/AI is the thinnest bridge, almost entirely one-way (climate-tech sells tools to farmers; conservation NGOs use Indigenous and place-based knowledge as training data, not co-author). The triangle of all three is constructed by attendees moving across spaces rather than by host organisations. The connective tissue is biographical, not institutional.
4. The Live Edges Where the Field Is Reaching Forward
What's actually new in 2026, work happening at the frontier of these clusters that's worth tracking. Five threads recur across the briefs.
AI-augmented governance. T04 names this as the most genuinely novel governance experiment cluster of 2026. The vTaiwan/Polis lineage is the most serious thread: the December 2024 AI governance methodology was presented to Taiwan's National Human Rights Commission in March 2025 with feedback already reflected in the draft AI Basic Act, with hybrid in-person plus AI-facilitated online deliberation explicitly designed to counter polarisation, and generative AI summarising Polis discussions and visualising theme clusters. The Collective Intelligence Project's Alignment Assemblies have shaped Anthropic and OpenAI behavioral guidelines through citizen assemblies. OrgaLog (framing itself as Digital Governance Infrastructure) and Synchrony are newer (2024-2026) tools specifically for sociocratic structure maintenance with AI-supported transparency. Gitcoin's 2025 Steward Council model (10-15 elected stewards every six months operating alongside GTC-holder voting) and the October 2025 arXiv paper DAO-AI: Evaluating Collective Decision-Making through Agentic AI in Decentralized Governance extend the conversation into on-chain governance. The honest limit, named in T04: vTaiwan reported AI tools "performed less reliably in Traditional Chinese, reflecting their lack of high-quality training data." The structural ceiling on AI-augmented deliberation across non-English languages is real.
Bioregional Finance Facilities prototyping in Europe. T02 and T13 both surface the BWL collective (Ashoka, Commonland, OpEPA) as the most institutionally mature attempt to invent the finance infrastructure place-based regen actually needs. Eight active European labs, twenty-six weaving teams, three BFFs in 2026. The 4 Returns framework (inspiration, social, natural, financial) functions as a lingua franca among European bioregional practitioners, giving finance, ecology, and culture commensurable categories without flattening them into money. Whether the BFFs work at scale is unproven; the patient-capital math T09 documents is constraining (RSF Social Finance's 3-year notes at 3.50% annualized, not the 15-30% IRR the BCG modeling for regenerative landscapes shows). But the prototyping is live and the pattern is worth tracking.
Steward-ownership in legislative motion in DE/NL. T04 and T05 both surface this. The German coalition agreement of April 2025 commits to introducing a distinct Verantwortungseigentum legal form with an "irrevocable asset lock," a project live since 2020. In the Netherlands a draft for a Dutch steward-ownership legal form was submitted to the Minister of Justice in January 2025. The form is in active legal construction in 2026, not settled. Manelli, Pek, Waldkirch et al.'s 2025 Business & Society paper Beyond Ownership As Usual marks the field's arrival in mainstream management research. Bosch's 1942-designed structure is still load-bearing in 2026; Patagonia's 2022 conversion has produced $180M cumulative to Holdfast since 2022 and $14.7M in FY May 2024-April 2025 to climate activists. The legal layer is genuinely maturing.
Te Hiku Media / Papa Reo as production-grade Indigenous AI. T03, T07, and T11 all cite this as the most coherent working example of kaitiakitanga-as-data-license, a license that explicitly does not permit commercial AI training without iwi consent. ~92% accuracy on a model built on the Whare Kōrero archive (1,000+ hours, some speakers born in the 19th century) plus a crowdsourced 300-hour labeled dataset gathered in days from 2,500+ contributors. Now packaged as Papa Reo, a platform other small language communities can use to gather their own data and train their own models without exporting it to Big Tech. T11 names the limit honestly: the model is portable in principle but extremely difficult to instantiate in practice without the equivalent of a thirty-year media archive and an iwi-level governance structure.
The Bateson lineage actively writing new theory. T10 is the cleanest finding here: Nora Bateson's Warm Data and Symmathesy work is genuinely live in 2026, with Warm Data Labs running at the RSA London (September 2025), Centre for Systemic Studies (May 2025), Old Jedburgh Library (April 2026), keynoting GDI's European Trend Day March 2025, and an active certified-host network producing facilitators. This is one of the few frameworks in the brief where the theory is still being written rather than just transmitted, the cold-data critique of AI is one of the few substantive philosophical engagements with the AI/data extraction problem from a living-systems frame. T08 reinforces it by naming the Friston/Active Inference architecture work as the only framework actively generating new design vocabulary in the brief, the others are in maintenance, transmission, or critique mode.
Underneath these five threads, two infrastructure layers are inflecting. T07 names the IEEE 2890.1 "Recommended Practice for Provenance of Indigenous Peoples' Data" adopted January 2026 as the first international standard requiring disclosure of Indigenous data in research outputs, Local Contexts moving from advocacy artifact to load-bearing piece of research infrastructure. T12 and T19 both name the Open Collective transition (placed under the stewardship of the Open Finance Consortium in late 2024) as a quiet but significant move from VC-backed startup to commons-governed utility, currently serving 2,500+ open-source projects. Both are infrastructure-altitude moves that will outlast their current cycle.
5. The Honest Failures and Closures
The field's success literature is loud and its failure literature is thin. Reading across the tracks, the failures cluster.
The post-2020 burnout pattern in flat orgs. T04 cites the 2024-2025 burnout literature consistently showing about 52% of employees reporting burnout, with mid-level employees highest at about 54%. The pattern in self-managing organisations is variant: burnout shifts from the manager layer to everyone, because role-and-circle structures distribute managerial cognitive load across all members without distributing the recovery time. Few self-management evangelists name this. T09 reinforces it: Buurtzorg's 8% overhead is real, and so is the cognitive load on nurses doing what HR, scheduling, and middle management used to do. The trade is genuine, not free.
Cabin DAO wind-down 2025. T07 and T15 both treat this as required reading. Four years in, the formal DAO structure was wound down by community vote, treasury distributed proportionally to token holders in USDC. The most-cited lesson: digital coordination tools could not absorb the friction of physical-place stewardship. Land, tenancy, neighbour relationships, local infrastructure, these did not respond to token-weighted voting. The pattern that survived is the directory of physical regen-coliving nodes; the wiring (DAO governance, token mechanics) was discarded.
Schumacher closure cascade. T04, T05, T13, and T15 all reference the August 2024 closure of Schumacher College's academic programme. The Dartington Hall Trust held the land; the College was an operating entity; when the College ran sustained monthly losses the parent trust could not underwrite them. The Satish Kumar Foundation acquired the College's name and IP in July 2025 but had not secured a site as of late 2025. T05's structural reading is the most useful: a land trust does not, by itself, make the operations on the land viable. T13 extends it: with Findhorn restructured after the 2021 fire and the loss of its educational programmes, and Schumacher closed, the European intellectual layer that used to surround intentional-community practice has no current home. GEN's new 2026 platform is being framed as connection infrastructure, not editorial infrastructure.
Mondragon cross-border failure. T04, T05, and T09 converge on the same finding. Mondragon's largest industrial coops have become "coopitalist hybrids", cooperative core in the Basque Country, capitalist subsidiaries with non-member wage workers abroad. No foreign subsidiary has been converted into a cooperative. The 2013 Fagor bankruptcy and the December 2023 exit of Ulma and Orona (two of the largest member coops) are documented stress points. The 2025 De Gruyter article frames the contemporary tension precisely. The cooperative model does not scale across borders without losing its cooperative form. Empirically settled.
Buurtzorg cost-shift, not cost-save. T04 surfaces the buried lede in the 2024 Journal of Organization Design paper and 2024 PMC scoping review: when adjusted for patient case-mix, the cost savings shrink to about €330/patient/year against a total ~€15,500/patient/year, because medical follow-up costs are higher under Buurtzorg. In 2013 Regional Care Organisations refused to pay the additional costs and Buurtzorg posted its first loss. The model works; the savings narrative is misframed.
Holacracy's process-trap. T04's most portable lesson from the entire post-2020 self-management literature: treating process as substitute for trust. When the constitution becomes the only legitimate channel for grievance, informal repair work is delegitimised, and people leave. Holacracy survived its own evangelism phase by quietly shedding orthodoxy (Zappos retained the circular structure, killed the most ritualised meetings, reintroduced managers, what one ex-employee called "Holacracy-lite"). The interesting cases are no longer "pure Holacracy" but the hybrid descendants, companies that ran Holacracy 2015-2020, kept the role/circle abstraction, replaced governance meetings with lighter facilitation.
The DAO/Web3 conscious-coordination strand largely over. T05 and T07 both name this. Solana's Jupiter and Yuga Labs both abandoned their DAOs in 2025 citing "governance theater." What survived from the ReFi wave: outcome-verification primitives (Hypercerts, now rebuilt on AT Protocol after the Ethereum wiring became disposable), contributor-recognition systems (Coordinape), retroactive public goods funding (Optimism RetroPGF). What largely failed: the bet that on-chain mechanisms would fund ecological regeneration directly (KLIMA dropped 99.99% by April 2026; Toucan's BCT fell from $8.60 to $0.08; Moss's MCO2 fell from $20.56 to $0.10). The directory pattern survived where the token-governance pattern died.
Synthesis Institute collapse 2023, still casting a shadow. T07 names this as required keeping in any 2026 conscious-tech inventory. The collapse of the gold-standard psychedelic retreat/facilitator-training provider is still cited in 2025-2026 industry retrospectives as the moment trust in the broader psychedelic-services category broke. Roughly 300 in-progress students were left stranded. The recurring lesson: a values-led organisation in a fragile market, scaling on retreat-booking revenue with no bridge funding, collapses fast when bookings dip.
The pattern beneath the patterns, said directly: the cross-cluster work is the work that doesn't survive at institutional scale. OuiShare closed (T15), Berkana entered hibernation (T15), Cabin wound down (T15), Schumacher closed (T15), Wellspring Philanthropic Fund is winding down by 2028. The institutions that span the clusters are structurally fragile. The narrowing-back-to-one-cluster pattern is empirically the survival move.
6. The Silences (Where the Field Doesn't Talk)
What's conspicuously not being discussed in 2026, across the twenty tracks. Seven recurring silences worth naming.
AI inside ecovillage discourse. T02 is direct: regenerative design has a strong "low-tech, ancestral knowledge" identity that makes engaging AI tools feel like betrayal, but most practitioners are quietly using ChatGPT for grant writing while publicly skeptical. The honest conversation about what AI could mean for place-based work is largely unheld. T13 reinforces it: GEN has not metabolised regenerative-design theory or AI/cyberecology questions. T18 maps the structural reason: ecological designers operate at place and timescale (a watershed, a ten-year succession); AI builders operate at platform and release cycle. The unit of work doesn't translate.
Class composition in regen design. T02 names this as largely unspoken. Most permaculture designers are downwardly mobile college-educated white people. Most ranchers in the Savory network are not. Most ecovillage residents are former professionals. The cluster's class composition shapes what gets discussed and what doesn't, and it's almost never on the table. The pyramid-critique of PDC economics (you fund your work by teaching the next PDC cohort) lives in the field as gossip and not in print.
Grief infrastructure. T02 surfaces this directly: many practitioners are doing this work having watched a place they love be destroyed, or having burned out of mainstream careers, or having lost a community. There's almost no infrastructure for that grief inside the field. Joanna Macy's Work That Reconnects sits adjacent but isn't really integrated. T10 extends it: the Two Loops "hospice" role is undertheorized and increasingly needed (organizational closure, climate grief, AI displacement). The hospice work is the work the field needs and largely refuses to name as work.
Post-credential learning. T02 names the post-PDC decade as improvised. T10 names the same gap: a designer can spend two years in Theory U and emerge with a transformed inner stance and no tools for the legal/governance/financial restructuring her organisation now obviously needs. T12 extends it institutionally: there is no bridge between leadership-development (Presencing Institute, Mobius) and structural change (Purpose Foundation, sociocracy academies). No platform holds these as a single design problem.
Capital-access cost in steward-ownership proponent literature. T05 and T09 both flag this evasion. Almost all proponent literature understates the cost of the asset lock. Designers choosing these forms should be told plainly that they are choosing slower growth and a smaller addressable capital pool, not just "different" capital. There is no public dataset on failed steward-ownership conversions; the success cases are loud, the abandoned conversions and post-conversion dividend disappointments are not in the literature. This is the single most important missing evidence base for designers recommending the structure.
The succession question in regen orgs. T13 names it as unfaced infrastructure: three of the most important pedagogies in the ecological cluster are organised around single founders (Geoff Lawton at PRI, Dr Elaine Ingham at Soil Food Web School who died February 2026, Pamela Mang and Bill Reed at Regenesis in their seventies). Movement-wide there is no shared apparatus for knowledge transfer past the founders. T10 extends it: Capra is in his late 80s, Luisi died in 2021, no clear successor framework has emerged from the lineage. T20 confirms it from the literature side: no successor book has displaced Designing Regenerative Cultures (2016), which is itself a tell, the field hasn't produced its next-generation textbook.
The binding-metaphysics question when indigenous principles enter design literacy. T08, T10, and T11 all converge on this. When a kincentric principle enters a sociocratic charter or a regen-AI protocol, does the binding obligation come with it, or does only the form travel? T08 calls this "the field's central evasion." T11 names the operational form: most engagement asks permission rhetorically and proceeds; the Honorable Harvest's "abide by the answer" is the test that distinguishes engagement from extraction, and most published "engagement" in conscious-design has no observable instance of a no being abided. T10 reinforces it: most designers using these frameworks have no sustained relationship with the lineages they cite. The frameworks travel as language; the protocols (kin obligations, place-bound transmission, ceremony) do not.
The integration of energy, watershed, and mantle-temperature reality missing from AI-sovereignty discourse. T03 names this as striking: the energy, watershed, and mantle-temperature framing of AI infrastructure is almost completely absent from the alignment debates. Even the "AI sovereignty" conversation treats it as compute geopolitics, not as ecology. T14 echoes it across all eight strata: ecology-as-ontology-of-AI is not the organising frame for a single platform. The cyberecology-shaped move, where infrastructure throughput (mineral, water, carbon) becomes a first-order design constraint, is structurally absent.
7. What's Adjacent to IoC's Vocabulary But Not Yet Connected
A repeated pattern in the briefs: independent arrivals at the same move that do not recognize each other. Naming these matters because the convergence is invisible to the people inside it.
SRAGI's three-tier framework arriving independently. T03 surfaces this directly. SRAGI ("SpiritRune Artificial Generated Intelligence") is an explicitly philosophical compass mapping degenerative/sustainable/regenerative across tech-bio-culture, the closest independent arrival at a three-tier extractive/sustainable/regenerative AI model that mirrors the three-tier distinction in IoC's foundation work. Different vocabulary, same architecture. They are not citing each other. The Movate "Beyond Generative" framing and the Regen AI Institute "cognitive alignment" framing are also taking the term in different directions, mostly without cross-citation. The vocabulary is up for grabs and being grabbed badly.
The relational-alignment paper trail without a shared name. T03 cites the 2026 Alignment Forum and arXiv work explicitly reframing alignment as "a relational and evaluative judgment that depends on whose values are prioritised." One author calls solving-alignment a category mistake; another proposes replacing the phrase entirely with "Successfully Navigating AI Co-evolution." The papers are converging at the EA Forum, the Alignment Forum, arXiv, and the Tandfonline "Values in science and AI alignment research" 2026 piece, but the relational camp is winning the philosophy and losing the funding, and they have no shared name for themselves yet. The slot for naming is open.
The Bateson-Yunkaporta-Kimmerer-Friston-IIT convergence. T08's most direct finding: there is no shared cross-cluster name for the relation-as-unit principle. S3 calls it "the circle." Permaculture calls it "the guild." AI calls it "the protocol." Bateson calls it "symmathesy." Indigenous traditions call it "kin." The absence of a shared name is itself preventing the convergence from being recognized. Cyberecology is one of the few terms attempting the cross-domain naming move, and its weakness in the literature is also its strategic opening, almost no one has claimed the slot.
The five sub-bubbles in tech-AI building parallel frameworks. T03 documents the five sub-bubbles (open-weights/sovereignty, Indigenous-protocol-aware, trustworthy/responsible-AI, AI-for-nature, contemplative-tech) sharing almost no common reading list, no common conference, no shared vocabulary. T14 confirms institutionally that Mozilla touches the ecology frame via "open infrastructure" but doesn't theorise it; CHT explicitly humanises it; AI for Good operationalises it as "AI for environment" use-case; DAIR comes adjacent through worker-and-community rootedness; Wisdom 2.0 brushes it via interbeing teachings. None of them is making the underlying ontological move that all of them are gesturing at.
Living systems migrating into psychology and care literature. T17 surfaces this as a marker worth tracking: MDPI's 2026 "Emotional Regulation as Relational Infrastructure: A Living Systems Perspective" indicates that "living systems" is being adopted in psychology/public health journals. The migration of vocabulary out of regenerative-design and into mental-health and care-infrastructure literature suggests "living systems" may have a longer half-life than "regenerative" as a precise term. This is a vocabulary moving into adjacency without anyone naming it as such.
The Pendleton-Jullian/Brown Design Unbound lineage and Berkana's Two Loops and Snowden's Cynefin all describing the same epistemic move (designing conditions for emergence rather than designing emergence) without cross-citation. T08 names this. Snowden's March 2026 12-part series re-grounding Cynefin in "anthropo-complexity" is a partial admission that the framework had been hollowed by overuse, but the cross-framework synthesis of why these three lineages keep arriving at the same move is not in the literature.
The Friston/Active Inference architecture work. T08 names Designing Ecosystems of Intelligence from First Principles (Friston, Ramstead, Kiefer et al., 2024) as a serious technical attempt to articulate a relational, multi-scale design grammar for AI that "reads as cyberecology in mathematical clothing." Almost no one in the conscious-design field is reading it; almost no one in the active-inference field is reading Bateson or Yunkaporta. The convergence is real and the bridge does not exist. T10 confirms: Active Inference is the only framework actively generating new design vocabulary in the entire brief.
The honest read: the convergence is happening across at least seven independent streams in 2026, with the same underlying ontological move, mostly without cross-citation, mostly without a shared name. The field is unified at ontology, fractured at vocabulary. The vocabulary slot is open and closing.
8. What's Already Here vs What Doesn't Exist Yet (Adjacent Platforms Mapped)
T12 through T16 give the cleanest possible picture of what already serves each audience cut and what nobody serves.
Methodology-specific hubs are strong. T12 maps the org/governance layer: Sociocracy For All runs the Sociocracy Academy with a one-year certification track and looks like a small university for one methodology. Sociocracy 3.0 runs its own Academy with Creative Commons-licensed pattern library. Greaterthan operates as a global self-managed steward-owned collective doing actual organizational change consulting. Purpose Foundation/Purpose Economy runs the steward-ownership network across DE/US/UK/NL with bi-annual global gatherings. The Presencing Institute runs Theory U/u-Lab MOOCs at 100k+ enrolment scale. The Inner Development Goals network has emerged as the most successful recent attempt to give "inner work" legitimacy in mainstream org settings. T13 maps the ecological layer: Permaculture Research Institute, Regenesis Group with The Regenerative Practitioner Series at 1,000+ practitioners across 40+ countries, Capital Institute and the Regenerative Communities Network, Project Regeneration, Bioregional Weaving Labs, GEN, Soil Food Web School, Land to Market/Savory, Demeter. T14 maps the conscious-tech layer: Mozilla Foundation Trustworthy AI plus MozFest, Center for Humane Technology with 17,000+ Foundations of Humane Technology participants from 130+ countries, AI for Good (UN ITU) with 50+ UN agencies as partners, Indigenous Protocol AI Working Group plus Old Ways New, DAIR plus the critical-AI lab cluster, Wisdom 2.0 and the contemplative-tech layer, HuggingFace and EleutherAI, Berkman Klein and the academic AI-and-society centres.
Cross-cluster hubs are thin and mostly aspirational. T15 is brutally clear on this. The cross-cluster hub is mostly a category mistake. Almost every organisation surveyed serves one cluster well and gestures at the others. The few that explicitly try to span all three do so by abstracting upward into "transformation" or "sensemaking" or "regeneration" or "metamodernity" until the abstraction stops being useful to any specific designer trying to build something on Tuesday morning. Perspectiva names the bridge well ("systems, souls, society") but serves thinkers and reflective practitioners more than makers. Life Itself's Sensemaking Studio is the most explicit recent attempt at multi-modal consultancy for the meta-crisis-aware client, but the bridge runs from sensemaking into organisations, not from ecological design into tech. Capital Institute/RCN crosses ecology and finance and community organising but tech is barely present. Ecoversities Alliance is the strongest educational cross-cluster network operating but doesn't reach the conscious-tech designer audience at all.
Sensemaking-platforms-for-designers are absent. T16's most direct finding: no one has built a sensemaking platform with both editorial voice AND structured cross-reference. Substacks have voice, no structure. Are.na has structure, no voice. PKM has structure but private. Wikipedia has structure but no voice. The combination is genuinely empty space. Are.na's 18K paying members at ~$5/mo, deliberately refusing AI integrations, ads, and algorithmic feeds, is the most philosophically aligned reference point in the entire research round, and what it does NOT do, synthesise, argue, take positions, is exactly what a sensemaking platform should do. The Liminal Web cluster (Perspectiva, Kainos, Resonant World, Sentiers) functions as a de facto sensemaking commons via Substack recommendation graphs, but there is no shared canon, no cross-referenceable corpus, no way to ask "what has this network collectively concluded about X."
The agentic research commons attempts are structurally invisible to the public web. T07 and T16 both note: no 2026 source confirms any project called "Bridging Worlds" by Ostara, or a "Possible Planet Lab," or "Localism Labs Bridging Worlds" in the agentic-research-commons space. Either these are private/pre-launch, or the names have changed. Murmurations is the most directly aligned existing infrastructure, a federated protocol for the regenerative economy with the CoBot AI search tool launching March 2026, but it solves directory discovery, not thinking. The interpretive layer over the federated data does not exist.
The honest gap in T16's vocabulary: the "single-author commons" is a category no one has defined. All commons literature assumes multi-author. But many of the most influential intellectual sites in the last twenty years have been single-author corpora that functioned as commons (LessWrong started this way, Slate Star Codex, Marginal Revolution, Stratechery). The category description that doesn't exist is the one that names what such a thing is, what makes it work, what kills it, what it owes its readers and the field it serves.
T19 confirms the tooling layer is bimodal and improvised. Better-funded conscious orgs (Dark Matter Labs, larger foundations) build custom mapping infrastructure; smaller practitioners use a Google Sheet, a Notion page, and a hand-drawn map. The middle is thin. Pattern-language decks (GroupWorks 91 patterns, Wise Democracy Pattern Language, Liberating Voices, Sociocracy 3.0 patterns) are the field's actual methodology UX, yet almost none have first-class digital surfaces. The most cited "design tools" in this field are not software at all.
9. What This Synthesis Can't See
Honest gaps in coverage worth naming.
The twenty tracks are observation from outside. They infer reading patterns from public traces, citation patterns from syllabi and Substack recommendation graphs, and practitioner numbers from triangulation, not census. T20 says it directly: confidence is medium on rotation patterns, lower on relative weights between authors, lowest on the AI-builder audience whose reading habits are the hardest to surface from outside. This applies to the synthesis too. Where a track named precise numbers (Mondragon's 70,085 employees in 2024, Patagonia's $180M cumulative to Holdfast since 2022, Are.na's 18K paying members at November 2025) the numbers are sourced. Where a track triangulated practitioner counts (the under-100 serious specialists in steward-ownership advisory globally, the IPAI core circle of ~40 named people, the Mozilla Fellows at 10 globally per year) the triangulation is best-effort, not definitive.
Several specific blind spots. The Latin American organizational-design tradition, Buurtzorg-inspired, ToP/ICA International, is named in T01 and not researched in depth in any track. The same goes for the Indian self-help-group governance lineage and African ubuntu-based decision practices, named in T12 and not surveyed. The conscious-tech builder world outside the North Atlantic, small São Paulo / Mexico City strands around Indigenous data sovereignty named in T03, the responsible-AI-for-development crowd in Nairobi and Bengaluru, surfaces as a reference but not as a research target. Reading lists across all three audiences skew Anglo-American with Australian Aboriginal exceptions (T20 names the Escobar/Mbembe/Glissant/Wynter absence directly).
What the synthesis can't see about its own framing: it is reading the field from the lens of someone trying to map it, which is itself a particular position inside the field. Practitioners inside any single methodology will read this and find their tribe rendered too thinly. The five-bubble framing of tech, the five-cluster framing of ecological, the multi-lineage framing of governance, each of these is a synthesis-level abstraction that practitioners inside any single bubble may experience as flattening. The trade is necessary for the mapping to be possible, and worth naming as a trade.
The agentic research commons attempts named as private/pre-launch/non-indexed (Ostara, Localism Labs Bridging Worlds, Possible Planet Lab) cannot be researched from web alone. If they exist they are doing something the synthesis can't see. T16 flags this directly.
The synthesis also can't see the next twelve months of vocabulary capture. T17 estimates the regen-vocabulary slot is closing on a window comparable to where "sustainable" was in 2008-2010, but the actual rate of capture by Movate-style enterprise framings versus IT for Change-style commons framings is not predictable from current data. Same for the cyberecology slot. The 18-24 month window noted in CURRENT.md is real and the rate is unknowable.
10. The Real Open Questions
Questions about the FIELD that the next round of writing or research might answer.
Will "regenerative" survive as precise vocabulary, or will the field migrate to compound forms (just-and-regenerative, kincentric regeneration) or to entirely different vocabulary (living systems, relational, more-than-human)? T17's three options for serious holders are visible in the field but none has won. Wahl is holding the term and defining it every use; Forum for the Future has added "just"; the design-research literature is migrating toward kincentric and place-based vocabulary. Living Building Challenge keeps the rigor at the cost of staying small. The next 18-24 months probably decide this. The question that follows: when the term breaks, what name do the practitioners who hold the depth use?
Will the AI-in-governance experiments mature into useful infrastructure or fragment? vTaiwan/Polis, Collective Intelligence Project's Alignment Assemblies, OrgaLog and Synchrony, Gitcoin's Steward Council, the DAO-AI arXiv research, these are five live experiments in 2026 with no shared evaluation framework and no carrier for honest cross-experiment learning. T04 names this as a frontier; whether it becomes a maturing field or a list of attempts is unknown.
Where will the displaced Wellspring grant capacity actually land? T20 and CURRENT.md both surface that Wellspring Philanthropic Fund is winding down by 2028. The existing patient-capital networks (RSF, Slow Money) are demographically aging and not trivially expandable. The question of who funds patient capital in 2035, named in T09 as "largely unaddressed in the public literature," is real. The succession question for funders is the funder analogue to the succession question for founders.
What replaces Schumacher as the integrative residential learning container? T15 names that nothing in 2026 has filled the institutional gap left by Schumacher's August 2024 closure and OuiShare's 2024 wind-down. Ecoversities Alliance is the strongest educational cross-cluster network operating but explicitly outside tech. Edge Esmeralda is the most experimental popup-village model but tech-centric in gravity. Whether the field needs a residential container at all, or whether the cross-cluster work happens at meta-altitude (writing, federated commons) rather than in residential form, is genuinely contested.
Can the "single-author commons" category be defined cleanly enough to be discussable? T16 names this as a category that exists in practice (LessWrong-as-Yudkowsky, Slate Star Codex, Marginal Revolution, Stratechery) but lacks theory. The questions of what makes such a thing sustainable, what kills it, what it owes its readers, how it survives the substitution problem (Wikipedia losing 8% of human visitors after improving bot detection in May 2025), these are open in 2026 and the answers shape what becomes possible at this scale.
Will the convergence on relation-as-unit get a shared name, or will the methodology-tribes continue to rediscover the move in their own dialects? T08's central finding is that the move is happening across S3, permaculture, multi-agent AI, Bateson, Indigenous traditions, IIT, Active Inference, and the absence of a shared name is preventing the convergence from being recognized. The field is unified at ontology, fractured at vocabulary. Whether one of the candidate names (cyberecology, symmathesy, kincentric, relational ontology, ecology of intelligence) wins, or whether the convergence stays distributed across vocabularies, will probably be decided by which of these names a serious writer uses with discipline over the next several years.
What does "right relationship" with Indigenous knowledge look like at the operational layer of design organizations? T11's most uncomfortable open question. Most engagement asks permission rhetorically and proceeds. The shuumi-tax-style mechanism is replicable; the operating-budget question (what is the studio-level analog of the shuumi land tax) is rarely addressed. The deeper question, whether non-Indigenous practitioners should be using "regenerative" at all when "kincentric" or specific tradition-rooted terms exist, is asked in design-research literature, barely asked in 2026 mainstream regenerative-business literature.
Is the Bateson-Friston bridge real, or two parallel languages that look similar from outside? T08 and T10 both surface this. Friston's Designing Ecosystems of Intelligence from First Principles reads as cyberecology in mathematical clothing; Bateson's symmathesy and warm data work names what is lost when relationality is reduced to procedure. The two communities barely read each other. Whether the convergence is genuine ontological alignment or surface resemblance is empirically open and will be answered by whoever does the careful translation work.
What survives the AI substitution problem for small intellectual sites? T16 names this as the cultural blind spot the Liminal Web cluster has not, as of mid-2026, publicly grappled with. Wikipedia is at least public about the "AI ate our traffic" crisis. Substack writers are not talking about it. The defensive posture (authored voice that doesn't compress well, relational/contextual material that loses meaning when extracted, explicit licensing posture, building reader-relationship as the value) is articulable but not yet practiced at the field level.
The honest meta-observation. None of these questions are answerable from inside any single methodology. They are field-level questions, and the field-level vantage is exactly what's missing across the twenty tracks. That is the most useful single finding the synthesis can offer: the questions have to be asked from outside any tribe, and there is no tribe whose perspective the asking belongs to.