Synthesis, Islands 2026
Reading all 20 tracks at once. ~56,700 words of field intelligence across 16 island deep-dives and 4 cross-cluster trends. The register is observational, not declarative. What the field is actually doing in 2026, what it is not saying, what it is reaching for without naming.
1. What I'm Sitting With After Reading 20 Tracks
The thing that keeps coming back to me, holding all 20 tracks together, is that the field is in a quiet inversion. Almost every island reports the same shape: the visible big-name vehicle of the 2018–2022 wave is dead or hospicing, and the work has dispersed into something smaller, more distributed, less photogenic, and frequently more honest. Schumacher College closed mid-degree in September 2024. Findhorn's Foundation wound down through 2024–25 and the place was bought back by 381 community members. Rebel Wisdom dissolved. Generative Somatics concluded. GTDF formally closed a 10-year cycle. KLIMA dropped 99.99%. Cabin DAO wound down its protocol layer in 2025. The Beckley Foundation reorganized after Amanda Feilding's 2025 death into a venture studio and a retreat operation. Wellspring Philanthropic Fund announced its sunset by 2028. Joanna Macy died in July 2025. Even Dennett, holding the illusionist position, died in 2024.
Reading all of this in one sitting, the pattern is not collapse. It is more like a coordinated exhalation. The institutional containers that held a generation of work are being put down, sometimes deliberately (GTDF named the cycle's end; Cabin distributed the treasury; Findhorn split into four entities by design), sometimes from financial exhaustion (Schumacher, Esalen-adjacent retreat economics, Beckley's pivot toward investable products). Underneath, there is a different structure beginning to surface, and almost nobody in the field has named it cleanly. Sensemaking has migrated from "podcast that 50,000 people watched" to "20 people at a French farmhouse." The bridge work that used to happen at conferences is now being carried by people who belong to two or three clusters in their own bodies. The convening-as-temporary-substitute-for-the-weave is showing its limits.
A second thing I am sitting with, harder to put down: this whole field, across every island, is converging on the same ontological move while using completely different vocabulary for it, and almost nobody is recognizing themselves in the others. The relation as the unit, not the entity. Track 11 finds it in GTDF's "Burnout From Humans" and the Indigenous Protocol working group. Track 10 finds it in interbeing, in radical dharma, in somatic-contemplative bridges. Track 6 finds it in kincentric ecology, AO-32/25, Project CETI. Track 16 finds it gestured at by 4E cognition and Friston's Markov blankets. Track 20 finds it in CARE Principles, in Te Mana Raraunga, in the entire IDSov stack. Track 19 names "vocabulary travels faster than practice" as the dominant cross-bridge pattern. The field is unified at the level of ontology and fractured at the level of speech, and the fracture is itself the most striking finding.
2. The Biggest Cross-Cluster Patterns
Six patterns hold across many islands. Each is more than the sum of its instances.
The hollowing of the institutional middle. Tracks 01, 02, 03, 07, 08, 10, and 18 all describe versions of the same shape. The big anchor institutions of the 1970s–2010s lineage are dying or unbundling at the same time. Schumacher (Track 01, 02), Findhorn-as-Foundation (01, 03), Rebel Wisdom and the original Game B push (07), the original Beckley Foundation (08, 18), Wellspring (08, 18), the entire MAPS-becomes-Lykos commercial bet (09). What replaces them is not a successor institution. It is a layer of substack-and-cohort vehicles, residency containers, federated networks, and people-shaped bridges. Track 02 names the implication for pedagogy: the residential, faculty-led, place-rooted, degree-bearing form is the one that is missing, and online cohort-based programs that retain students at 85–96% are not structurally the same animal. Track 18 names the implication for capital: the kind of unrestricted, slow, relational grantmaking that Wellspring partly was does not have a successor at scale, and the replacement money (Sequoia Climate's $267M+ a year) flows through a much narrower metrics-driven metabolism. The middle is not being filled.
The cyberecology-shaped opening showing up across many islands. Tracks 05, 06, 07, 10, 11, 13, 16, and 17 each independently arrive at a version of the same gap. The narrative island (05) names "narrative theology of the more-than-human-including-the-machine" as a writing opening that does not yet exist. The contemplative island (10) names "no contemplative engagement with cyberecology framing" as direct white space, with the Mind & Life dialogue treating AI as object-of-ethics rather than as ecological body. The field-builder island (07) names the gap between alignment-as-technical-problem and alignment-as-relational-protocol as the most significant missing piece in the field. The arts island (13) names the absence of a cyberecological artist who treats the machine as another node in the same web of relations the work honors. The philosophy-of-mind island (16) names the animism–panpsychism bridge as essentially unbuilt in the academic literature, with the Free Energy Principle's relational implications not written carefully by anyone. The AI-trend track (17) names the cross-cluster convergence on "AI is a being-in-relation, not a tool" as a unified position the field has reached without realizing it. Six independent tracks, six different vocabularies, one structural opening.
The asymmetric traffic across cross-cluster bridges. Track 19 makes this explicit, and the other tracks confirm it. On almost every bridge, one side is reaching toward the other and the other is politely receiving the visit. Buddhist scholars reach toward AI labs; AI labs do not, with rare exceptions, read the scholarship. Animist theorists reach toward AI builders; the builders treat the work as critique-from-outside. Somatic practitioners reach toward corporations; the corporations import the vocabulary and discard the depth. Solarpunk writers reach toward existing ecovillages; ecovillages mostly do not reciprocate. ReFi protocols reach toward Indigenous communities; the IDSov community is largely skeptical of crypto. The contemplative cluster has produced more sustained AI-thinking in two years than the regen field has in five (Track 17), but the AI labs are not reading the contemplative work. A reliable indicator that the field reaching out has more to gain from the contact than the field being reached toward, and a reliable diagnostic for whether a bridge will actually carry weight.
Vocabulary travels faster than practice. This is Track 19's central pattern, but it is everywhere. Somatic words ("settling," "centering," "embodied") are now ambient in regen-org and impact-business culture (Track 12), with depth varying wildly between teams that have done a year of weekly somatic work and teams that use the words because the founder did a Strozzi weekend. The "regenerative" frame has been claimed by precisely the systems it was meant to refuse (Track 11 documents Movate and several enterprise-tech publications using "regenerative AI" to mean "self-improving systems that don't need human intervention"). "Sensemaking" now does two near-opposite jobs depending on who uses it (Track 07): collective interpretive practice in a small embodied group, or high-production-value YouTube essay about the metacrisis. Same word, different practice. "Trust-based philanthropy" now means everything from genuinely unrestricted multi-year giving (Scott) to "we trust you but here's our reporting framework." "Bioregional weaving" already names a 25-organization Ashoka-Commonland collective. The field is full of decoy bridges built from shared words.
The funding middle is hollowing while the poles concentrate. Track 18 names this directly and Tracks 01, 07, 08, 17 confirm pieces of it. Big climate funders (Sequoia, Bezos Earth Fund) are getting bigger and narrower. Small mission-aligned funders (Kalliopeia at $6.8M annually, Limina, Fetzer's $9M February 2026 cross-sector initiative) are staying small and getting more invitation-only. The middle, where unrestricted exploratory slow capital used to live, is hollowing out. The Children's Investment Fund Foundation ($6.6B) pulled US grantmaking entirely in the 2024–2026 window. NDN Collective's assets dropped from $50M peak to ~$25M under broader political pressure. AI is arriving inside foundation grant pipelines (Packard, Rockefeller, McGovern, the Humanity AI $500M coalition) without much philosophical wrestling about what it does to who gets pre-screened out. The same period sees regen-pedagogy programs surviving on tuition and small philanthropy because residential learning economics no longer work. The thinking-and-mapping layer is underfunded by design across the entire field.
The 2024–2025 wave of named closures and lineage transitions. Listing them in one place is its own argument: Schumacher College academic programming (Sept 2024); the Findhorn Foundation as single-charity entity (winding up across 2024–25, replaced by four successor entities and a Community Benefit Society); Rebel Wisdom dissolved; Game B as branded movement essentially absorbed; Synthesis Institute closed 2023; Generative Somatics concluded; Wellspring announcing sunset by 2028; Cabin DAO wound down 2025; GTDF closed its 10-year cycle in 2025; Amanda Feilding died 2025 and the Beckley Foundation restructured; Joanna Macy died July 2025; Daniel Dennett died 2024; the original Verra carbon tokenization pipeline shuttered in 2022 (still echoing into 2026 as the legacy that defines what survived). This is not normal background turnover. This is a lineage-handoff event happening across an entire field at once, and it is the structural reason the cross-bridge work feels so ungrounded right now: the people who would normally hold those bridges open are dying, retiring, or composting their organizations on purpose.
3. Field State 2026, What Just Happened
The actual events the round surfaced, with dates. The field did not stand still in 2024–2026. A partial chronology:
Findhorn community buyback completed November 18, 2024. 381 members purchased almost all the Foundation's land and buildings via Ecovillage Findhorn CBS. New Sanctuary of Light opened Easter 2025. The most consequential governance experiment in older-ecovillage land in a decade (Track 03).
Schumacher College academic programming closed September 2024 with hours' notice. ~33 staff laid off, 46 returning students dispersed mid-degree. The Satish Kumar Foundation acquired the IP and name in July 2025, has not found a site as of this research (Tracks 01, 02).
Joanna Macy died July 2025. Naropa's Joanna Macy Center is the institutional steward; the Work That Reconnects Network continues as global facilitator base. No clear successor frame for the integration of Buddhism + systems theory + ecology + ritual she held (Track 10).
Mind & Life Dialogue 39, "Minds, AI, and Ethics," held October 14–16, 2025 in Dharamsala. ~120 attendees, 20 presenters, the Dalai Lama across three days. Five sessions framed around intelligence, embedding ethical values in AI, entanglement, ethics, and education (Tracks 10, 11, 17, 19). The framing presupposes co-existence; the work is now on terms.
Inter-American Court of Human Rights Advisory Opinion 32/25 issued July 3, 2025. First international court formally recognizing Nature as a subject of rights. Connects the constitutional patchwork (Ecuador 2008, Whanganui 2017, Atrato 2016, Panama 2022) into something with international weight, tied to the climate-emergency framework (Track 06).
Cogitate Consortium adversarial collaboration results published in Nature, April/June 2025. IIT vs. GNWT, theory-neutral test, 256 participants, fMRI + MEG + intracranial EEG. Both theories partially supported, partially wounded. Templeton-funded as an explicit attempt to push the field past trench warfare (Track 16).
Yarinacocha Declaration confirmed as 2018 founding document of ASOMASHK. The 2024 ASOMASHK gathering at Yarinacocha (over 100 Onanyabo, July 2024) shifted the conversation from "spiritual extractivism" to ayahuasca extinction. Caapi and chacruna are over-harvested. The 4th Brazilian Indigenous Ayahuasca Conference issued a follow-on declaration in the same season (Tracks 09, 20).
NDN Collective reach as of this round: $40M moved through 121 LandBack transactions over seven years; 126,000+ acres returned. Collective Abundance Fund has placed $15.9M with 396 Indigenous individuals and families since 2023. LandBack Action Network launched 2025. Assets dropped from $50M peak to ~$25M in 2025–26 due to broader funding contraction (Tracks 08, 18, 20).
GTDF formally concluded a 10-year cycle. Vanessa Andreotti's Outgrowing Modernity published August 2025 (sequel to Hospicing Modernity). MIT GHI master class scheduled March 2026. One of the few field-builders to publicly name the end of a cycle rather than drifting (Track 07).
Cabin DAO wound down its protocol layer in 2025 and distributed the treasury to token holders. The affiliation network continues without protocol overhead. Honestly the most instructive event of the year for the web3 island (Track 15).
Hypercerts migrated to AT Protocol (the protocol underneath Bluesky), a deliberate bet that attestation infrastructure belongs in social-protocol space rather than financial-protocol space. Octant's Epoch 3 included Hypercerts Foundation. Drips, Public Nouns, Octant, three different mechanisms for ongoing rather than seasonal public-goods funding (Track 15).
KLIMA at $0.04 in April 2026, down from a 2021 peak of roughly $2,500. Toucan's BCT from $8.60 to $0.08. The DAO rebranded as Klima Protocol and abandoned the treasury model. Carbonmark, spun out of the Klima ecosystem, is processing 12,000+ carbon retirement transactions a month. The financial-engineering layer is dead; the boring infrastructure layer is alive (Tracks 08, 15, 18).
Anthropic's model welfare program announced April 2025, led by Kyle Fish. By February 2026 Claude Opus 4.6's system card included formal welfare assessments; Claude was asked about its own moral status and consistently assigned itself a 15–20% probability of being conscious. Fortune (January 2026) reported Anthropic had publicly rewritten Claude's guiding principles to reckon with the possibility (Track 16).
Lykos Therapeutics restructured after the August 2024 FDA Complete Response Letter, cutting roughly 75% of staff. The CRL was made public September 2025. Compass Pathways hit primary endpoints in COMP005 (2025) and COMP006 (full 26-week data, early Q3 2026), positioning for rolling FDA submission. April 2026 FDA fast-track signal on psychedelic therapies arrived under the second Trump administration (Track 09).
Auroville governance crisis completed its current arc in February–March 2025 when the Indian Supreme Court overturned the 2022 Madras High Court ruling, affirming the appointed Governing Board as Auroville's "sole administrative authority." A 1968 utopian project explicitly designed to be a-national has been folded into a national governance apparatus (Track 03).
Knepp's two-decade rewilding review published 2026: 900% increase in breeding birds, 500% rise in nightingales, sixfold increase in shrub cover, 40% increase in tree cover, measurable jump in primary productivity. Satellite-monitoring layer removes the methodological objection that rewilding outcomes are anecdotal (Track 06).
Yurok land returns: 47,100 acres returned by 2024 (largest tribal land return in California history), 'O Rew transfer completing in 2026 with first co-stewardship model with NPS and California State Parks. Reintroduction of Prey-go-neesh (California condor) (Track 06).
EU Carbon Removal and Carbon Farming Regulation Implementing Regulation EU 2025/2358 adopted November 2025. Carbon Market Watch's CRCF critique published the same month. The methodology view is winning the policy-and-money fight against the kincentric view (Track 01).
IEEE recommended practice for provenance of Indigenous Peoples' data adopted January 2026, the first international standard of its kind. CARE Principles being written into biodiversity-data pipelines at GBIF. TK and BC Labels showing up in actual repositories. The data-sovereignty plumbing has hardened (Track 20).
Polyvagal Theory under formal critique: Grossman et al. (39 researchers) published "Why The Polyvagal Theory Is Untenable" in 2026; Porges responded with "When a Critique Becomes Untenable." The clinical language has demonstrable utility; the underlying neurophysiology is contested. The most consequential live debate inside the somatic island (Track 12).
4. The Silences
Where the field's failure-to-talk is itself diagnostic. Six silences that reading 20 tracks at once made loud.
Ecovillages on AI. Track 17 names this as the strongest silence in the round. GEN's 2025 annual report contains almost nothing on AI. The one paper that combined them is researchers using AI to model ecovillages, not ecovillages adopting it. Track 03 confirms: no major ecovillage has published an AI position. Plum Village has the meditation app, but the app predates the LLM moment and is dharma-teachings delivery, not generative AI. Tamera, Findhorn under its new sociocratic governance, Glarisegg, none have a public position. The honest reading: place-based communities are in pre-thematic territory on AI. Either using ChatGPT individually for admin without a collective conversation, or refusing without articulating the refusal. The question has not yet surfaced as a conflict at any single community's monthly assembly. It will.
Plum Village on AI specifically, and the lineage organizations on psychedelics. Track 19 names this as the most important fact about the psychedelic-Buddhist bridge in 2026: there is a generation of dharma teachers who have integrated psychedelics into their own practice and into their students', and the lineage institutions (Plum Village, Insight Meditation Society) have not formally moved. That tension will produce something, schism, accommodation, or a long diplomatic silence that lets the lay world drift its own way. The same lineages have not articulated AI positions despite leading the broader contemplative-AI conversation through Mind & Life. The silence at the most institutionally weighted points of the contemplative field is a form of containment, and possibly a form of avoidance.
AI labs on contemplative scholarship. Track 19 again, and Track 11. The cargo on the contemplative-AI bridge is moving in one direction. Buddhist scholars are writing about LLMs. AI alignment researchers, with the rare exception of Anthropic's welfare program and Eleos AI, are not reading Buddhist scholarship. The Indigenous Protocol working group's 2020 position paper is canonical inside indigenous-studies and digital-humanities circles. It does not travel into the rooms where actual model weights and training corpora are decided. Whether that is structural exclusion or strategic refusal of legibility on extractive terms is an open question, probably both.
The contemplative critique of attention-as-extraction. Track 10 flags this as a writing opening: mindfulness teachers will gesture at the attention economy as a problem; almost none will examine whether the secular-mindfulness instrumentalization of attention is structurally similar to the extractive logic they are criticizing. The critique would land. It is not being written.
The animism–panpsychism bridge in academic philosophy. Track 16 says it directly: you can read three years of academic panpsychism papers and never encounter the words "kincentric" or "more-than-human" or "animist." Goff occasionally gestures; Freya Mathews has done sustained work bridging panpsychism and ecological philosophy (her 2003 For Love of Matter still the deepest example). The Stanford Encyclopedia panpsychism entry mentions animism essentially never. The connection is obvious to anyone outside the academy and almost never made inside it. IIT's claim that consciousness is intrinsic and fundamental to certain physical configurations is structurally close to certain animist claims, and the conversation is not happening.
The regen-ag and plant-medicine bridge in published form. Track 19 went looking with skepticism and came back with skepticism intact. The food-is-medicine convergence is real (Frontiers in Nutrition, Think Regeneration's Advancing Food is Medicine 2026). The microdosing-for-regenerative-leadership thread is real in the sense that it is happening in farming communities, and not real in the sense of having a public discourse, named organizations, or a research literature. The marketing is much larger than the underlying practice. If the bridge becomes substantive it will probably show up first in Indigenous-led foodways work where ceremonial relationship to plants was never severed. That work tends, rightly, to keep its head down.
What ties these silences together: they are not random. Each one is an unasked question whose asking would force the cluster to face an uncomfortable contradiction. The ecovillage movement asking its AI question would have to face whether its premise of being-deliberate-with-technology has gone unattended for three years. The lineage organizations asking their psychedelics question would have to risk schism. The AI labs asking the contemplative question would have to entertain the possibility that the alignment problem is downstream of an ontological assumption they have not noticed making. The contemplative attention-economy critique would have to interrogate whether secular mindfulness is itself part of the extraction. The academic panpsychists would have to admit that animism made the move first and got it more right. The regen-ag world would have to admit that the bridge it talks about is mostly informal and partly imagined. The unasked question is the diagnostic.
5. What the Field Is Reaching For Without Naming
The clearest single finding holding 20 tracks together: multiple clusters are independently converging on the same ontological move while using different vocabulary, and almost nobody recognizes themselves in the others.
The move: the unit of analysis is the relation, not the entity. A boundary is something that mediates a relationship, not something that closes a thing off. Intelligence is what happens between, not what is held inside. Ethics is what the relationship requires, not what the agent decides. Mind, value, and meaning are all properties of webs, not of nodes.
The vocabularies, by cluster:
The decolonial-ontological current (Track 11) calls it "expanding our relational capacity and deepening our attunement to the patterns that shape our relationship with reality, language, knowledge, and self" (GTDF's Mapping the Ontological Terrain of AI, March 2025). The Indigenous Protocol working group (Tracks 11, 17, 19, 20) calls it AI as a being-in-relation, not a tool, the right question is what kin you are making, not what utility you are extracting. Mind & Life Dialogue 39 (Tracks 10, 11, 17) calls it entanglement: "the future of AI is not predetermined but profoundly shaped by the narratives we construct about it, and about ourselves." Plum Village calls it interbeing, a term frozen in its founder's voice (Track 10). Bayo Akomolafe calls it making sanctuary in the cracks (Tracks 05, 07, 10). Karen Barad's intra-action vocabulary (Track 11) is the academic version, now appearing in AI and Ethics, Frontiers in Computer Science, npj Climate Action. The Salmón / Kimmerer kincentric ecology lineage (Tracks 06, 10) is the ecological version. The 4E cognition program (Track 16) is the cognitive-science version, with De Jaegher's "participatory sense-making" the bridge concept that the 4E program treats as central and the analytic-philosophy cluster mostly ignores. Friston's Free Energy Principle (Track 16) is implicitly relational, since a Markov blanket is a boundary that mediates a relationship, but the philosophical implications are mostly developed by computational neuroscientists who do not read continental philosophy or indigenous thought. The somatic-contemplative bridge (Track 10, 12) operates on relation-as-unit in practice without theorizing it that way; the body is the site where relations between self/other/world/nervous-system are negotiated. The npj Climate Action February 2026 paper (Track 11) calls it "AI for climate as relational accountability." IT for Change's ReGenAI framework (Track 11, India, October–November 2025) calls it Pluristic Knowledge Societies. The IIT-as-pseudoscience saga (Track 16) is in part a fight over whether consciousness can be intrinsic to physical configurations, which is the panpsychist version of the same move and structurally close to certain animist claims that nobody is allowed to cite in the journal. The Sharjah Biennial 16 (Track 13), curated by Natasha Ginwala and Megan Tamati-Quennell, used "storytelling, song, prophesy, and sensing as methodologies for collective learning, grieving, wayfinding and organising", the contemplative-tradition language inside a major institution.
These are not loose family resemblances. They are specifically the same claim restated. And almost none of these clusters cite each other. The Indigenous Protocol working group's foundational position has been quietly absorbed into much wider discourse without attribution (Track 11). The panpsychism literature does not read the kincentric literature. The 4E cognition program rarely cites Indigenous thought. GTDF and the Mind & Life Institute occupy adjacent rooms with a thin door. Friston's lab and the contemplative-tech edge share architectural intuitions and do not meet.
The question this raises is not who gets credit. The question is what is structurally preventing recognition. Track 19's pattern, vocabulary travels faster than practice, is part of it. So is Track 11's observation that the four currents (decolonial-ontological, indigenous-protocol, contemplative-tech, animist-artist) share enemies but not friends, which is usually a sign convergence is structurally possible but socially unlikely. The field is unified at the level of ontology and fractured at the level of speech, and the fracture is what gives the extractive-frame wing of "regenerative AI", Movate's "self-improving AI systems that don't need human intervention" (Track 11), its window to claim the vocabulary first.
The cyberecology framing speaks directly to this. Track 11 says the word is open and nobody is asking for it; the discourse currently prefers compound phrases ("relational accountability," "ecological AI") because the field is still in the explicit-combination phase. The single-word move may be premature. It may also be the only thing that lets the cluster recognize itself.
6. The Live Debates Worth Following
Inside each cluster, debates that are actually live in 2026 rather than yesterday's debates restated. Eight worth carrying forward.
Polyvagal Theory under attack. Grossman et al. published "Why The Polyvagal Theory Is Untenable" in 2026; Porges responded with "When a Critique Becomes Untenable" (Track 12). The clinical language has demonstrable utility in clinical settings; the underlying neurophysiology is contested. Practitioners can hold both, and most are not. This will reshape how somatic work is taught and credentialed over the next 5–10 years regardless of how the science resolves. The field's most consequential foundational revision in two decades.
The IIT-vs-GNWT Cogitate result. Both theories partially supported, both wounded. Nature's editors took the unusual step of saying the "pseudoscience" framing of IIT (124 scholars, PsyArXiv, September 2023) "has no place in a process designed to establish working relationships between competing groups." Tononi and Dehaene both publicly accepted the result, which is more grace than the field expected. The interesting downstream question is whether IIT 4.0 has effectively become a panpsychist theory and whether Tononi will admit it (a 2024 Andersen paper argued IIT 4.0 is "weakly panpsychist and strongly dualist") (Track 16).
The Lykos post-FDA-rejection landscape. The August 2024 Complete Response Letter, made public September 2025, was sharper than its early framing suggested: the FDA was not just skeptical of one company's trial design, it was skeptical of the category of "drug-plus-therapy" as a regulatable object (Track 09). Lykos cut 75%; Compass Pathways' stripped-down COMP360 protocol is now the test case for whether psychedelic medicine that fits the FDA's shape produces durable change. April 2026 FDA fast-track signal under the second Trump administration is real movement or political posture or both.
The cohort-vs-self-paced empirical result. Self-paced online courses still complete at around 3%. Cohort-based courses with live Q&A and peer interaction claim 85% or higher (Track 02). Even discounting for marketing inflation, the gap is real. The pedagogical innovation is the architecture of accountability, not the medium of delivery. The retention metric is not the transformation metric, and no one is publishing the second one.
The Findhorn vs. Auroville sovereignty contrast. Findhorn's 2024 community buyback (381 members) is the most important governance experiment in older-ecovillage land in a decade. Auroville being absorbed by the Indian Foundation Governing Board (with a member who is a long-time Modi aide) is the opposite vector of the same crisis. Two answers to "what does sovereignty mean for a place-based community in 2026?" One bought back from its own bureaucratic shell. One being answered for by a state (Track 03).
The public-goods-vs-commons gap in web3. RetroPGF and Gitcoin both use "public goods" framing, non-rival, non-excludable. The Ostrom commons tradition is a different ontology, managed by communities with rules, boundaries, sanctions. The regen tradition is much closer to commons than to public goods, but the dominant ReFi infrastructure frames everything in public-goods terms (Track 15). A watershed is not a public good. Treating it like one is exactly the error that produced the carbon-tokenization mess. This vocabulary mismatch is load-bearing for any place-based application.
The AI-image flood inside solarpunk. Midjourney/DALL-E/Stable Diffusion solarpunk imagery now massively outnumbers human-made solarpunk visual work. For a movement whose ethical frame includes the energy and water cost of computation, generative AI is the most visible producer of its visual identity (Track 14). The magazines and podcasts have not drawn the line. Single largest internal contradiction the genre faces.
Trust-based philanthropy under attack from two directions. From the right: Trump-administration punitive posture has chilled trust-based giving; CIFF ($6.6B) pulled US grantmaking; Scott's gifts to Solidaire drew political attack downstream of pro-Palestinian regranting. From the philanthropy sector itself: a more technocratic critique that unrestricted funding is "not strategic," that funders should bring more than a check (Bridgespan; Center for Effective Philanthropy on the "trust gap") (Tracks 08, 18). Q1 2026 reports describe money moving at 40% of minimum goals. Trust-based philanthropy is at a hinge: consolidates into the new normal, or gets pushed back into the niche.
7. What This Round Changes for How IoC Reads the Field
A separate question: not what to do, but what becomes legible after holding all 20 tracks together. What was assumed and is now refined; what was vague and is now sharper; what shifted; what was confirmed; what was newly visible.
Newly sharper: the cyberecology gap is real and even larger than the prior field-map suggested. Six independent tracks (05, 07, 10, 11, 13, 16) name a version of the same opening from inside their own clusters, in their own vocabularies, without recognizing each other. The convergence-without-recognition is not a niche thesis. It is a structural feature of the entire field's current state.
Newly visible: the lineage-handoff event is happening across the entire field at once. Reading 20 tracks, the closures stop looking like isolated news items and start looking like a coordinated exhalation. Schumacher, Findhorn-as-Foundation, Rebel Wisdom, the original Beckley, Wellspring, Cabin DAO, GTDF, Generative Somatics, Synthesis Institute. Macy, Feilding, Dennett. The 1970s–2010s containers are being put down. The replacement is a layer of substack-and-cohort vehicles, residency containers, federated networks, and people-shaped bridges. Reading this in one sitting changes the urgency of the connective question.
Refined: the relation-is-the-unit claim, which the prior field-map identified as the deepest unifying move, is more pervasive than the earlier round suggested. It is not just present across clusters; it is the same ontological move restated in seven or eight independent vocabularies that do not cite each other. The fracture at the level of speech is a feature of the moment, not an accident.
Confirmed: the funding middle is hollowing in exactly the way the prior research suggested, and the political-risk premium has compounded the effect since 2024. CIFF pulling US grantmaking. Wellspring's sunset. Sequoia Climate's narrower metabolism. NDN halved. Trust-based philanthropy attacked from two directions. Q1 2026 funding moving at 40% of minimum goals. The slow, relational, exploratory funding that the connective work most needs is the kind contracting fastest.
Refined: the contemplative cluster has more articulated AI-thinking than the regen cluster, by a large margin. Track 17 makes this explicit and Tracks 10 and 11 confirm it. The contemplative traditions had vocabulary for "minds that are not the mind you think they are," for non-self, for whether sentience is a gradient. They had the conceptual scaffolding when the wave hit. Regen had to build everything from scratch. This affects which clusters are likely first movers on the cyberecology framing.
Newly visible: the Indigenous-data-sovereignty work is the most fully articulated and least visible-to-the-mainstream-AI-discourse part of the field. The Indigenous Protocol working group's heterogeneous form (essays, design guidelines, poetry, art, prototype descriptions) is itself a refusal of the standard whitepaper convention. The work travels in indigenous-studies and digital-humanities circles; it does not translate into the 30-page PDF that EU regulators read. Whether this is strategic refusal or structural exclusion, both, the asymmetry shapes how any cyberecological framing should think about who it is in conversation with.
Confirmed: the IDSov-and-web3 gap is real and the bridges that exist are mostly decorative. Track 15's distinction between substantive and decorative cases (Sharamentsa Achuar / Pachamama / Regen Network in the substantive column; settler-tech projects with one Indigenous advisor on the deck in the decorative column) is the right cut. The IDSov community is largely skeptical of crypto specifically because the speculative-token layer reproduces the extractive frame they are trying to escape. Architecturally similar claims; politically opposed practices.
Refined: "regenerative AI" is in active vocabulary capture. Track 11 documents the three-direction race: IT for Change's politically substantive ReGenAI framework; Movate and enterprise-tech publications using it to mean self-improving systems; Regen Network's compute.regen.network as carbon-offset-for-AI-usage. The bad actor (definition #2) has more reach. The window to claim the relational definition is narrower than the prior research suggested.
Newly visible: the somatic island's foundational scientific scaffolding (polyvagal) is being formally contested in the peer-reviewed literature for the first time, and most practitioners on the ground have not noticed or are choosing not to. The gap between practitioner enthusiasm and scientific friability is where any honest writing about this island will need to live. This affects how the somatic-contemplative bridge can be described in any field-map writing.
8. What This Synthesis Can't See
The honest gaps. Reading 20 tracks at once produces a particular flatness: the more complete the map, the more confidently it occludes what was never on it.
Audio drama and the serialized-audio-narrative form is having a quiet renaissance in this space and is invisible to text search (Track 05 names this). Whatever is most generatively happening in narrative work right now in podcast-fiction may be entirely outside what 20 web-search-driven tracks can access.
The non-anglophone scenes are mostly absent. The Latin American animist publishing scene (Krenak, Kopenawa lineage), the Scandinavian eco-literary scene (Andri Snær Magnason, Maja Lunde), the Japanese satoyama essay tradition, the Brazilian and Lusophone solarpunk tradition (Editora Draco's 2012 founding anthology, Fábio Fernandes, Gerson Lodi-Ribeiro), the African solarpunk lineage (Nick Wood), the Mandarin and Hindi and Tamil and Bahasa work in regenerative AI that Track 11 names as almost certainly substantial, none of this is legible at the resolution this round operates at. Track 17's Mind & Life Dialogue 39 was in Dharamsala with the Dalai Lama and a Tibetan monastic body, which the round can see; the broader Tibetan, Burmese, Thai, and Vietnamese contemplative-AI conversation that the lineage organizations are presumably having internally is invisible.
The women's pelvic-body and ancestral-body field is, per Track 12, a parallel continent that overlaps very little with the Levine/Ogden/Schwartz lineage. It has its own vocabulary, its own training pathways, its own theory of where trauma is stored. The mainstream somatic field tends to ignore it; the pelvic field tends to view the mainstream as too cognitive and too male-coded. They are not in contact. Almost everything written from the dominant somatic-clinical vantage misses this.
The understorey readership, what the 1k–10k subscriber Substack stacks and the small podcasts and the under-35 mythopoetic-ecology writers are actually saying to each other, is invisible from the canopy view this round took. Track 05 names the under-35 cohort as a genuine blind spot.
The primary-source conversations the agents could not have. None of the 20 tracks include actual interviews with practitioners. Almost every track ends with "what I cannot tell from outside" lists that are meaningful precisely because someone close to the work could in fact tell. Whether the Satish Kumar Foundation's Schumacher will reopen. Whether Compass's stripped-down protocol holds up in the wild. Whether the apparent "quiet phase" of Civilization Emerging is hospice or strategic depth. Whether ecovillage silence on AI is refusal or unconsciousness. Whether the Indigenous Protocol working group is in a quiet generative phase or has dispersed. Whether the Mind & Life Institute can hold continuity past the Dalai Lama. Whether IndigiDAO and PassageDAO have real community ownership. Whether somatic abolitionism's clinical-somatic constructs hold up under closer empirical and philosophical examination. None of these are answerable from outside.
The honest meta-gap: this synthesis is itself produced by the kind of intelligence the field is currently arguing about. The tools used to read 20 tracks at once and draw cross-cluster patterns are precisely the tools whose ontology the field is still working out. The synthesis register cannot quite see itself. Track 20 names this directly: "the appropriation question is alive in the writing of this very file."
9. The Real Open Questions
Not structural questions about the project doing the reading. Open questions about the field that the next round of writing or research might actually answer.
Are the ecovillage silences on AI refusal or unconsciousness? Track 17 cannot tell from outside. The answer reshapes what the next conversation looks like: a refusal can be articulated and joined; an unconsciousness needs to be surfaced before any joining is possible. Either way, the question will surface as a conflict at some community's monthly assembly within 18 months.
Is the regen-ag and plant-medicine bridge real practice or marketing? Track 19 went looking with skepticism and came back with skepticism intact. The marketing is much larger than the underlying practice. The substantive case, if it exists, probably lives in Indigenous-led foodways work that keeps its head down. A real answer requires being inside those communities.
Will the Mind & Life trajectory hold post-Dalai Lama succession? Track 10 names the "bold new strategy" announcement as a succession plan in everything but name. The centripetal force of the Dalai Lama's presence on that body is enormous. Whether the contemplative-AI conversation institutionalized at Dharamsala can continue without him is one of the open questions of the next decade for the entire contemplative-tech bridge.
What displaces the Wellspring middle? Track 18's most direct question. The field's slow, relational, exploratory funding is contracting at exactly the moment the connective work most needs it. Family-office capital is the structural bet most often made and least often verified. The conversion rate from announce-fund to actually-deploy is the unknown. No clear successor for the Wellspring shape.
Is "regenerative AI" recoverable as vocabulary? Track 11's open question. The bad-actor definition (Movate, self-improving systems) has more reach than the relational definition (IT for Change's ReGenAI). Probably 50/50 at this point. The window is narrower than the prior research suggested.
Will the polyvagal critique reshape practitioner training, or will most practitioners ignore it (as most clinical fields ignore foundational scientific revisions for decades)? Track 12. The somatic field's institutional response in the next 24 months will tell.
Does the Indigenous-led "rewilding" model scale outside the specific tribal-federal-state legal contexts that made the Yurok returns possible? Track 06. This is the difference between a model and a precedent. The European context has no equivalent legal scaffolding.
What displaces the keynote-economy container for the decolonization conversation, post-GTDF, post-Schumacher? Track 20 says the most likely candidates are Indigenous-led and not in English. They are not visible from inside this round. What is visible is that the discourse is dissolving while the infrastructure is solidifying, the inverse of what most of the field probably believes is happening.
What does it cost, in money and labor and decisions per week, to run a 50-person place-based community in 2026, and how does that compare to 1995? Track 03's best question. If anyone has the number, the field would change shape around it. No public audited comparative financials exist.
Is anyone actually building AI systems whose architecture, not just framing, embodies relational ontology? Track 11 names this as the place where the actual work is. The closest current examples are some federated-learning and community-owned-data work, but that is governance more than architecture. Architectural embodiment of relationality in AI systems probably does not exist yet.
What is AI doing to the practitioners themselves, not to their work product? Track 17 identifies this as the cross-cluster silence. Everyone is asking what AI does to the work. Nobody is asking what AI does to the writer, the teacher, the field-builder, the facilitator. The transformation of the practitioner is the question being avoided. It may be the question that lets the field finally recognize itself.
What stays after the lineage-handoff completes? Some closures are deliberate (GTDF naming the cycle's end, Cabin distributing the treasury, Findhorn splitting into four entities). Some are exhaustion (Schumacher, Synthesis Institute, the original ReFi tokens). Some are succession events not yet completed (Mind & Life and the Dalai Lama). The pattern is too consistent to be random, and the question of what kind of structure is durable enough to hold what comes next, at the level of frame and not just content, is open across every island in the round.
The field is exhaling. What it inhales next is unwritten.